Nuclear Medicine Textbook: Editors

Duccio Volterrani
Paola Anna Erba

Ignasi Carrió
H. William Strauss
Giuliano Mariani
Editors
Nuclear Medicine
Textbook
Methodology and Clinical Applications
123
Nuclear Medicine Textbook
Duccio Volterrani • Paola Anna Erba
Ignasi Carrió • H. William Strauss
Giuliano Mariani
Editors
Nuclear Medicine Textbook

Methodology and Clinical Applications
Editors
Duccio Volterrani Paola Anna Erba
Regional Center of Nuclear Medicine Regional Center of Nuclear Medicine
Department of Translational Research and Department of Translational Research and
Advanced Technologies in Medicine and Surgery Advanced Technologies in Medicine and Surgery
University of Pisa University of Pisa
Pisa Pisa
Italy Italy
Ignasi Carrió H. William Strauss

Hosp. Sant Pau, Nuclear Medicine Department Molecular Imaging and Therapy Department
Autonomous University of Barcelona Cornell University Weill
Barcelona New York, NY
Spain USA
Giuliano Mariani
Regional Center of Nuclear Medicine
Department of Translational Research and
Advanced Technologies in Medicine and Surgery
University of Pisa
Pisa
Italy
ISBN 978-3-319-95563-6 ISBN 978-3-319-95564-3 (eBook)

https://doi.org/10.1007/978-3-319-95564-3
© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
There are few up-to-date books dealing with the Nuclear Medicine specialty as a discipline per
se, i.e., intended as complete textbooks providing the rationale for understanding the physical,
chemical, biological, and pathophysiological bases of all the applications of nuclear medicine
in the routine clinical practice, including diagnosis and therapy.
In addition to the expertise in therapy with radionuclides, the modern specialist in nuclear
medicine must be a specialist in multimodality imaging. This requirement is the guiding con-
cept throughout this multiauthor textbook (written by expert authors from Europe and the
United States), which presents the spectrum of information required to become a practitioner
of nuclear medicine and molecular imaging.
The text presents the basic science of radiation physics, interaction of radiation with bio-
logic systems, measuring devices for radiation detection and radionuclide imaging, synthesis
and biodistribution of radiopharmaceuticals, and approaches to data recording and analysis
that comprise the clinical applications of nuclear medicine in diagnosis and therapy. Each of
the 51 chapters in this 1300 page text contains numerous graphs, tables, diagrams, and patient
illustrations designed to meet the needs of students training for a career as practitioners, tech-
nologists, physicists, and/or data analysts.
Each chapter provides detailed learning objectives, comprehensive text, and visual material
to provide the reader with the information necessary to understand the topic. In addition to the
basic science and clinical material, chapters on Computed Tomography and Magnetic
Resonance Imaging provide foundation knowledge to employ these important complementary,
imaging techniques in hybrid imaging.
To supplement the basic science and clinical applications of radionuclide imaging, the
authors present current guidelines from international organizations on approved indications to
perform specific procedures, recommendations on appropriate techniques to perform and
interpret the procedures as well as recommendations from the International Commission on
Radiation Protection to perform these procedures with the lowest recommended radiation bur-
den to the patient and the clinical staff.
In addition, case presentations in the clinical chapters illustrate the specific application of the
radiopharmaceutical procedures. The clinical case presentations are supplemented by additional
patient material, presented in Chaps. 47 and 48. These cases allow the reader to combine the
clinical history, laboratory findings, radiographic and radionuclide findings to get acquainted
with the most typical presentations for diagnosis or therapeutic plan to manage the patient.
Developing this text has required the work of many people with different areas of expertise.
The authors and editors want to offer special thanks to the numerous individuals from our
publisher, Springer Publishing Co., for their work in typesetting, layout, and printing, to bring
this project to fruition.
Pisa, Italy Duccio Volterrani

Pisa, Italy Paola Anna Erba
Barcelona, Spain Ignasi Carrió
New York, NY H. William Strauss
Pisa, Italy Giuliano Mariani
April 2019
v
Contents
Part I Basic Science
1 Fundamentals of Natural and Artificial Radioactivity

and Interaction of Ionizing Radiations with Matter�� 3
Alberto Del Guerra and Daniele Panetta
2 Single-Photon-Emitting Radiopharmaceuticals�� 21
Federica Orsini, Erinda Puta, Federica Guidoccio, and Giuliano Mariani
3 Positron-Emitting Radiopharmaceuticals �� 57
Piero A. Salvadori, Elena Filidei, and Assuero Giorgetti
4 Radiopharmaceuticals for Therapy�� 99
Federica Orsini, Federica Guidoccio, and Giuliano Mariani
5 Methods and Instrumentation for Measuring Radioactivity�� 117
Maria Evelina Fantacci
6 General-Purpose Gamma Cameras, Dedicated Gamma Cameras,
and Gamma-Probes for Radioguided Surgery�� 137
Roberto Pani, Federica Guidoccio, Raffaele Scafè, Pat Zanzonico,
and Giuliano Mariani
7 Image Acquisition and Processing with Gamma Cameras
Including Integrated SPECT/CT and Dedicated Gamma Cameras�� 173
Duccio Volterrani, Federica Guidoccio, Giulia Puccini, and Sara Mazzarri
8 Principles of CT and MR imaging�� 187
Christian Bracco, Daniele Regge, Michele Stasi, Michela Gabelloni,
and Emanuele Neri
9 PET/CT and PET/MR Tomographs: Image Acquisition and Processing�� 199
Nicola Belcari, Ronald Boellaard, and Matteo Morrocchi
10 Essentials of Quantitative Imaging with PET �� 219
Adriaan A. Lammertsma
11 Principles of Radiation Biology and Dosimetry for Nuclear
Medicine Procedures�� 235
Massimo Salvatori, Amedeo Capotosti, and Luca Indovina
12 Radiation Protection for Patients �� 261
Lawrence T. Dauer, Sören Mattsson, Eliseo Vañó, and Yoshiharu Yonekura
13 Radiation Protection for Personnel and the Environment�� 273
Antonio C. Traino
14 Essentials of CT Image Interpretation�� 281
Davide Caramella, Matteo Revelli, and Alessandro Villa
vii
viii Contents
15 Essentials of MR Image Interpretation�� 317

Davide Caramella and Fabio Chiesa
16 Image-Guided and Radioguided Surgery�� 351
Francesco Giammarile, Sergi Vidal-Sicart, Federica Orsini,
Renato A. Valdés Olmos, and Giuliano Mariani
Part II Organ/Apparatus-Specific Applications
17 Radionuclide Imaging for Non-tumor Diseases of the Brain�� 391

Duccio Volterrani, Giampiero Giovacchini, and Andrea Ciarmiello
18 Hybrid Imaging for Tumors of the Brain�� 413
Giampiero Giovacchini, Mattia Riondato, Patrizia Lazzeri, Elisa Borsò,
Valerio Duce, Rossella Leoncini, Elisabetta Giovannini, and Andrea Ciarmiello
19 Hybrid Imaging of the Head and Neck Region �� 431
Alejandro Fernández and Valle Camacho
20 Radionuclide Imaging of Cardiovascular Disease�� 449
Matteo Bauckneht, Flavia Ticconi, Roberta Piva, Riemer H. J. A. Slart,
Alberto Nieri, Silvia Morbelli, Paola Anna Erba, Cecilia Marini,
H. William Strauss, and Gianmario Sambuceti
21 Radionuclide Imaging of Benign Pulmonary Diseases �� 499
Federica Guidoccio, Edoardo Airò, and Giuliano Mariani
22 Hybrid Imaging for Tumours of the Chest�� 523
Roberto C. Delgado Bolton and Adriana K. Calapaquí Terán
23 Hybrid Imaging for Breast Malignancies�� 543
Federica Padovano, Giuliano Mariani, and Marco Ferdeghini
24 Hybrid Imaging and Radionuclide Therapy of Musculoskeletal Diseases �� 571
Paola Anna Erba, Martina Sollini, Roberta Zanca, Roberto Boni, Lesley Flynt,
Elena Lazzeri, Giuliano Mariani, and Torsten Kuwert
25 Hybrid Imaging of Melanoma and Other Cutaneous Malignancies�� 645
Montserrat Estorch
26 Hybrid Imaging and Radionuclide Therapy in Hemato-oncology�� 655
Paola Anna Erba, Martina Sollini, Roberto Boni, and Sara Galimberti
27 Hybrid Imaging and Radionuclide Therapy for Thyroid Disorders�� 707
Federica Guidoccio, Gayane Aghakhanyan, and Mariano Grosso
28 Hybrid Imaging in Non-thyroidal Endocrinological Disorders�� 749
Duccio Volterrani, Federica Guidoccio, and Giuliano Mariani
29 Hybrid Imaging and Radionuclide Therapy of Neuroendocrine Tumors�� 767
Duccio Volterrani, Lisa Bodei, and Federica Guidoccio
30 Radionuclide Imaging of the Nephro-Urinary Tract�� 785
Duccio Volterrani, Federica Orsini, and Federica Guidoccio
31 Functional Imaging for Benign Conditions of the Gastrointestinal Tract�� 809
Gayane Aghakhanyan, Elisa Fiasconaro, Elisa Tardelli, Mariano Grosso, and
Italia Paglianiti
32 Hybrid Imaging for Malignant Conditions of the Gastrointestinal Tract�� 841
Joan Duch and Albert Flotats
Contents ix
33 Radionuclide Therapy for Tumors of the Liver and Biliary Tract�� 859
Federica Guidoccio, Giuseppe Boni, Duccio Volterrani, and Giuliano Mariani
34 Hybrid Imaging for Gynecologic Malignancies�� 881
Elisa Lodi Rizzini, Elena Tabacchi, and Cristina Nanni
35 Hybrid Imaging for Male Malignancies�� 899
Akram Al-Ibraheem, Abdullah S. Al Zreiqat, Serena Chiacchio, and
Abedallatif A. AlSharif
36 Nuclear Medicine in Pediatrics�� 925
Pietro Zucchetta and Diego De Palma
37 Hybrid Imaging of the Peripheral Lymphatic System �� 951
Paola Anna Erba, Roberto Boni, Martina Sollini, Andrea Marciano,
Rossella Di Stefano, and Giuliano Mariani
38 Molecular Guidance for Planning External Beam Radiation Therapy�� 977
Federica Orsini, Giovanna Pepe, Arturo Chiti, Giuseppe Roberto D’Agostino,
Annibale Versari, Carlo Cavedon, Marco Ferdeghini, Paola Anna Erba, and
Martina Sollini
Part III Practice and Procedures (with Supplemental Online Files)
39 Radiopharmacy/Radiochemistry for Conventional Single-Photon

Emitting and Therapeutic Radiopharmaceuticals�� 1009
Alessandra Boschi and Adriano Duatti
40 Radiopharmacy/Radiochemistry for Positron-Emitting
Radiopharmaceuticals, Including Quality Assurance and
Process Validation Principles�� 1019
Piero A. Salvadori
41 Recommendations for Conducting Clinical Trials with
Radiopharmaceuticals �� 1039
Clemens Decristoforo and Serge K. Lyashchenko
42 Standard Operating Procedures for Quality Control of
Gamma Cameras �� 1051
Angela Vaiano
43 Standard Operating Procedures for Quality Control of PET/CT
and PET/MR Tomographs�� 1061
Alessandra Zorz and Alessandro Scaggion
44 Standard Operating Procedures for Quality Control of
Instrumentation for Radioguided Surgery�� 1079
Angela Vaiano
45 Technical Aspects of Dual-Energy X-Ray Absorptiometry and
Quantitative Computed Tomography for Assessment of Bone
Mineral Density and Body Composition�� 1085
Pat Zanzonico
46 Current Practical Guidelines for the Most Common Nuclear
Medicine Procedures�� 1099
Irene Marini, Onelio Geatti, and H. William Strauss
x Contents
47 Teaching Cases in Nuclear Medicine: Non-oncological Applications�� 1139

Venanzio Valenza, Isabella Bruno, Daniela Di Giuda, Germano Perotti,
Fabrizio Cocciolillo, Valerio Lanni, Lucia Leccisotti, Daria Maccora, and
Valentina Scolozzi
48 Teaching Cases in Nuclear Medicine: Oncological Applications�� 1199
Laura Evangelista, Lucia Setti, Anna Rita Cervino, Gianluigi Ciocia,
Lea Cuppari, Riccardo Vicinelli, and Emilio Bombardieri
49 Radiological Anatomy with CT: What the Nuclear Physician
Should Know When Reading a PET/CT Scan�� 1241
50 Radiological Anatomy with MR: What the Nuclear Physician
Should Know When Reading a PET/MR Scan �� 1257
51 Interpreting and Reporting the Results of Radionuclide Tests�� 1279
H. William Strauss and Federica Orsini
Glossary�� 1297
Index�� 1305
Contributors
Gayane Aghakhanyan Regional Center of Nuclear Medicine, Department of Translational

Research and Advanced Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy
Edoardo Airò Fondazione CNR/Regione Toscana Gabriele Monastero, Pisa, Italy
Akram Al-Ibraheem Department of Nuclear Medicine, King Hussein Cancer Center,
Amman, Jordan
Abedallatif A. AlSharif Department of Radiology and Nuclear Medicine, School of
Medicine, Jordan University Hospital, University of Jordan, Amman, Jordan
Abdullah S. Al Zreiqat Department of Nuclear Medicine, Jordanian Royal Medical Services,
Amman, Jordan
Matteo Bauckneht Nuclear Medicine, IRCCS Policlinico San Martino, Genoa, Italy
Nicola Belcari Department of Physics “E. Fermi”, University of Pisa and INFN, Pisa, Italy
Lisa Bodei Molecular Imaging and Therapy Service, Department of Radiology, Memorial
Sloan Kettering Cancer Center, New York, NY, USA
Ronald Boellaard Department of Radiology and Nuclear Medicine, VU University Medical
Center, Amsterdam, The Netherlands
Emilio Bombardieri Nuclear Medicine Department, Humanitas Gavazzeni, Bergamo, Italy
Scientific Direction, Humanitas Gavazzeni, Bergamo, Italy
Giuseppe Boni Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa, Italy
Roberto Boni Nuclear Medicine Service, “Papa Giovanni XXIII” Hospital, Bergamo, Italy
Elisa Borsò Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Alessandra Boschi Department of Morphology, Surgical and Experimental Medicine,
University of Ferrara, Ferrara, Italy
Christian Bracco Department of Medical Physics, Institute for Cancer research and
Treatment, Candiolo, Turin, Italy
Isabella Bruno Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli
IRCCS, Roma, Italy
Adriana K. Calapaquí Terán Department of Pathology, University Hospital Marques de
Valdecilla (HUMV), Santander, Cantabria, Spain
Valle Camacho Nuclear Medicine Department, Hospital de la Santa Creu i Sant Pau,
Barcelona, Spain
Amedeo Capotosti Department of Physics, Catholic University of the Sacred Heart, Rome,
Italy
xi
xii Contributors
Davide Caramella Diagnostic and Interventional Radiology, Department of Translational

Carlo Cavedon Nuclear Medicine Unit and Medical Physics Unit, University Hospital of
Verona, Verona, Italy
Anna Rita Cervino Nuclear Medicine and Molecular Imaging Unit, Veneto Institute of
Oncology IOV – IRCCS, Padua, Italy
Serena Chiacchio Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa,
Italy
Fabio Chiesa Unit of Radiology, ASL 5 “Spezzino”, Sarzana, La Spezia, Italy
Arturo Chiti Nuclear Medicine, Humanitas Cancer and Research Center IRCCS, Milan, Italy
Department of Biomedical Sciences, Humanitas University, Milan, Italy
Andrea Ciarmiello Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Gianluigi Ciocia Nuclear Medicine Department, Humanitas Gavazzeni, Bergamo, Italy
Fabrizio Cocciolillo Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli
IRCCS, Roma, Italy
Lea Cuppari Nuclear Medicine and Molecular Imaging Unit, Veneto Institute of Oncology
IOV – IRCCS, Padua, Italy
Giuseppe Roberto D’Agostino Radiotherapy and Radiosurgery, Humanitas Cancer and
Research Center IRCCS, Milan, Italy
Lawrence T. Dauer Department of Medical Physics, Memorial Sloan Kettering Cancer
Center, New York, NY, USA
Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
Clemens Decristoforo Department of Nuclear Medicine, Medical University of Innsbruck,
Innsbruck, Austria
Roberto C. Delgado Bolton Department of Diagnostic Imaging and Nuclear Medicine,
University Hospital San Pedro and Center for Biomedical Research of La Rioja (CIBIR),
Logroño, La Rioja, Spain
Diego De Palma Nuclear Medicine Department, “H. Circolo” Varese, Varese, Italy
Alberto Del Guerra Department of Physics “E. Fermi”, University of Pisa and National
Institute of Nuclear Physics (INFN), Pisa, Italy
Daniela Di Giuda Institute of Nuclear Medicine, Università Cattolica del Sacro Cuore, Rome,
Italy
Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy
Rossella Di Stefano Cardiovascular Disease Unit, Department of Surgical, Medical,
Molecular Patology and Medical Emegencies, University of Pisa, Pisa, Italy
Adriano Duatti Department of Chemical and Pharmaceutical Sciences, University of Ferrara,
Ferrara, Italy
Valerio Duce Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Joan Duch Nuclear Medicine Department, Hospital de la Santa Creu i Sant Pau, Barcelona,
Spain
Paola Anna Erba Regional Center of Nuclear Medicine, Department of Translational
Contributors xiii
Montserrat Estorch Nuclear Medicine Department, Sant Pau Hospital, Autonomous

University of Barcelona, Barcelona, Spain
Laura Evangelista Nuclear Medicine and Molecular Imaging Unit, Veneto Institute of
Oncology IOV – IRCCS, Padua, Italy
Maria Evelina Fantacci Department of Physics “Enrico Fermi”, University of Pisa, Pisa,
Italy
Marco Ferdeghini Nuclear Medicine Unit and Medical Physics Unit, University Hospital of
Verona, Verona, Italy
Alejandro Fernández Nuclear Medicine Department, Hospital de la Santa Creu i Sant Pau,
Barcelona, Spain
Elisa Fiasconaro Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa,
Italy
Elena Filidei Nuclear Medicine Unit, Fondazione CNR/Regione Toscana Gabriele
Monasterio, Pisa, Italy
Albert Flotats Universitat Autònoma de Barcelona, Barcelona, Spain
Nuclear Medicine Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain
Lesley Flynt Department of Nuclear Medicine, M.D. Anderson Cancer Center, Houston, TX,
USA
Michela Gabelloni Diagnostic and Interventional Radiology, Department of Translational
Sara Galimberti Hematology Unit, Department of Clinical and Experimental Medicine,
University of Pisa, Pisa, Italy
Onelio Geatti Nuclear Medicine Center, University Hospital of Udine, Udine, Italy
Francesco Giammarile Nuclear Medicine and Diagnostic Imaging Section, Division of
Human Health, Department of Nuclear Sciences and Applications, International Atomic
Energy Agency, Vienna, Austria
Assuero Giorgetti Nuclear Medicine Unit, Fondazione CNR/Regione Toscana Gabriele
Monasterio, Pisa, Italy
Giampiero Giovacchini Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia,
Italy
Elisabetta Giovannini Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Mariano Grosso Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa,
Italy
Federica Guidoccio Regional Center of Nuclear Medicine, Department of Translational
Luca Indovina Department of Medical Physics, Fondazione Policlinico Universitario A.
Gemelli, Rome, Italy
Torsten Kuwert Clinic of Nuclear Medicine, Friedrich-Alexander-University of Erlangen-
Nürnberg, Erlangen, Germany
Adriaan A. Lammertsma Department of Radiology and Nuclear Medicine, Amsterdam
UMC, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
xiv Contributors
Valerio Lanni Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli

IRCCS, Rome, Italy
Elena Lazzeri Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa, Italy
Patrizia Lazzeri Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Lucia Leccisotti Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli
IRCCS, Rome, Italy
Rossella Leoncini Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Elisa Lodi Rizzini Metropolitan Nuclear Medicine, AOU S. Orsola-Malpighi, Bologna, Italy
Serge K. Lyashchenko Department of Radiology, Memorial Sloan Kettering Cancer Center,
New York, NY, USA
Daria Maccora Institute of Nuclear Medicine, Università Cattolica del Sacro Cuore, Rome,
Italy
Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy
Andrea Marciano Regional Center of Nuclear Medicine, Department of Translational
Research and Advanced Technologies in Medicine, University of Pisa, Pisa, Italy
Giuliano Mariani Regional Center of Nuclear Medicine, Department of Translational
Cecilia Marini Nuclear Medicine, Department of Health Sciences, University of Genoa,
Genoa, Italy
CNR Institute of Bioimages and Molecular Physiology, Milan, Italy
Irene Marini Nuclear Medicine Institute, Università Cattolica del Sacro Cuore and Nuclear
Medicine Center, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy
Sören Mattsson Medical Radiation Physics, Lund University, Malmö, Sweden
Sara Mazzarri Regional Center of Nuclear Medicine, Department of Translational Research
and Advanced Technologies in Medicine and Surgery, University of Pisa, Pisa, Italy
Silvia Morbelli Nuclear Medicine, IRCCS Policlinico San Martino, Genoa, Italy
Nuclear Medicine, Department of Health Sciences, University of Genoa, Genoa, Italy
Matteo Morrocchi Department of Physics “E. Fermi”, University of Pisa and INFN, Pisa,
Italy
Cristina Nanni Metropolitan Nuclear Medicine, AOU S. Orsola-Malpighi, Bologna, Italy
Emanuele Neri Diagnostic and Interventional Radiology, Department of Translational
Alberto Nieri Nuclear Medicine, Department of Health Sciences, University of Genoa,
Genoa, Italy
Federica Orsini Nuclear Medicine Unit, “Maggiore della Carità” University Hospital,
Novara, Italy
Federica Padovano Nuclear Medicine Unit and Department of Diagnostics and Public
Health, University of Verona, Verona, Italy
Italia Paglianiti Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa,
Italy
Daniele Panetta CNR Institute of Clinical Physiology, National Research Council, Pisa, Italy
Contributors xv
Roberto Pani Department of Medical and Surgical Sciences and Biotechnologies, Sapienza

University of Rome, Rome, Italy
Giovanna Pepe Nuclear Medicine, Humanitas Cancer and Research Center IRCCS, Milan,
Italy
Germano Perotti Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli
IRCCS, Rome, Italy
Roberta Piva Nuclear Medicine, Department of Health Sciences, University of Genoa,
Genoa, Italy
Giulia Puccini Regional Center of Nuclear Medicine, Department of Translational Research
Erinda Puta Nuclear Medicine Unit, “Maggiore della Carità” University Hospital, Novara,
Italy
Daniele Regge Department of Surgical Sciences, University Hospital “Città della Salute e
della Scienza”, University of Turin, Turin, Italy
Department of Radiology, Institute for Cancer Research and Treatment, Candiolo, Turin, Italy
Matteo Revelli Department of Diagnostic Imaging and Laboratory Medicine, AUSL Reggio
Emilia – IRCCS, Reggio Emilia, Italy
Mattia Riondato Nuclear Medicine Department, “S. Andrea” Hospital, La Spezia, Italy
Piero A. Salvadori PET/Cyclotron Unit, CNR Institute of Clinical Physiology, Pisa, Italy
Massimo Salvatori Nuclear Medicine Institute, Fondazione Policlinico Universitario A.
Gemelli, Rome, Italy
Nuclear Medicine Institute, Catholic University of the Sacred Heart, Rome, Italy
Raffaele Scafè Department of Molecular Medicine, Sapienza University of Rome, Rome,
Italy
Gianmario Sambuceti Nuclear Medicine, IRCCS Policlinico San Martino, Genoa, Italy
Nuclear Medicine, Department of Health Sciences, University of Genoa, Genoa, Italy
Alessandro Scaggion Medical Physics Department, Veneto Institute of Oncology IOV –
IRCCS, Padua, Italy
Valentina Scolozzi Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli
IRCCS, Rome, Italy
Lucia Setti Nuclear Medicine Department, Humanitas Gavazzeni, Bergamo, Italy
Riemer H. J. A. Slart Biomedical Photonic Imaging Group, University of Twente, Enschede,
The Netherlands
Martina Sollini Department of Biomedical Sciences, Humanitas University, Milan, Italy
Michele Stasi Department of Medical Physics, Institute for Cancer research and Treatment,
Candiolo, Turin, Italy
H. William Strauss Molecular Imaging and Therapy Service, Memorial Sloan Kettering
Cancer Center, New York, NY, USA
Elena Tabacchi Metropolitan Nuclear Medicine, AOU S. Orsola-Malpighi, Bologna, Italy
Elisa Tardelli Regional Center of Nuclear Medicine, University Hospital of Pisa, Pisa, Italy
xvi Contributors
Flavia Ticconi Nuclear Medicine, Department of Health Sciences, University of Genoa,

Genoa, Italy
Angela Vaiano Health Physics Unit of Prato and Pistoia, “San Jacopo” Hospital, Pistoia, Italy
Renato A. Valdés Olmos Nuclear Medicine Section and Interventional Molecular Imaging
Laboratory, Department of Radiology, Leiden University Medical Centre, Leiden, The
Netherlands
Nuclear Medicine Department, Onze Lieve Vrouwe Gasthuis Hospital, Amsterdam, The
Netherlands
Venanzio Valenza Institute of Nuclear Medicine, Università Cattolica del Sacro Cuore,
Rome, Italy
Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli IRCCS, Roma, Italy
Eliseo Vañó Department of Radiology, Complutense University of Madrid, Madrid, Spain
Annibale Versari Nuclear Medicine Unit, Department of Oncology and Advanced
Technologies, AUSL-IRCCS Reggio Emilia, Reggio Emilia, Italy
Riccardo Vicinelli Postgraduate Specialty School in Nuclear Medicine, “Milano Bicocca”
University, Milan, Italy
Sergi Vidal-Sicart Nuclear Medicine Department, University Hospital Clinic, Barcelona,
Spain
Alessandro Villa Unit of Radiology 2, ASL 5 “Spezzino”, Sarzana, La Spezia, Italy
Duccio Volterrani Regional Center of Nuclear Medicine, Department of Translational
Yoshiharu Yonekura National Institute of Radiological Sciences, Chiba, Japan
Roberta Zanca Regional Center of Nuclear Medicine, Department of Translational Research
Pat Zanzonico Memorial Hospital Research Laboratories, Department of Medical Physics,
Memorial Sloan Kettering Cancer Center, New York, NY, USA
Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY,
USA
Alessandra Zorz Medical Physics Department, Veneto Institute of Oncology IOV – IRCCS,
Padua, Italy
Pietro Zucchetta Nuclear Medicine Department, University Hospital of Padua, Padua, Italy
Part I
Basic Science
Fundamentals of Natural and Artificial
Radioactivity and Interaction 1
of Ionizing Radiations with Matter
Alberto Del Guerra and Daniele Panetta
Contents
1.1 Brief History of Radiation and Radioactivity 4
1.1.1 Wilhelm Röentgen 4
1.1.2 Henry Becquerel 4
1.1.3 Marie and Pierre Curie 4
1.1.4 Ernest Rutherford 4
1.1.5 International Commission on Radiological Protection 4
1.2 Physical Basis of a Radiation Field 5
1.3 Radioactive Decays 6
1.3.1 Electron Energy Levels 6
1.3.2 Physical Principles of the Radiation Sources 6
1.3.3 The Nuclear Decays 7
1.4 Natural Sources of Radiation 9
1.5 Artificial Sources of Radiation 12
1.6 Interaction of Charged Particles with Matter 14
1.7 Interaction of Neutral Particles with Matter 15
References 19
Learning Objectives • To acquire the knowledge of the various types of decays

• To build up the competence to describe the main discov- (alpha, beta, gamma), laws of decays, decay chains, and
eries that paved the way to the modern science of radia- equilibrium
tion and radioactivity and their numerous applications • To become familiar with the concept and the description
• To become acquainted with the ICRU and SI units for of natural occurring radioisotopes, terrestrial radioactiv-
quantitative description of radiation sources, radiation ity, radioactive families, and cosmic rays
fields, and interaction of radiation fields with matter • To be able to understand the working principles and the
application of the generators of electromagnetic radia-
tion, nuclear reactions and generators of radionuclides,
and cyclotrons
A. Del Guerra (*) • To understand the physical principles of Coulomb inter-
Department of Physics “E. Fermi”, University of Pisa and National actions, light particles versus heavy particles, stopping
Institute of Nuclear Physics (INFN), Pisa, Italy
e-mail: alberto.del.guerra@unipi.it
power, and range
• To be able to describe the interaction of photons with mat-
D. Panetta
CNR Institute of Clinical Physiology, National Research Council,
ter (photoelectric effect, Compton scattering, e−/e+ pair
Pisa, Italy production) and the interaction of neutrons
© Springer Nature Switzerland AG 2019 3

D. Volterrani et al. (eds.), Nuclear Medicine Textbook, https://doi.org/10.1007/978-3-319-95564-3_1
4 A. Del Guerra and D. Panetta
1.1 rief History of Radiation

B the first electrically positive charged, the second negative
and Radioactivity charged, and the third neutral. The first three letters of the
alphabet α (alpha), β (beta), and γ (gamma) were associated
1.1.1 Wilhelm Röentgen with these rays.
X-rays were discovered by Wilhelm Röentgen on Christmas

Eve 1895, and the official announcement was made on 1.1.4 Ernest Rutherford
December 28, 1895. The use of X-rays for medical purposes
was rapidly accepted: in Europe many hospitals used X-rays Ernest Rutherford, during his studies for the understanding
for the diagnosis of trauma and fractures and for locating of the atomic structure, discovered that alpha particles pen-
bullets in soldiers. X-rays were used in absence of knowl- etrate only a few tens of micrometers in aluminum, while
edge of the potential detrimental effects of ionizing radia- beta particles are characterized by a 100-fold more penetrat-
tion and their interaction with living matter. Nevertheless, ing power. Using electric and magnetic fields, he verified that
the idea that X-rays could be used to treat tumors was intro- in the electric field the beta particles are strongly deflected
duced in January 1896 (within 1 month after the announce- toward the positive pole, the alpha particles are deflected to a
ment of the discovery), when a Chicago electrician, Emil lesser degree toward the negative pole, while the gamma ray
Grubbe, started using X-rays to radiate a woman with trajectory is not affected by the presence of the field. As a
relapsing breast cancer, becoming the first radiotherapist in result, it was understood that the beta particles have a nega-
history. tive charge, the alpha particles have a positive charge (and a
greater mass than the beta particles), and the gamma rays are
electrically neutral. The deflection studies of such particles
1.1.2 Henry Becquerel crossing thin layers of material (gold sheets) then led to the
formulation of the atomic structure with a central positive
Henry Becquerel in 1896 was carrying out a series of exper- charge core (i.e., the nucleus), where almost all the mass is
iments to investigate phenomena related to the lumines- concentrated and charged (negative) electrons rotating
cence and/or phosphorescence of certain materials. In around it: the Rutherford’s model of the atom was
particular, he was looking for a possible relationship with established.
the X-ray phenomenon, which he had come to know a few
months earlier. In fact, the presentation of the work on
X-rays performed by W.C. Röentgen had taken place at the 1.1.5 International Commission
beginning of 1896 at the Académie Des Sciences in Paris. on Radiological Protection
During his experiments, Henry Becquerel found that ura-
nium salts placed casually beside photographic plates, com- Several decades after their discovery, it was realized that X-rays
pletely protected from light, caused their darkening. He could produce far superior effects than those originally hypoth-
hypothesized that such salts would emit unknown rays that esized: it was understood that their “administration” for medi-
were able to penetrate the plate container and very similar to cal purposes could not be prolonged because of safety issues
X-rays. He had thus discovered the phenomenon of and had to be regulated. The process of evaluating radiation
radioactivity. damage was started through the most apparent effects: the
emergence of a radiation erythema was considered the typical
warning signal. However, this alarm index was only a qualita-
1.1.3 Marie and Pierre Curie tive element, and it was necessary to wait until 1928 to have a
standardized unit of measure of the intensity of a radiation
Marie and Pierre Curie, in continuing the experiments of beam by using the induced ionization analysis. The International
Henry Becquerel, discovered 2 years later that uranium salts Commission on Radiological Protection (ICRP) was estab-
were not the only ones to emit penetrating radii but that lished in 1928 at the International Radiology Congress (the
other substances, such as thorium, had the same property. original name being “International X-ray and Radium
Marie Curie suggested these substances to be defined Protection Committee”). The ICRP is still an emanation of the
“radioactive” (from the Latin word “radium,” i.e., ray). In International Society of Radiology, but its range covers not
the continuation of the studies with a new mineral, the pitch- only the medical field but also all aspects of protection against
blende, the Curies succeeded in finding new elements, polo- ionizing radiation. The committee issues recommendations on
nium, and radium, whose “radioactivity” was much higher the basic principles of radioprotection. These recommenda-
than that of thorium. In particular, using radium, for the first tions, in turn, are transposed by other international and national
time, they identified and classified three types of radiation: organizations in the form of specific scientific documents and,
1 Fundamentals of Natural and Artificial Radioactivity and Interaction of Ionizing Radiations with Matter 5
then, by international, national, and regional government Δt through a sphere of cross-sectional area S centered in
bodies, through applicable laws and decrees. P. When S becomes infinitely small, Φ becomes
F = dN / dS éë m -2 ùû (1.1)

Key Learning Points
• Wilhelm Röentgen, Henry Becquerel, and Marie Since the number N is a stochastic quantity, in the above
and Pierre Curie together with Ernest Rutherford equation, the expectation value of N must be considered for the
are the authors of the main discoveries that paved computation of Φ. The time derivative of Eq. (1.1) represents
the way to the modern science of radiation and the number of particles passing through P in the unit time, i.e.,
radioactivity and to their numerous applications. the fluence rate (or flux density) which is indicated by ϕ:
• The International Commission for Radiation
Protection is the scientific body responsible for f = d / dt ( dN / dS ) = d 2 N / ( dt dS ) éë m -2 s -1 ùû . (1.2)

issuing the recommendations on the basic princi-
ples of radioprotection. Sometimes the information about the number of particles
is not sufficient. Also the total energy transported by parti-
cles requires a specific quantitative parameter. Let us denote
by R the total kinetic energy of the N particles through P:
1.2 Physical Basis of a Radiation Field
R = åEi ni , i = 1, ¼, M , (1.3)
i
The term radiation is applicable to every physical quantity
that is able to transport energy through space and time. The where ni is the number of particles having energy Ei, M is the
source of the radiation could be due to transformations of total number of energies available in the field, and åni = N
i
unstable nuclei or interactions of particles with nuclei. If the is the total number of particles. The distribution of particles
particles have enough energy, the electromagnetic (EM) ni in each energy level is referred to as the energy spectrum
radiation and charged particles can ionize the matter they of the radiation. Since a finite number of energies are present
interact with. For this reason, these types of radiation are among all the N particles, the spectrum is a discrete energy
called ionizing radiation. spectrum. In a most general case, the energy of the particles
The radiation could be emitted from either the transfor- in a field can have a value in a continuous interval between 0
mations of an unstable nucleus (such as beta electrons, and Emax. In this case, the summation in Eq. (1.3) is replaced
α-particles, γ photons, neutrons, and fission fragments) or by by an integral, and the energy spectrum is called continuous.
secondary processes (such as internal conversion electrons Examples of discrete energy spectra are the photon field of a
or Auger electrons, X-rays, and UV photons). The radiation γ-emitting radionuclide (e.g., 99mTc) or the particle field of an
is emitted in all directions (4π geometry). Particles with an α-emitter (e.g., 210Po), whereas examples of continuous
electric charge can directly ionize atoms through multiple energy spectra are the bremsstrahlung field, the particle field
Coulomb collisions: they are called directly ionizing radia- of a β-emitter (e.g., 3He), or the particle field of a positron
tion. Neutral particles and EM radiation (e.g., photons) can emitter (e.g., 11C). The energy fluence Ψ and energy fluence
also ionize atoms but only through sparse interactions. These rate ψ, measured in the SI in J m−2 and J m−2 s−1, are defined,
latter interactions produce secondary charged particles that, respectively, as:
in turn, lead to direct ionization of the surrounding atoms.
Hence, neutral particles are called indirectly ionizing Y = dR / dS éë Jm -2 ùû , (1.4)
radiation.
A photon is a quantum of electromagnetic energy, i.e., the
y = d / dt ( dR / dS ) = d 2 R / ( dt dS ) éë Jm -2 s -1 ùû . (1.5)
elementary unit of the energy transported by an electromag-
netic wave (in analogy, the electron charge is a quantum of
electric charge). Because of the wave-particle duality of
quantum mechanics, the photon can be treated as a particle
with its own energy (E = hν). Key Learning Points
For a quantitative description of a radiation field, we will • Radiation with sufficient energy can ionize matter,
follow the convention adopted by the International whether directly or indirectly according to its physi-
Commission of Radiation Units and Measurements (ICRU), cal nature.
in Ref. [1]. A radiation field could be thought of as a portion • Several quantities are identified by ICRU to define
of space in which particles move along individual trajectories. quantitatively a radiation field in terms of the num-
At a point P in the field, the particle fluence (or flux), Φ, is ber of particles passing through a given portion of
defined as the number N of particles passing in a time interval space or in terms of the energy that they carry.
1.3 Radioactive Decays  = 0,1, ¼, n - 1 (1.8)
1.3.1 Electron Energy Levels In addition, for each value of ℓ, there are (2ℓ + 1) possible
values of the quantum magnetic number, m (the projection of
To understand the motion of electrons around the nucleus, it ℓ on an arbitrary axis z). Finally, for each m, the electrons can
may be useful to start from the atomic model of Rutherford. take two different spin states. It follows that for each value of
In this model it is assumed that the electrons move on ellipti- ℓ, the maximum number of electrons is given by
cal orbits around the nucleus, with a motion similar to that of
the planets around the Sun. If the eccentricity of the orbits is N e (  ) = 2 ( 2 + 1) (1.9)

neglected, the motion of each electron can be assimilated to
a uniform circular motion. In this case, the centripetal force By combining Eqs. (1.8) and (1.9) and knowing that the
is given by the Coulomb’s force between the nucleus and the sum of the first k numbers is equal to k (k + 1)/2, it is found
electron. For a given radius of the orbit, the electron potential that the maximum number of electrons in a given energy
energy in an atom with atomic number Z is given by level is
n -1
Ep = -1 / 4pe 0 Ze 2 / r (1.6) N e ( n ) = å2 ( 2 + 1) = 2n 2 (1.10)
 =0
where the minus sign indicates that the electron is attracted namely
by the nucleus. According to the previous equation, the nega-
tive potential energy of the electron increases by increasing r • Ne = 2 for K orbitals
until it goes to zero when its distance from the nucleus • Ne = 8 for L orbitals
becomes very large with respect to the atom size. Suppose • Ne = 18 for M orbitals
now that one wants to “extract” an electron from the atom, • Ne = 32 for N orbitals
i.e., to take it from an initial distance r from the nucleus • Ne = 50 for P orbitals
(where the potential energy is equal to Ep) to a very large
final distance, that is, to the “infinite” (where the potential
energy is zero). If we assume that the kinetic energy of the 1.3.2 P
hysical Principles of the Radiation
electron is the same at the two positions, both initial and Sources
final, the energy required to extract the electron is
The transfer of energy from the radiation to the atoms of a
Efinal - Ep = 0 - Ep = Ep (1.7) medium causes its excitation or ionization. In the excitation

process, the transferred energy is sufficient only to change
To displace the electron from the atom, that is, to ion- the atom state to an energized energy level. On the other
ize the atom, it is necessary to provide an energy equal to hand, ionizing radiation can produce the ionization (electron
or greater than the absolute value of Ep. For this reason, expulsion) of the atoms of the medium. A biological medium
this energy is called electron energy. The atomic model of exposed in a field of ionizing radiation becomes the site of a
Rutherford is a classical model: in fact, it cannot explain series of processes, originating from the transfer of energy
many phenomena that can only be described by a quantum from radiation to the medium, which manifest themselves
model. In the early twentieth century, it was replaced by with various effects. It is necessary to relate the effects pro-
the Bohr model (semiclassical model), which in turn was duced with the physical characteristics of the radiation field
further refined by modern quantum mechanics. In the and those of the irradiated medium. To do this, the character-
present atomic model, each electron is in a particular istics of the various types of particles and radiations, their
quantum state, characterized by a certain energy value interactions with the material crossed, and the physical quan-
and other physical quantities (such as angular momentum tities and units used to describe the characteristics of the
and spin). Not all energies are allowed: only a few “dis- radiation fields and material media must be known with
crete” or, better yet, “quantized” values are available. respect to the interactions with ionizing radiation.
These energy values represent
the so-called energy levels The term “radiation” is usually used to describe appar-
uniquely distinguished from the main quantum number. ently very different phenomena such as light emitting by a
The energy of electrons in each energy level increases lamp, heat from a flame, or X-ray from a medical diagnostic
with increasing n. For historical reasons (inherited by machine. A common characteristic of all these phenomena is
spectroscopy), the energy levels with n = 1, 2, 3, 4, 5 are the transport of energy into space without a material propa-
given by the letters K, L, M, N, P, respectively. For each n gation medium. Energy can be transported in vacuum or in
value, the electrons can assume different values of the dielectric materials through electromagnetic waves, i.e.,
orbital quantum number ℓ: oscillations of the electric field and the magnetic field that
Frequency (Hz) 104 108 1012 1015 1016 1018 1020
Type of
Radio frequency (RF) Microwaves Infrared Visible Ultraviolet X-rays Gamma rays
radiation
Wavelength
(m) ∼10–2 ∼10–3 ∼10–5 ∼5 × 10–7 ∼10–8 ∼10–10 ∼10–12
Fig. 1.1 The electromagnetic spectrum. Longer wavelengths correspond to smaller frequency due to the inverse relationship stated in Eq. (1.12).
In this figure, the waves are not drawn in scale for graphical reasons (adapted from Ref. [2])
propagate in space. Electromagnetic waves are characterized accelerated by an electrical potential difference of 1 V. The
by a constant speed in the vacuum and in a medium as kinetic energy acquired by an electric charge that is accelerated
by a uniform and constant electric field is given by
v vacuum = c = 2.99729 ´ 108 m / s (1.11a)

Ec = q ´ DV (1.14)

v medium = c / n (1.11b)
Therefore, by knowing that the charge of the electron is
respectively, where n is the refractive index of the medium. ≈1.602 × 10−19 C, using Eq. (1.14) and the definition of elec-
The frequency of an electromagnetic wave, indicated by the tron volt, we get the following conversion value:
symbol ν, is defined as the number of field oscillations in the
1eV = 1.602 ´ 10-19 [C ´ V ] = 1.602 ´ 10-19 J (1.15)
time unit at a given location of the space. The w
avelength, indi-
cated by λ, is the distance between two consecutive maximum
field strength (peaks) positions at a given instant (see Fig. 1.1).
The relation that correlates speed, frequency, and wavelength is
1.3.3 The Nuclear Decays
n =c/l (1.12)
For the atom to be ionized due to wave interaction, the Atoms in nature can be stable or unstable. An atom is stable
wavelength must have a similar value to the size of the atom. if its nucleus does not undergo spontaneous transformations.
Therefore the wavelength λ must be of the order of 10−10 m. This means that in a stable atom, the number of protons and
The frequency of such a wave can be calculated with Eq. neutrons contained in the nucleus maintains their nuclear
(1.12), obtaining ν = c/λ ≈ 3 × 1018 Hz. Considering that the energy levels unchanged over time. Unstable atoms are those
visible light frequency is in the order of 1014 Hz (about in which the nucleus spontaneously undergoes transforma-
10,000 times lower), it is clear that it is not possible to ionize tions. Such spontaneous nuclear transformations are called
the atoms with this type of radiation. “decays.” A nuclear decay is always accompanied by the
It is useful to reformulate the concept in terms of energy emission of some form of radiation: for this reason, unstable
transmitted by electromagnetic radiation. Electromagnetic elements are also called radioactive. Following one or more
waves propagate within “packets,” called photons, which can transformations, a radioactive nucleus always decays into a
be treated as energy particles equal to stable nucleus, which can belong to a chemical species equal
to or different from the initial one. The nuclear decay is a
E = hn (1.13)
probabilistic phenomenon. It is not possible to determine the
where h = 6.62 × 10−34 [J s] is the Planck constant. For a photon instant in which a given unstable nucleus will undergo a
to ionize an atom, it is necessary that its energy is greater than transformation. However, we can quantitatively describe the
or equal to the binding energy of the atomic electrons involved temporal evolution of a very large set of nuclei by the follow-
in the interaction. To express numerically the energies involved ing law of radioactive decay:
in the ionization and excitation processes, it is convenient to
use a special unit called electronvolt (eV). One electronvolt is
N ( t ) = N 0 exp ( -l t ) (1.16)
defined as the kinetic energy acquired by an electron that is
In the previous equation, N0 represents the number of A ( t ) = -dN / dt = - N 0 ( -l ) exp ( -l t ) = l N ( t ) = N ( t ) / t

radioactive nuclei contained in the material (sample) consid-

ered at a given initial moment t0 = 0; N(t) represents the num- (1.19a)
ber of radioactive nuclei contained in the sample at a time t A ( t ) = A0 exp ( -l t ) (1.19b)
from the initial instant (i.e., the number of nuclei that at time
t has not yet undergone any transformation); and λ is the The activity measurement unit is the Becquerel (Bq). An
decay constant, which represents the average number of activity of 1 Bq corresponds to a nuclear transformation (dis-
nuclear decays occurring in the unit of time. The constant λ is integration) per second. Another measure of activity, of his-
characteristic of each radioactive isotope and does not depend torical interest but also very much used in practice, is the
on the amount of matter contained in the radioactive sample Curie (Ci), defined as the activity of 1 g of radium-226. The
under examination. It has the inverse dimension of time and is conversion between Bq and Ci is obtained by the following
measured in s−1. Equation (1.16) describes the following con- equivalence:
cept: the number of unstable nuclei, contained in a given
1 Ci = 3.7 ´ 1010 Bq (1.20)
radioactive material sample, decreases exponentially over
time; τ is defined as the average lifetime of a given nuclear There are various types of nuclear decay. The α-decay is
species, i.e., the inverse of the decay constant λ: the spontaneous particle spraying by heavy nuclei, character-
ized by an excess of protons; α is a positively charged heavy
t = 1/ l (1.17)
particle, formed by two protons and two neutrons, identical
“Average lifetime” means the average time between the to a helium-4 nucleus (4He). Some examples of α-decaying
production (either natural or artificial) and the decay of a radionuclides are 226Ra, 222Rn, and 210Po. The β− and β+
given radioactive nucleus. This time varies considerably from decays consist of the emission of an electron (e−) and of a
isotope to isotope and can assume values between fractions of positron (positive electron, indicated by e+) by a nucleus,
second and billions of years. Another frequently used quan- respectively. Some elements that undergo decay β are 90Sr
tity, strictly related to the average lifetime, is the half-life, t1/2, (β−) and 18F (β+). The γ-decay consists in the emission of a
indicating the time required to halve the number of radioac- photon by the nucleus originally in an excited state. Photons
tive nuclei contained in the sample (see Fig. 1.2): emitted in the γ-decay are physically indistinguishable from
X-rays, though the latter are produced by de-energizing pro-
t1/ 2 = t ´ ln 2 (1.18) cesses of atomic electrons. X photons have generally lower
energies than γ photons. The distinction between photons of
The activity A(t) is defined as the average number of type X or γ is based only on the mechanism by which they
nuclear decays occurring at time t in the unit of time. were generated (atomic or nuclear) and not on the energy
According to the definition, the activity coincides with the they carry. Radionuclides may have different allowed routes
decay rate (i.e., the derivative) of N(t), changed by sign: of nuclear de-excitation. When more decays are possible for
a given nuclear species, we use the term “branching ratio”
N(t)/N0 (BR) of a decay mode (or decay channel) to denote the ratio
1.0 between the number of decays of that mode and the total
0.9 number of disintegrations. For instance, the BR for β+ decay
0.8 of 18F is 96.86%, whereas the same nuclide can decay also by
0.7 electron capture (EC) with a BR of 3.14%.
0.6
N=N0/2 0.5
0.4
0.3 Key Learning Points

N=N0/4
0.2 • Atomic electrons are arranged in orbitals with dis-
N=N0/8
0.1 crete energy levels and angular momentum accord-
0.0 ing to quantum mechanics.
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
• The atomic nuclei can be stable or unstable.
t/τ
t=t1/2 t=2t1/2 t=3t1/2 • Unstable nuclei can reach stability by undergoing
single or (more frequently) chains of transforma-
Fig. 1.2 Exponential decay law. N0 is the number of radioactive nuclei tions (i.e., decays) leading to the emission of alpha
contained in a sample at the initial instant t0 = 0; N(t) is the number of particles (4He nuclei), beta particles (electrons or
non-decayed nuclei at time t; and τ and t1/2 indicate the average lifetime
and the half-life of the considered nuclear species, respectively (repro- positrons), or gamma photons (i.e., EM waves).
duced from Ref. [2])
1.4 Natural Sources of Radiation with half-life of 1.4 × 1010 year and ending to the stable
isotope lead-208. All elements of the radioactive series
Radioactive isotopes reach stability by radiation emission. undergo α or β decay; hence all the modifications of the
Such isotopes can be naturally occurring in rocks, in the atmo- mass number through the series occur in multiples of 4. For
sphere, or in water, or they can be artificially produced. In the this reason, the thorium family is sometimes referred to as
first case, they are classified as natural sources of radiation. the A = 4n series, meaning that all the elements of the series
Some natural sources of interest in the radioprotection field have a mass number that can be obtained by multiplying 4
are uranium-238 (238U), potassium-40 (40K), radium-226 by an integer number (e.g., n = 58 for thorium-232, n = 57
(226Ra), and radon-222 (222Rn). In particular, 222Rn is one of the for radium-228, and so on). With this convention, the so-
major sources of radiation risk for the population, as it is a called uranium family is referred to as A = 4n + 2 family.
colorless and tasteless gas that can emerge from the subsoil to This series has uranium-238 as progenitor (half-life
reach high concentrations in basements and poorly ventilated 4.47 × 109 year) and ends into the stable isotope lead-206.
areas. The radioactive isotopes present in the rocks are referred Radium-226 and its daughter radon-222 belong to this
to as primordial sources of radiation, as they have been pro- decay chain. Also the uranium-actinium (or simply actin-
duced in cosmic events prior to the formation of Earth. Other ium) family (A = 4n + 3) ends to a stable isotope of lead
natural-type radiations are cosmic radiation, mainly consisting (lead-207); the name of this series follows from the histori-
of high-energy protons and cosmogenic radiation such as trical name of 235U, “actinouranium” (half-life 7 × 108 year),
tium (hydrogen-3) and carbon-14 being produced by the inter- with lower natural abundance with respect to 238U. It also
action of cosmic radiation with the stable nuclei present in the helps to avoid ambiguity with the uranium family that starts
atmosphere or on the Earth. The main properties of all known from a uranium isotope as well. A family with A = 4n + 1
isotopes, whether they are stable or unstable, natural or artifi- also exists (the neptunium series), but it cannot be included
cially produced, are organized in the Nuclide Table (see in the present list of natural sources of radiation because
Fig. 1.3a). their members can only be produced artificially by nuclear
In this table, the various nuclides are ordered in a matrix reactions.
where the rows are characterized by an equal number of The rate of production of members of a radioactive fam-
protons and the columns are characterized by an equal num- ily depends on the availability of their parents; hence, in a
ber of neutrons. Moving from left to right, the N value given segment of a decay chain A → B where A and B are
increases, while going from the bottom up, there are Z both unstable, the rate of emission of radiation of the
increasing numbers. Stable isotopes in the Nuclide Table daughter nucleus B is depending on the decay rate of its
form the so-called stability curve (see Fig. 1.3b). By observ- parent A. When t1/2,A is much longer than t1/2,B, after a suf-
ing this curve, it is noted that for lightweight elements ficient amount of time, the amount of daughter nuclei NB
(Z < 20), most of the stable nuclei contain an equal number of reaches a constant value, due to the fact that the rate of its
protons and neutrons and are therefore near the Z = N line. production equals the rate of its decay. In formulae, this can
For heavier elements there is a deviation from this symmetry, be show by considering that the time derivative of NB is
and the stable nuclei are those with an excess of neutrons. given by
This is due to the necessity of having more attractive nucleon-
dN B
nucleon “strong” force to counterbalance the proton-proton = lA N A - lB N B (1.21)
dt
Coulomb repulsive force.
All elements from Z = 84 (polonium) to Z = 92 (ura- The first term on the right-hand side has a positive sign
nium) are radioactive, whereas all elements with Z > 92 because NB grows up as the parent nucleus A undergoes dis-
(transuranic elements) are not naturally occurring and can integration. The more general solution of the above differen-
only be produced artificially (see next section). Bismuth-209 tial equation is
(Z = 83), which has been long regarded as the heaviest sta-
ble nucleus, has been recently found to be a metastable iso- lA
tope undergoing α-decay, even though with an incredibly
N B = N 0, A
lB - lA
( e-lAt - e-lBt ) + N0,Be-lBt (1.22)

long half-life of 2 × 1019 years, i.e., more than one billion
times the age of the universe [4]. Elements from polonium where N0,A and N0,B are the initial numbers of parent and
to uranium are linked to each other by decay chains, also daughter nuclei, respectively.
called radioactive “families” or radioactive series (see Let us suppose to start from a pure sample of A, so that
Fig. 1.4). Each family has a radioactive progenitor, decay- N0,B = 0. If t1/2,A ≫ t1/2,B then λA ≪ λB (see Eqs. (1.17) and
ing into radioactive daughters and so on until a stable con- (1.18)) and hence λB − λA ~ λB. Furthermore, if the observation
figuration is reached. The three main radioactive families time t is long enough (i.e., t ≫ t1/2,B), we have e - lBt  e - lA t .
have a naturally occurring radioactive progenitor. The tho- Under these approximations, the solution for NB can be
rium family starts from thorium-232, undergoing α-decay rewritten as follows:
100
80
N
60
Z=
40
20 Stable nuclei
0 20 40 60 80 100
b N
Fig. 1.3 (a) Excerpt from “Table of Radioactive Isotopes” (Ref. [3]). (b) The stability curve. For increasing Z, the stable nuclei are those charac-
terized by an excess of neutrons (adapted from Ref. [2]). The plotted data mimics the real distrubution of the stable nuclei
lA
NB = NA (1.23) and hence AB = AA. In other words, at secular equilibrium the
lB
activity of the daughter equals the activity of the parent.
In the above equation, we have exploited the equation Because we have assumed a very long half-life of the parent
N A = N 0, A e - lA t . This situation is called secular equilibrium. species, at secular equilibrium, this activity remains constant
In terms of activity, it’s easy to show that multiplying both with good approximation until some processing is done to
sides of Eq. (1.23) by λB we get break the equilibrium condition (e.g., uranium ores are
extracted during mining). As an example, radon-222
lB N B = lA N A (1.24) (t1/2 = 3.82 days) is in equilibrium with its parent radium-226
α
A
X Ancestor
A
β– X Stable
Z
238 234 235
U U U 92 (Uranium)
234M 231Pa
91 (Protactinium)
232Th 228Th 234Th 230Th 231Th 227Th

90 (Thorium)
228Ac 227Ac
89 (Actinium)
228Ra 224Ra 226Ra 223Ra 88 (Radium)
223Fr 87 (Francium)
220Rn 222Rn 219Rn

86 (Radon)
218At 219At 215At

85 (Astatine)
216Po 212Po 218Po 214Po 210Po 215Po 211Po

84 (Polonium)
212Bi 214Bi 210Bi 215Bi 211Bi
83 (Bismuth)
212Pb 208Pb 214Pb 210Pb 206Pb 211Pb 207Pb

82 (Lead)
208Tl 210Tl 206Tl 207Tl
81 (Thallium)
206Hg
80 (Mercury)
Thorium family Uranium family Actinium family

(A=4n) (A=4n+2) (A=4n+3)
Fig. 1.4 The three radioactive families with naturally occurring ancestors (thorium-232, uranium-238, and uranium-235). The half-lives and
branching ratios are not reported for graphical reasons. The reader can refer to Ref. [5] for a complete decay scheme of all these elements
(t1/2 = 1620 years) if the radium-containing sample is left the parents’ one even though the ratio between them remains
undisturbed for t ≫ t1/2 (222Rn) (typically 4–5 weeks). constant over time. This condition is called transient equilib-
When the parent’s half-life is not much greater than the rium. By starting from a pure A sample, the time required to
observation time, the daughter’s half-life is not treated as reach the maximum activity of the daughter B is given by
negligible, and its activity does not reach a constant value
after long times. In this case, Eq. (1.23) (which is valid after æl ö
ln ç B ÷
long times of observation) must be rewritten as follows: l
tmax = è A ø. (1.27)
lB - lA
lA
NB = NA . (1.25)
lB - lA A common example of transient equilibrium in nuclear

medicine is that of molybdenum-99 (the production of 99Mo is
Multiplying both sides by λB, we get the following equa- discussed in the next section, as it is only produced artificially),
tion in terms of activity: undergoing β− decay with half-life of 65.94 h. Its daughter
99m
Tc is in an excited state, and its stability is reached by
lB γ-decay with a half-life of 6 h. By starting from a pure sample
AB = AA . (1.26)
lB - lA of 99Mo, the maximum activity of 99mTc is reached after 22.8 h.

Another important primordial radioactive element is
The previous equation implies that, after a sufficient potassium-40. This isotope has a half-life of 1.25 × 109 year
amount of time, the daughter’s activity becomes greater than and undergoes either β+ decay to calcium-40 (branching ratio
89.28%) or electron capture to argon-40 (branching ratio abrupt abrasion of electrons on the anode. The spectrum of
10.72%). Approximately 120 ppm of the total amount of the emitted radiation is composed of a continuous compo-
potassium found in nature is in form of 40K. This is important nent (braking radiation, in German “bremsstrahlung”) and
from the point of view of nuclear medicine because for a man another discrete (“lines”) component (characteristic of the
weighting 70 kg, about 16 mg of 40K are contained in the anode material). In the case of sources used in computed
body giving rise to a background body activity of ca 4.3 kBq. tomography (CT), radiogenic tubes must be able to with-
Besides radiation coming from radioactive elements pres- stand substantial workloads. In these cases, the anode is
ent in rocks and soil, conveniently organized in radioactive rotated very quickly so that the electron beam does not
families as explained above, radiation also comes from out- always hit the same metal area so as to prevent the overheat-
side the Earth’s atmosphere and can be either originated by ing of a single disk area and to promote thermal dissipation.
the Sun or from outside the solar system. The solar compo- X-ray tubes for medical applications are limited to acceler-
nent of such cosmic radiation is mainly composed by high- ating voltages <400–500 kV. For therapy application, X-ray
energy protons (primary radiation) that produce secondary beams should have energies in the range of 6–25 MeV to avoid
radiation (photons, electrons, neutrons, pions, and muons) skin burning and optimizing the treatment of deep tumors [6].
when interacting with the molecules of the atmosphere. The For this purpose, electron beams are accelerated by means of
primary cosmic ray component coming from outside the linear accelerators (LINACs). LINACs employ resonant cavi-
solar system has in general higher intensity, and its spatial ties in which alternating electric fields oscillating at radio fre-
distribution is isotropic. quency (RF) provide “kicks” to the electron bunches in such a
way that the kinetic energy of the bunch is increased at each
passage from one cavity to the next one. In medical LINACs,
Key Learning Points typical frequencies in the RF field are in the order of few giga-
• Radioactive sources can be naturally occurring, hertz, and the final kinetic energy at the beam exit is in the
such as those present in rocks (terrestrial radioactiv- order of few megaelectronvolts to few tens of megaelectron-
ity) or coming from outside the Earth (cosmic rays). volts. After some focusing steps, the accelerated electrons are
• Naturally occurring radioisotopes all belong to then directed to the LINAC’s head where a target of high-Z
radioactive decay chains (or families) starting from material is used to produce X-rays via bremsstrahlung. Unlike
thorium-232 (thorium family), uranium-238 (ura- low-energy electron accelerators such as radiogenic tubes
nium family), or uranium-235 (actinium family). where the useful X-ray beam is reflected off a thick anode,
megavoltage X-ray beams are generally employing transmis-
sion targets. In this way, the useful photon beam is mainly
directed along the same direction of the accelerated electron
1.5 Artificial Sources of Radiation beam. For radiotherapy applications, the LINAC’s head is
mounted on a rotating gantry in order to allow flexibility in the
The natural sources of radiation are of little interest in medi- choice of the beam angle with respect to the patient’s position
cine. Most natural radioisotopes have a lifetime too long to on the treatment table. In medical LINACs, the electron beam
be safely used in therapeutic or diagnostic fields. Furthermore, can also be used right away for treatment (i.e., without con-
it is also difficult to obtain them with the levels of purity verting it to a photon beam) for special types of treatment
required for this type of application. It is preferable to artifi- where the radiobiology and type of interaction of electron
cially produce the desired radioisotopes using particle accel- beams can be advantageous (e.g., superficial tumors or “dose
erators that allow the nuclei contained in a target to be boost” to tumors already treated with photons).
activated in a controlled manner. LINACs can also be used to accelerate protons and ions for
A very common artificial radiation source in the medical radionuclide productions. Nevertheless, this latter task is gener-
field is the radiogenic tube, used in radiodiagnostics to pro- ally accomplished by means of cyclotrons. The cyclotron was
duce X-rays (Fig. 1.5). In a radiogenic tube, electrons emit- invented in 1929 by Ernest O. Lawrence, who received the
ted by a cathode by thermal effect are initially accelerated by Nobel Prize in Physics in 1939 “for the invention and develop-
an electrical potential difference (typically 40–140 kV ment of the cyclotron and for results obtained with it, especially
depending on the specific diagnostic application) and, subse- with regard to artificial radioactive elements.” Similar to
quently, are abruptly braked in a collision with a heavy metal LINACs, also in cyclotrons the particles are accelerated by
anode. The cathode consists of one or two tungsten filaments mean of electric fields oscillating in the RF range (10–30 MHz).
that are heated by Joule effect at temperatures of 1500– The electric field is applied between two D-shaped electrodes
2600 °C. The anode can be tungsten or molybdenum (the (hence called “dees”), where a magnetic field provides the cen-
latter is used for mammography tubes) and is designed to tripetal force sufficient to keep the particle in a semicircular tra-
optimize the X-ray escape from the exit window and maxi- jectory (see Fig. 1.6). After each passage between the two dees,
mize thermal dispersion. X-rays are emitted following the the particle gains kinetic energy due to the effect of the RF volt-
Fig. 1.5 Basic scheme of a a

radiogenic tube for the
production of kilovoltage Rotating anode
X-rays (a), along with a (+)
typical X-ray energy spectrum
encountered in diagnostic
radiology (b) (adapted from Tube envelope
Ref. [2])
Rotor
Electrons
Cathode (–)
Exit window
X-ray beam
b
120 kVp, 2.7 mm Al
5×106
Number of photons at 20 cm (mA–1 s–1 mm–2)
4×106
3×106
2×106
1×106
0
0 20 40 60 80 100 120
hν (kev)
age, and the radius of the semicircular trajectory increases pro- protons (H+). The conversion of H− ions to protons is performed
portionally, thus creating a spiral shape trajectory. When the at the end of the acceleration process by means of stripping foils
radius of the spiral trajectory approaches the radius of the dees, [7]. This mechanism allows having more efficiency in the
the accelerated particle beam is extracted and sent to a target extraction of the accelerated particle beam.
designed in such a way that the desired nuclear reaction occurs. Targets are specifically designed in order to have the high-
The most common reaction useful for PET imaging consists in est production yield of the desired radionuclide for the avail-
bombarding with protons a liquid target containing 18O atoms able proton energy. In some case, deuterons (d) are
(water-18O), which yields 18F and neutrons. The notation for accelerated instead of protons: for instance, the 15-oxygen
such reaction is 18O(p,n)18F. In most modern cyclotrons for med- (β+ emitter, with half-life of 122 s) can be produced in cyclo-
ical applications, negative H ions (H−) are accelerated instead of tron via the reaction 14N(d,n)15O.
the reaction 98Mo(n,γ)99Mo [8]. Due to the relatively small

a
number of research reactors worldwide producing 99Mo,
there is currently an active research on alternative production
technologies in order to avoid future shortages of this radio-
nuclide [9].
Key Learning Points

• Due to the long half-life of most naturally occurring
radioisotopes, artificially produced radioactive
sources are preferred for medical applications.
• These types of radiations are produced with electro-
static generators (X-ray tubes, LINACs), or by
nuclear reactions by means of cyclotron-accelerated
b Beam extraction
particles (e.g., widely used for the production of
(to target)
PET radionuclides).
• Generators of radionuclides (such as the 99Mo/99mTc
generator) are obtained by nuclear fission in nuclear
reactors.
1.6 I nteraction of Charged Particles

with Matter
When the particles of a radiation field hit a target material

(e.g., the human body or a radiation detector), they can trans-
fer part or all of the transported energy to the target atoms.
Fig. 1.6 (a, b) Configuration of “dees” in a commercial cyclotron for This is the basic mechanism exploited in all radiation detec-
medical application (PETTrace™, General Electric). The magnetic tors to build a measureable signal (most often, an electric
field B (perpendicular to the plane of the dees) provides the centripetal signal), but it’s also at the base of the biological effects (e.g.,
force required to steer the protons (or deuterons) during acceleration.
when targets are living organisms) of radiations. However,
After acceleration due to the alternating voltage applied between the
two dees, the particle beam is extracted and directed toward the target these effects are outside the scope of this chapter and will not
where specific nuclear reactions (depending on the desired radionu- be treated. Several types of interaction can occur leading to
clide) occur. The drawn spiral trajectory (bottom) is just a graphical the deposition of radiation energy to matter, depending on
example and can differ from the actual one
(1) the type of radiation (i.e., charged or neutral, heavy or
light particles), (2) the energy spectrum of the radiation field,
Specific nuclear reactions are also employed to produce and (3) the chemical and physical properties of the traversed
radionuclide generators, even though such reactions are not matter.
practicable for hospital accelerators and can only be per- Charged particles interact with electrons and nuclei of the
formed on nuclear reactor facilities. The more common route atoms in the target via electric (Coulomb) forces. The stop-
of 99Mo production is by fission of 235U in research nuclear ping power, S, allows us to describe quantitatively the energy
reactors. A nuclear fission is a process in which a heavy dE lost by the incident particles per unit length in the target
nucleus splits into two or, more rarely three, nuclear frag- material, dx:
ments. Besides the major nuclear fragments, also lighter par-
dE é J ù
ticles (such as neutrons and/or alphas) are emitted, with a S =- (1.28)
positive energy balance. We can denote the production of dx êë m úû
99-molybdenum through this route with 235U(n,f)99Mo, Even though in the SI, the stopping power is expressed in
where the letter “f” refers to the “fission products.” The yield units of J m−1, a more common unit of measure is keV μm−1.
is relatively low, as the fission of uranium-235 gives rise to The stopping power has slightly different expressions for
molybdenum-99 with a probability of 6%. Alternatively, heavier particles (protons, α-particles, and larger ions), com-
enriched targets of 98Mo can be bombarded with neutrons via pared to lighter particles (electrons or positrons). For heavier
particles (e.g., protons, α’s, ions), the ratio of S to the mate- Table 1.1 Positron energies and ranges for the most common β+ emit-
rial density ρ (referred to as the mass stopping power) has the ters used in PET (adapted from Ref. [10])
following dependence on the relevant variables: Isotope Emean (keV) Emax (keV) Rmean (mm) Rmax (mm)
18
F 252 635 0.660 2.633
S ZZ p C 390 970 1.266 4.456
11
µ 2 , for b < 0.96 (1.29a) 13

N 488 1190 1.730 5.572
r b 15
O 730 1720 2.965 9.132
68
Ga 844 1899 3.559 10.273
S 64
Cu 1280.6 2926 6.077 16.149
µ Z p2 Z ln b , for 0.96 < b < 1 (1.29b)
r 82
Rb 1551 3378 7.491 18.603
where Zp is the atomic number of the primary particle, Z is

the atomic number of the target material, and β = v/c is the energy of the beta spectrum (Rmean) and for the energy end
particle velocity. point (Rmax).
For light charged particles, a non-negligible amount of
energy is spent by particles in processes that end up in the
emission of secondary EM radiation (“radiative” processes): Key Learning Points
we hence distinguish between “collision” and “radiative” • Charged particles interact with atomic electrons and
components of the stopping power, denoted by Sc and Sr, nuclei of the target material via Coulomb-type
respectively, where S = Sc + Sr. Below there are reported the collisions.
main dependencies of Sc and Sr on the relevant variables: • The stopping power describes the average kinetic
Sc Z energy lost by collision per unit length.
µ 2 (1.30a) • The range is defined as the mean value of the mate-
r b
rial thickness traversed by the particle in a given
Sr material.
µ Z 2b 2 , for Ekin  mp c 2 (1.30b)
r
where Ekin is the kinetic energy of the particle, and mpc2 is its
rest mass. 1.7 I nteraction of Neutral Particles
After having lost most of its kinetic energy travelling with Matter
through a medium, a charged particle could reach thermal
equilibrium with surrounding atoms. The mean value of the Uncharged particles (e.g., photons and neutrons) can only ion-
path travelled by each particle of a charged particle beam in ize matter in an indirect way. When a photon enters the patient’s
a target material is defined as the range of the particle in that body, it interacts with the atomic electrons (or, in certain cir-
medium (Fig. 1.7). The range is a function of the energy and cumstances, with the nuclei) with mechanisms depending
charge of the particle and of the atomic number and density mainly on the photon energy E = hν and the Z of the material.
of the target material. In the energy range relevant for radiology and nuclear medicine
Table 1.1 reports the values of the range in mm for the (i.e., from tens of keV to around 1 MeV), mainly four types of
most used β+ emitters in nuclear medicine for the mean photon-matter interactions can occur (see Fig. 1.8).
Protons / ions Electrons / positrons
100 100
Transmission (%)
Transmission (%)
90 90
80 80
70 70
60 60
Incident 50 50
Detector
beam 40 40
30 30
20 20
10 10
0 0
L R L R50 Rex L
Fig. 1.7 Operative definition of range. For heavy particles, the average challenging, and besides the average range (here denoted by R50) also an
range R is conveniently defined as the 50% of the falling edge of the “extrapolated range” can be defined as shown in the right panel (adapted
transmission curve. The definition of range for light particles is more from Ref. [2])
Rayleigh scattering Photoelectric absorption

Fluorescence X-ray photon
L
Scattered M
photon K
Primary
Excited Primary
photon Atom
atom photon
Secondary electron
(or photo-electron)
Compton scattering Pair production
Scattered
photon Electron (e–)
Electric field
of the nucleus
Nucleus
Primary Primary
photon photon
Secondary electron Positron (e+)
(or Compton electron)
Fig. 1.8 Main interactions of photons with matter in the energy range of interest for diagnostic radiology and nuclear medicine
The Rayleigh scattering (or coherent scattering) occurs tering). It results in the emission of a secondary electron (due
when an incident photon of low energy interacts with the to the kinetic energy transferred from the primary photon to
whole target atom. The scattered photon has the same wave- the target electron) and on a secondary, scattered photon with
length (and hence, the same energy) of the primary one. No different direction and lower energy than the primary one
net transfer of energy to matter arises from this type of scat- (see Fig. 1.8). The term “incoherent” arises from the fact that
tering. The incident photon is removed from the primary each scatter event involves a single electron instead of the
beam as it is scattered within a narrow cone with respect to atom as a whole. The energy Escatter = hν′ of the scattered
its original direction. photon is a function of the initial energy hν of the photon and
If the primary photon energy is high enough to ionize the the scattering angle θ:
target atom, it can be completely absorbed while an electron
of the target atom is ejected with kinetic energy: hn
hn ¢ = (1.32)
hn
E pe = hn - E b (1.31) 1+ (1 - cos q )
me c 2

where the ejected electron is referred to as “photoelectron.”
In the previous equation, Eb is the binding energy of the elec- where me is the mass of the electron. At higher energies, pho-
tron orbital before the collision, which depends on the Z of tons can interact with the electric field of the atomic nucleus
the material, on the type of orbital (s, p, d, etc.), and on its leading to the production of an electron-positron pair
level (K, L, M, etc.). This process is called photoelectric (Fig. 1.8), in agreement with the Einstein’s equation for
effect (or photoelectric absorption). This process can only mass-energy equivalence, E = mc2. This process is allowed
occur if hν > Eb. only if the initial energy hν of the primary photon is greater
Photoelectric absorption has a low probability to occur on than the total rest mass energy of the two output particles:
weakly bound electrons. In this case, the interaction can be
hn > me + c 2 + me - c 2 = 2me c 2 = 1022 keV (1.33)
modelled as a relativistic collision between a photon and a
free (i.e., unbound) electron at rest. In most cases, this type The difference between the initial photon energy and the
of interaction occurs with electrons of external orbitals, and threshold energy is shared between the electron and the posi-
the process is called Compton scattering (or incoherent scat- tron being created in form of kinetic energy. The nucleus also
shares a small (in most cases, negligible) fraction of recoil The cross section is expressed in m2, although a common
energy, in such a way that the conservation laws for energy unit is the barn (1 b = 10−28 m2). The linear attenuation coef-
and momentum are preserved. ficient is obtained by multiplying the total cross section by
All the three interaction processes described above (pho- the number of atoms per unit volume:
toelectric absorption, Compton scattering, and e+e− pair pro-
NA r
duction) result in the emission of one or more charged m= s tot (1.35)
particles, which in turn can ionize the surrounding atoms via A
Coulomb-type interactions as explained in the previous sec- where A is the mass number, ρ is the material density, and NA
tion. In this sense, photons in the energy range covered in the is the Avogadro number. The dependence of the linear atten-
context of nuclear medicine and diagnostic radiology are uation coefficient on Z and hν is given by the interaction
called “indirectly ionizing” radiations. cross section for the individual interaction types and can be
Let us now define some useful parameter for a macro- seen in Fig. 1.9 for photon beams interacting in water (top)
scopic description of the photon beam interaction with mat- and lead (bottom).
ter. The atomic or nuclear cross section, σ, represents the The linear attenuation coefficient (μ) has the dimension of
probability for a photon to undergo a given interaction on a the inverse of a length, so its SI unit is m−1. The parameter μ
single atom (or nucleus). The overall probability of interac- is a function of the incident photon energy and of the Z and ρ
tion is the sum of the cross sections for all the possible types of the target material.
of interaction: When a monoenergetic photon field with initial fluence
Φ0 traverses a homogeneous slab of material with linear
s tot = ås i (1.34)
attenuation coefficient μ and thickness x, the resulting
i
a
H2O, Zeff=7.51
3
10 Total
Rayleigh
2
10 Photoelectric
Compton
1
10 Pair production
100
µ/ρ (cm2/g)
10-1
10-2
Nuclear Medicine
10-3
Diagnostic radiology
Radiotherapy
10-4
10-5
10-6 0
10 101 102 103 104
hν (keV)
Fig. 1.9 Energy dependence of the total and interaction-specific linear attenuation coefficient for photons in water (a) and lead (b) [11]. In lead,
the electron binding energies are within the photon energy range, leading to sharp edges in the photoelectric absorption cross section
b Pb, Z=82
L-edges
L1: 15.9 keV
103 L2: 15.2 keV Total
M-edges
L3: 13.0 keV
M1: 3.9 keV Rayleigh
102 M2: 3.6 keV
M3: 3.1 keV Photoelectric
M4: 2.6 keV K-edge
M5: 2.5 keV 88.0 keV Compton
101
Pair production
100
µ/ρ (cm2/g)
10-1
10-2
Nuclear Medicine
10-3
Diagnostic radiology
Radiotherapy
10-4
10-5
10-6
100 101 102 103 104
hv (keV)
Fig. 1.9 (continued)
(attenuated) fluence Φ after the slab is given by the Beer- tron enters the target nucleus, this latter one is set to an
Lambert law of exponential attenuation (see Fig. 1.10): excited state; the subsequent de-excitation is followed by the
emission of a prompt γ photon.
F = F0 exp ( - m x ) (1.36) For fast neutrons, most of the energy loss of the primary

particle is due to elastic scattering. The highest fraction of
Neutrons can also ionize atoms in an indirect way, but, neutron energy is lost through elastic collision with target
instead of interacting with atomic electrons, they are more particles of approximately the same mass (i.e., hydrogen
likely to interact with atomic nuclei via elastic or inelastic nuclei). More generally, the average energy loss for elastic
nuclear scattering. Nuclear reactions often occur when the collision is given by
target nucleus absorbs the neutron with the subsequent emis-
sion of various kinds of secondary radiations. 2 Ekin A
DE = (1.37)
( A + 1)
2
The atomic cross section of each type of interaction

strongly depends on the energy of the incident neutron, as
well as on the target material. For fast neutrons (with kinetic where A is the mass number of the target atom and Ekin is the
energy in the order of kiloelectronvolt or higher), the domi- initial kinetic energy of the neutron. Hence, the most effec-
nating interaction is the scattering. In elastic scattering (n,n), tive materials for slowing down neutrons (or to moderate
the mechanism of indirect ionization comes from the fact them) are those with high hydrogen content.
that the incident neutron transfers kinetic energy to the recoil At room temperature, i.e., with mean kinetic energies of
nucleus. If instead the collision is inelastic (n,n′), the neutron kT ≃ 0.025 eV for T = 300 K (where k is the Boltzmann con-
enters the target nucleus which then re-emits it at a different stant), the neutron absorption dominates over the scattering
angle with respect to the original direction. When the neu- in some materials. The absorption cross section of thermal
a dx neutrons is particularly high for Boron-10. For instance,

nuclear reactions like 10B(n,α)7Li are often exploited in a
special type of radiation therapy called Boron neutron cap-
ture therapy (BNCT).
µ
attenuated beam
Incident beam
References
x 1. ICRU. Report 85. Fundamental quantities and units for ionizing

radiation (revised). J ICRU. 2011;11(1):1–31.
2. Volterrani D, Erba PA, Mariani G, editors. Fondamenti di medicina
b Φ(x)/Φ0
nucleare. Tecniche e applicazioni. Milan: Springer; 2010.
1.0 3. Firestone RB, Baglin CM, Chu SYF. Table of Isotopes. 8th ed.
0.9 New York, NY: Wiley and Sons, Inc; 1999.
4. de Marcillac P, Coron N, Dambier G, Leblanc J, Moalic
0.8
J. Experimental detection of α-particles from the radioactive decay
0.7 of natural bismuth. Nature. 2003;422:876–8.
0.6 5. Chu SYF, Ekström LP, Firestone RB. The Lund/LBNL Nuclear
Φ=Φ0/2 0.5
Data Search, Version 2.0, February 1999. http://nucleardata.
nuclear.lu.se/toi/nucSearch.asp.
0.4 6. Podgorsak EB, editor. Radiation oncology physics. A handbook for
0.3 teachers and students. Vienna: International Atomic Agency; 2005.
Φ=Φ0/4
0.2 7. IAEA. Technical report series no. 465. Cyclotron produced radio-
Φ=Φ0/8 nuclides: principles and practice. Vienna: International Atomic
0.1
Energy Agency; 2008.
0.0 8. IAEA. Technical document no. 1065, Production technologies for
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 molybdenum-99 and tecnetium-99m. Vienna: International Atomic
µx Agency; 1999.
x/HVL=1 x/HVL=3 9. Gould P. Medical isotope shortage reaches crisis level. Nature.
x/HVL=2 2009;460:312–3.
10. Jødal L, Le Loirec C, Champion C. Positron range in PET imaging:
Fig. 1.10 (a) Operative definition of the linear attenuation coefficient an alternative approach for assessing and correcting the blurring.
(see text for details). (b) Exponential attenuation curve as a function of Phys Med Biol. 2012;57:3931–43.
the attenuation (μx). The half-value layer (HVL) is the thickness for 11. Boone JM, Chavez AE. Comparison of x-ray cross sections

which the initial photon fluence is halved for diagnostic and therapeutic medical physics. Med Phys.
1996;23:1997–2005.
Key Learning Points

• In the energy range relevant for medical applica-
tions, neutral particles interact with atomic elec-
trons and nuclei of the target material mainly by
coherent (Rayleigh) or incoherent (Compton) scat-
tering, photoelectric absorption, or e+/e− pair
production.
• Some of these interactions lead to the production of
secondary charged particles, which in turn are
responsible for almost all the ionizations: for this
reason, neutral particles are called indirectly ioniz-
ing radiations.
• Neutrons interact mainly with the nuclei of the tar-
get material through elastic or inelastic collisions
(in case of fast neutrons) or by neutron capture or
absorption (for thermal neutrons).
Single-Photon-Emitting
Radiopharmaceuticals 2
Federica Orsini, Erinda Puta, Federica Guidoccio,
Contents
Overall Background
2.1 22
General Localization Mechanisms for Single-Photon-Emitting Radiopharmaceuticals
2.2 23
2.3 Tc-Labeled Radiopharmaceuticals
99m
28
2.3.1 99mTc-Sodium Pertechnetate 29
2.3.2 Tc-Bisphosphonates
99m
30
2.3.3 Tc-Diethylene Triamine Penta-Acetic Acid (99mTc-DTPA)
99m
31
2.3.4 Tc-Mercapto-Acetyl-Triglycine (99mTc-MAG3)
99m
32
2.3.5 Tc-Dimercaptosuccinic Acid (99mTc-DMSA)
99m
33
2.3.6 Tc-Radiocolloids
99m
33
2.3.7 Tc-Mannosyl-DTPA-Dextran (99mTc-Tilmanocept)
99m
36
2.3.8 Tc-Macroaggregated Albumin (99mTc-MAA)
99m
36
2.3.9 Tc-Hexa-Methyl-Propylene-AmineOxime (99mTc-HMPAO)
99m
37
2.3.10 99mTc-Ethylenediylbis-Cysteine-Diethylester (99mTc-ECD) 38
2.3.11 99mTc-Hexakis-2-Methoxy-2-Isobutyl-Isonitrile (99mTc-Sestamibi) 39
2.3.12 99mTc-6,9-Bis(2-Ethoxyethyl)-3,12-Dioxa-6,9-Diphospha-Tetradecane (99mTc-Tetrofosmin) 40
2.3.13 99mTc-Labeled Iminodiacetic Acid (IDA) Derivatives (99mTc-IDA) 41
2.3.14 99mTc-Sulesomab 42
2.3.15 99mTc-EDDA/HYNIC-[Tyr3]-Octreotide (TOC) and Other Somatostatin Analogs 42
2.4 Radiopharmaceuticals Labeled with Indium-111 44
2.4.1 111In-DTPA-Octreotide (111In-Pentetreotide) 44
2.4.2 In-DTPA
111
45
2.5 Radioiodinated Imaging Agents 45
2.5.1 123I/131I-Sodium Iodide 45
2.5.2 I-Ortho-Iodo-Hippuric Acid (123I-Hippuran)
123
46
2.5.3 Radioiodinated Meta-Iodo-Benzyl-Guanidine (MIBG) 47
2.5.4 I-N-ω-FluoroPropyl-2βCarbomethoxy-3β-(4-Iodophenyl)-nor-Tropane (123I-FP-CIT)
123
49
2.5.5 I-Iodobenzamide (123I-IBZM)
123
50
Tl-Chloride
2.6
201
51
Ga-Citrate
2.7
67
51
2.8 Lung Ventilation Radiopharmaceuticals 52
2.8.1 99mTc-DTPA Aerosol 52
2.8.2 Tc-Technegas
99m
53
2.8.3 Xenon-133 53
F. Orsini (*) · E. Puta

Nuclear Medicine Unit, “Maggiore della Carità” University
Hospital, Novara, Italy
e-mail: federica.orsini@maggioreosp.novara.it
F. Guidoccio · G. Mariani
Regional Center of Nuclear Medicine, Department of Translational
Research and Advanced Technologies in Medicine and Surgery,

22 F. Orsini et al.
2.9 Other Radiopharmaceuticals 54

2.9.1 Perspectives in Molecular Imaging Based on Single-Photon-Emitting Radiopharmaceuticals 54
References 55
Learning Objectives Radiopharmaceuticals consisting of only the radionuclide

• Learn the general definition of “radiopharmaceutical” and in its ionic form usually concentrate in specific target tissues/
“radiolabeling.” organs because either they are radioisotopes of a naturally
• Learn the main physical and chemical characteristics of occurring element of biologic interest (such as thyroid imag-
radionuclides emitting γ-rays commonly employed for ing with 123I− or 131I−—which undergoes the same physio-
single-photon imaging, including 99mTc, 111In, 123I, 131I, logic distribution as nonradioactive native iodine-127) or
201
Tl, and 67Ga. they share chemical similarities/analogies with native ele-
• Distinguish the different mechanism(s) of localization of ments of biologic interest (such as the 201Tl+ ion—mirroring
the most important radiopharmaceuticals employed in the physiologic distribution of the native K+ ion). For more
conventional nuclear medicine imaging. complex radiopharmaceuticals, the radionuclide is linked to
• Understand the favorable physical characteristics of 99mTc the main core/carrier molecule through a chemical process
for gamma camera imaging and of the 99Mo/99mTc genera- denominated “radiolabeling.”
tor for distribution logistics. The diagnostic information provided by scintigraphic
• Become familiar with the wide spectrum of available images derives from the specific distribution of a radiophar-
radiopharmaceuticals commonly used in clinical practice maceutical within the body, usually injected intravenously
for gamma camera imaging, with special attention to their (less often by inhalation, orally, interstitially, intracavitarily,
clinical indications and in vivo pharmacokinetics. or intra-arterially).
• Become familiar with the estimates of radiation dosime- The main parameters derived from serial scintigraphic
try to normal tissues/organs for each diagnostic images are delivery and clearance of the tracer from the
radiopharmaceutical. organ or tissue. Pathophysiologic changes induce modifica-
• Understand the concept of “molecular target” and of tions of these parameters (especially retention and/or the
“theranostics.” absence of retention of the radiopharmaceutical in the tissue/
organ under examination).
The radionuclides most frequently used for diagnostic
2.1 Overall Background applications in conventional nuclear medicine are the iso-
topes of technetium, iodine, indium, gallium, and thallium.
Nuclear medicine images depict anatomic, functional, and Their main physical characteristics are summarized in
metabolic processes in tissue. To create these images, specific Table 2.1.
compounds called radiopharmaceuticals are administered to
the patient. A radiopharmaceutical generally consists of a
biochemical core, which confers certain biological properties
that dictate its biodistribution. This biochemical core is chem-
Table 2.1 Main physical characteristics of radionuclides used for
ically linked to the radionuclide emitting the radioactive sig-
single-photon imaging, arranged according to increasing half-life of
nal that can be detected from outside the body, usually in the physical decay
form of γ-rays (also called “photons”); in some cases, it is the
Energy of major photon
ionic form per se of the radionuclide that determines its Radionuclide Decay T1/2 emissions (keV)
biodistribution. 99m
Tc 6.0 h 141 (89%)
Imaging with single-photon-emitting radionuclides 123
I 13.2 h 159 (83%); 528 (1%)
(radioisotopes of elements) produces both planar images and 111
In 2.80 days 171 (90%); 245 (94%)
single-photon emission computed tomography (SPECT) 201
Tl 3.04 days 68–82 (88%); 135 (3%); 167 (10%)
using a gamma camera. Radiopharmaceuticals labeled with
67
Ga 3.26 days 93 (39%); 185 (21%); 300 (17%)
positron-emitting radionuclides are used for positron
133
Xe 5.25 days 81 (36.5%)
131
I 8.0 days 284 (6%); 365 (82%); 637 (7%)
emission tomography imaging (as described in Chap. 3 of 51
Cra 27.7 days 5 (20%); 320 (10%)
this book “Positron-emitting radiopharmaceuticals”),
I
125 a
59.9 days 27 (14%); 31 (26%); 36 (7%)
while radiopharmaceuticals emitting predominantly β- or a
Not useful for actual external imaging in humans; used for specific
α-particles are used for therapeutic purposes (as described in application involving counting of biological samples with well-type
Chap. 4 of this book). γ-counters
2 Single-Photon-Emitting Radiopharmaceuticals 23
Radiopharmaceuticals labeled with 99mTc are used for The chemical and physical characteristics of a radiophar-
about 85% of diagnostic in single-photon imaging. This maceutical are the main factors determining its accumulation
radionuclide is widely employed clinically because: and retention in normal and diseased tissues of the organism.
This chapter classifies radiopharmaceuticals according to
• It can be obtained from commercial 99Mo/99mTc genera- radionuclide used for labeling and to the main mechanism(s)
tors in local radiopharmacies. of tissue localization responsible for the specific distribution
• Many kits are available to produce the radiopharmaceuti- properties of the imaging agent.
cal on-site with readily available equipment.
• The metastable radionuclide produces a 140 keV photon
in 88% of its nuclear decays (a conversion electron is pro-
duced in 12%). Key Learning Points
• Its γ-energy (140 keV) is favorable for detection by • Radiopharmaceuticals usually consist of a combi-
gamma cameras (whose detectors have excellent effi- nation of a biologic directing agent and of a radio-
ciency for γ-energies between 100 and 200 keV). nuclide emitting photons that can be detected
• Its relatively short half-life of 6 h represents a good com- outside the body using gamma cameras for
promise between the quality of images and radiation bur- imaging.
den to the patient. • The diagnostic information provided by radiophar-
maceuticals derives from their specific distribution
Biological-metabolic alterations typically occur at an within the body in physiologic and pathologic
early stage of disease prior to morphologic alterations in conditions.
shape, size, structure, or motion of the tissue/organ. On the • Alterations occurring at the molecular level detect-
other hand, this high sensitivity of nuclear medicine investi- able by functional imaging typically occur at an
gations for early changes is rarely matched by equally high early stage of disease, prior to morphologic altera-
specificity for a given disease. For instance, the increased tions in shape, size, structure, or motion of the tis-
metabolism of glucose in a particular region of the body can sue/organ.
indicate the presence of neoplastic cells with high prolifera- • The most common radionuclide used for single-
tive activity, but it may also indicate a condition of intense photon imaging is 99mTc; other radionuclides
inflammation due to infection or immunologic attack. employed in conventional nuclear medicine include
Since radiopharmaceuticals are typically used in very I, I, In, 67Ga, and 201Tl.
131 123 111
small mass amounts (in the order of micro-, nano-, or pico- • Since radiopharmaceuticals are typically used in
moles), they generally do not cause any disturbance of the very small mass amounts, they generally do not
biological system under investigation (except for possible cause any disturbance to patients, except for possi-
radiobiological effects due to the radioactive emission). ble radiobiological effect.
Quality control procedures applied to nonradioactive • Quality control procedures applied to nonradioac-
drugs are applicable to radiopharmaceuticals. Quality control tive drugs must be applied also to
tests that ensure the purity, potency, product identity, biologic radiopharmaceuticals.
safety, and efficacy of radiopharmaceuticals are mandatory • The quantity of radiopharmaceutical used in a spe-
for their clinical use. Moreover, because of the radiation com- cific procedure should be “as low as reasonably
ponent, radionuclidic and radiochemical purity must also be achievable” (ALARA), especially in pediatric
tested. The equipment and procedures necessary for these patients.
quality control tests are described in details in Chaps. 39 and • Radiopharmaceuticals can be classified according
40 of this book (for single-photon emitting and for positron- to the radionuclide used for labeling and to the main
emitting radiopharmaceuticals, respectively). mechanism(s) of tissue localizations.
The quantity of radioactivity administered varies accord-
ing to the procedure. Examinations based on the use of ion-
izing radiation must be optimized so as to deliver a radiation
dose “as low as reasonably achievable” (ALARA) consistent 2.2 eneral Localization Mechanisms
G
with the diagnostic goal. for Single-Photon-Emitting
Metabolism, biodistribution, and excretion of drugs are Radiopharmaceuticals
different in children from those in adults; the European
Association of Nuclear Medicine (EANM) and the Society Localization of most radiopharmaceuticals is not limited to
of Nuclear Medicine and Molecular Imaging (SNMMI) have one simple mechanism but also depends on the mode of
developed pediatric dosage cards that take into account the administration and delivery to the tissue. For some radio-
age and body weight for determining the activity of radiopharmaceuticals there is more than one single localization
pharmaceuticals for administration to children [1]. mechanism involved.
24 F. Orsini et al.
The principle of compartmental localization or dilution The mechanical trapping mechanism is also used for the
is at the basis of the applications of nondiffusible radio- preliminary evaluation of liver perfusion before trans-arterial
pharmaceuticals. This definition refers to the situation radionuclide therapy with 90Y-labeled microspheres in
where the molecules of interest are distributed in an patients with inoperable liver malignancies; in this case,
enclosed volume (called a compartment). In case of uni- 99m
Tc-MAA is injected directly into the arterial branches of
form radiopharmaceutical dispersion, the principle of com- the hepatic artery.
partmental dilution is exploited for calculation of the The mechanism of chemisorption is used in bone scans. For
whole-body erythrocyte mass (red blood cells labeled with example, 99mTc-labeled diphosphonates accumulate at the sur-
51
Cr) or of whole-body plasma volume (125I-radioiodinated face of newly formed calcium hydroxyapatite crystals and in
serum albumin). These methods are based on the principle the amorphous bone mineral matrix; their uptake is increased
of dilution of a known amount of radioactive substance in in areas with increased new bone formation (e.g., fracture heal-
an unknown distribution volume; in particular, by measuring, infection, primary and metastatic tumors) (Fig. 2.1).
ing the degree of dilution (i.e., change in concentration) of Chemisorption is quite strong, intermediate between chemical
the amount of radioactivity administered, it is possible to covalent binding and hydrogen binding (adsorption); hence the
calculate the volume in which the radiopharmaceutical has term chemisorption is used. In addition to chemisorption on the
been diluted. bone surfaces, there can also be chemisorption onto calcium
Red blood cells labeled with a radionuclide (e.g., 99mTc) phosphate crystals in dystrophic calcium deposits that precipi-
also allow to record in vivo images of the heart at systole and tate in certain soft tissues, e.g., as a consequence of severe
diastole and to measure chamber volume at the end of dias- hyperparathyroidism and hypercalcemia.
tole and systole determinig the left ventricular ejection Autologous radiolabeled leukocytes localize at sites of
fraction. infection based on chemotactic signals. The chemotaxix is the
In some other conditions, there can be an abnormal leak- characteristic movement or orientation of an organism or cell
age of contents from the compartment. For instance, scintig- (including leukocytes) along a chemical concentration gradient
raphy with 99mTc-labeled red blood cells can identify the site either toward or away from the chemical stimulus and tipically
of gastrointestinal bleeding as an area of increased radioac- occurs in the sites of inflammation and infection, involving
tivity accumulation at some point in the gastrointestinal immune cells, blood vessels, and molecular mediators.
lumen—where the radiolabeled erythrocytes do not normally Phagocytosis is the process whereby microparticles are
accumulate. Similarly, scintigraphy with a radiopharmaceu- internalized from the extracellular space into cells, a process
tical that normally undergoes hepatobiliary clearance and that is especially active in macrophages. Although macro-
excretion (e.g., 99mTc-mebrofenin) can visualize abnormal phages are ubiquitous cells, they are concentrated mostly in
biliary leakage into the abdominal cavity, e.g., after biliary the reticuloendothelial system, especially in the liver,
surgery. spleen, bone marrow, and lymph nodes, as well as at sites of
In other conditions, images can show, within a certain infection/inflammation. Phagocytosis is facilitated by sev-
tissue/organ, areas of radioactive concentration that are eral properties of the particles, such as size between 2.5 nm
reduced with respect to the surrounding (supposedly nor- and 1 μm, negative charge, and possible opsonization by a
mal) tissue/organ. Such areas are often the result of an class of macromolecules that include the complement frac-
obstruction in a compartmental space. For instance, venti- tions C3, C4B, and C5, as well as some α- and β-globulins.
lation lung scintigraphy after inhalation of the radioactive Phagocytosis is responsible for the accumulation of 99mTc-
gas xenon-133 can indicate the presence of an obstruction labeled particles in the nanocolloidal size range (e.g., albu-
in the lung airways as an area of reduced/absent radioac- min nanocolloids or sulfur colloid preparations) at different
tivity accumulation in the bronchi/bronchioles beyond the sites depending on the route of administration. In fact, these
point of obstruction. radiocolloids can be used for imaging the liver, spleen, and
Macroaggregated albumin (MAA) particles, ranging in bone marrow following systemic administration (i.e., an i.v.
diameter from 10 to 90 μm (with >90% of particles ranging injection). On the other hand, radiocolloids injected intersti-
from 10 to 40 μm), can be radiolabeled (usually with 99mTc) tially migrate through the lymphatic system. This mecha-
and injected intravenously to depict the distribution of nism is the basis of lymphoscintigraphy, a preliminary step
regional pulmonary blood flow. The particles become to, e.g., radioguided sentinel lymph node biopsy.
mechanically trapped in the capillary bed of the lungs, thus Many radiopharmaceuticals accumulate in cells through
enabling to utilize the principle of microembolization to transmembrane transport mechanisms (either passive or
identify the regions of pulmonary parenchyma with normal active transport). While passive transport does not require
perfusion. Regions with reduced perfusion (typically in case any energy expenditure and only involves diffusion of a
of thromboembolic disease) are depicted as areas with substance from areas of high concentration to areas of low
decreased/absent radioactivity. concentration (concentration gradient), on the other hand,
a b
Fig. 2.1 Examples of planar whole-body acquisitions of bone scintig- out any areas of abnormally increased tracer uptake. (b) The presence
raphy with 99mTc-MDP. (a) Normal pattern of skeletal scintigraphy of numerous focal areas with markedly increased tracer uptake through-
showing physiologic turnover of the mineral matrix of the bone, without the skeleton in a patient with multiple skeletal metastasis
active transport requires energy (in the form of ATP or an 99m

Tc-tetrofosmin) that cross cell membranes following both
electrochemical gradient of Na+ or H+) for the system to a concentration gradient and an electrochemical gradient.
function. These positively charged molecules are attracted by mole-
Passive transport generally occurs by diffusion, i.e., when cules or by structures in the cell that are negatively charged.
a substance moves from an area of high concentration to an Since diffusion is not a unidirectional process, intracellular
area of low concentration. A substance, typically lipophilic, accumulation typically involves some additional
crosses the phospholipid bilayer of cell membranes. Simple mechanism(s) of retention. In the case of 99mTc-tetrofosmin,
diffusion is influenced by pH/ionization state. In fact, many the radiopharmaceutical binds to intracellular polar metabo-
molecules may exist in either a neutral state or as a charged lites or complex carriers of the cytosolic components, while
ion, depending on pH. Hence, a molecule may be able to dif- 99m
Tc-sestamibi binds to the cell membrane of mitochondria.
fuse across a membrane in its non-ionized lipophilic form This process results in stable intracellular binding and mini-
but cannot diffuse across the same membrane in its ionized mizes washout of the tracer.
hydrophilic form. Membranes have small pores that limit the The tracers used to image cerebral perfusion, 99mTc-
size of molecules that can cross the membrane (molecular exametazime (99mTc-HMPAO) or 99mTc-bicisate (99mTc-
weight <80 kDa). It follows that passive diffusion is nonse- ECD), are electrochemically neutral compounds. Due to
lective, not competitively inhibited by similar molecules, their lipophilicity, these uncharged molecules cross the
and not subject to saturation. blood-brain barrier by passive diffusion following a con-
An example of simple diffusion as a mechanism of radio- centration gradient. Once in the intracellular space, the
pharmaceutical uptake is represented by 99mTc-labeled myo- radiopharmaceuticals undergo an enzymatic process (or
cardial perfusion agents (either 99mTc-sestamibi or chemical transformation) that renders them hydrophilic,
26 F. Orsini et al.
a b
Fig. 2.2 Examples of different thyroid conditions visualized by scin- nodule of the left thyroid lobe with complete inhibition of the remain-
tigraphy. (a) Scintigraphy with 123I-iodide: multinodular goiter with ing parenchyma uptake, typical of hyperfunctioning thyroid adenoma
both bilateral “cold” and “hot” nodules (image obtained using a pinhole (image obtained using a parallel-hole, high-resolution collimator, zoom
collimator). (b) Scintigraphy with 99mTc-pertechnetate: single “hot” factor 2)
therefore not capable of passively diffusing back to the A well-known example of active transport is the uptake of
extracellular space. This is therefore a process of “bio- iodide in the thyroid gland. Radioisotopes of iodine in ionic
chemical trapping” that causes radiopharmaceutical reten- form (123I−, 124I−, 125I−, 131I−) and other anions (99mTcO4−) are
tion into the neurons. transported into thyroid cells via the so-called Na+/I− sym-
When transmembrane diffusion of a substance involves porter (NIS); these radiopharmaceuticals are employed for
a specific carrier system, the phenomenon is called facili- morphofunctional evaluation of the thyroid gland (Fig. 2.2).
tated diffusion. Such carriers are generally selective and High concentrations of iodide in the blood, such as those
can be competitively inhibited by an excess amount of sim- resulting from, e.g., recent administration of iodinated X-ray
ilar molecules; in general, the process is also subject to contrast media, competitively inhibit thyroid uptake of these
saturation, when all the sites that activate transport are thyroid-imaging agents.
occupied. Another example of active transport is the transmembrane
Facilitated diffusion is the process responsible for trans- transfer of the ionic form of radioactive potassium or its ana-
port of the glucose analog [18F]FDG into cells. A series of log (201T1+) facilitated by the Na+/K+ transmembrane pump.
glucose transporters, GLUT1-GLUT5, are responsible for This pump system is active in all cells, to maintain the high
carrying glucose (or its analog 2-fluorodeoxyglucose) across intracellular concentration of potassium. Tl+ has an ionic
the cell membrane (see also Chap. 3 of this book). Once radius similar to K+ and is able to fit in the Na+/K+ pump.
inside the cells, an enzyme that physiologically initiates Uptake in the myocardium reflects myocardial perfusion and
intracellular metabolism of glucose (hexokinase) converts viability.
[18F]FDG into [18F]FDG-6-phosphate; this compound cannot The kidney has two mechanisms for transferring drugs
enter the glycolytic pathways that glucose-6-phosphate fol- from the blood into the urine: (1) glomerular filtration is a
lows, thus remaining trapped inside the cells. Also nonpolar special case of diffusion involving the transit of molecules
lipophilic anionic radiopharmaceuticals used for hepatobili- through pores, or channels, regulated by hydrostatic or
ary scintigraphy (derivatives of 99mTc-iminodiacetic acid) osmotic pressure gradients and by molecular size (a tracer
accumulate in hepatocytes through transmembrane carriers which is filtered, such as 99mTc-DTPA, is used clinically to
present on the hepatocyte cell surface; inside the hepato- measure the glomerular filtration rate), and (2) tubular
cytes, they are then conjugated and excreted into the bile. secretion is a special case of active transport out of glands
Active transport generally uses energy-dependent carriers directly into the renal tubules. A tracer that utilizes both
that transfer molecules against a concentration gradient, i.e., mechanisms, such as 99mTc-mertiatide (99mTc-MAG3),
from an area of low concentration to an area of high concen- undergoes partial glomerular filtration, followed by tubu-
tration. It is selective, can be competitively inhibited, and can lar secretion, resulting in a higher concentration in the
be saturated. kidney.
An additional mechanism of cellular interaction is based Radiolabeled antibody preparations have been developed
on the expression receptors on the surface of cell membranes. to recognize specific target epitopes expressed on the cell
Over 50 different classes of receptors have been identified, surface of tissues either physiologically or following malig-
and their number is continuously increasing. Radioactive nant transformation of cells. Accumulation of such radio-
drugs accumulating into cells through a receptor-mediated pharmaceuticals at certain sites of interest (tumor tissues) is
mechanism compete with the endogenous receptor-specific therefore based on an immunological mechanism.
ligands for binding to their respective receptor; therefore, the Radiolabeled monoclonal antibodies (or molecularly engi-
diagnostic or therapeutic use of these radiopharmaceuticals neered fragments, such as fab, fab′, and fc) are directed
is based on their high binding specificity (in the nanomolar against epitopes expressed on the surface of neoplastic cells
range) which, due to the limited availability of binding sites, or glycoproteins in the tumor stroma (extracellular matrix).
mainly depends on specific activity of radiopharmaceutical Antibody imaging can be used for therapy or diagnosis.
preparation. Specific activity of radiopharmaceuticals is When radiolabeled antibodies are employed for radioimmu-
defined as the radioactivity of a given per unit mass. In low notherapy of tumors, they are usually labeled with radionu-
specific activity radiopharmaceuticals, the amount of unla- clides emitting β-particles with suitable energies (as better
beled tracer is high, thus possibly resulting in saturation of detailed in Chap. 4 of this book “Radiopharmaceuticals for
the binding sites with the nonradioactive drug and conse- therapy”). When antibodies are used for diagnosis, they are
quent poor image quality. Single-photon radiopharmaceuti- labeled with the positron-emitting radionuclide 68Ga or sin-
cals accumulating in tissues/organs through a gle-photon-emitting 111In or 99mTc.
receptor-mediated mechanism include analogs of somatosta- Finally, there is a heterogeneous group of radiopharma-
tin (such as 111In-pentetreotide or 99mTc-EDDA/HYNIC- ceuticals that localize at specific sites (either extracellular or
TOC, used to investigate neuroendocrine tumors), intracellular) because they behave as ligands that bind with
99m
Tc-mannosyl-DTPA-dextran (a receptor-targeted agent high (and avid) chemical affinity to specific components
injected interstitially for lymphatic mapping that binds either in the extracellular space (e.g., the β-amyloid fibrils
avidly to the CD206 receptors for mannose on the surface of typical of Alzheimer’s disease) or in the intracellular space
macrophages and dendritic cells in lymph nodes), and (such as mitochondria, employed for imaging, e.g., myocar-
123
I-ioflupane, a chemical derivative of cocaine that binds to dial blood flow). Agents targeting intracellular components
presynaptic dopamine transporters (primarily located in the must be lipophilic to diffuse into cells (or cross the blood-
caudate and putamen) and is used to investigate patients brain barrier). Once in the cell, the agents need to either bind
with movement disorders (Fig. 2.3). Many other radiophar- to some intracellular component or undergo a reaction that
maceuticals are based on the receptor-mediated localization reduces the lipophilicity to trap the molecule in the cell.
mechanism, most of them labeled with positron-emitting Many of these imaging agents are derived from dyes used for
radionuclides. staining histological preparations.
Fig. 2.3 Axial SPECT

images obtained at different a
levels during receptor brain
scintigraphy with
123
I-ioflupane. (a) Physiologic
pattern of uptake by the
presynaptic dopamine
transporters, with normal
bilateral visualization of basal
ganglia. (b) Scintigraphy in a
patient with Parkinson’s
disease, showing an abnormal
pattern of reduced and
asymmetrical uptake in the b
basal ganglia
28 F. Orsini et al.
resin, which constitutes the stationary phase of the system usu-

Key Learning Points ally designed as a small glass column (about 40 mm long and
• Localization of most radiopharmaceuticals is not less than 10 mm wide). 99mTc is generated during the continu-
limited to one simple mechanism but also depends ous decay of 99Mo, and it is periodically removed from the
on the mode of administration and delivery to column by elution of the column with physiological saline.
tissues. 99m
Tc is produced in the chemical form of 99mTc-pertechnetate
• For some radiopharmaceuticals, there is more than anion (99mTcO4−); the chloride ions present in the saline solu-
one single localization mechanism involved. tion exchange with the 99mTcO4− ions in the column, thus per-
• The main principles of radiopharmaceutical accumu- mitting their elution in physiological saline.
lation include compartmental dilution of nondiffus- The regeneration process proceeds exponentially, and in
ible radiopharmaceuticals, leakage of compartmental about one half-life (6 h), 50% of the 99mTc is regenerated.
localization, mechanical trapping, chemisorption, After 24 h 99Mo and 99mTc reach a new balance, and the gen-
chemotaxis, phagocytosis, transmembrane transport erator is ready to be eluted again. However, since 99Mo has
(including active transport and passive transport— decayed in the meantime, the amount of 99mTc obtained with
either simple or facilitated diffusion), filtration, the second elution will be about 70% of the amount obtained
secretion, receptor-mediated mechanism, immuno- with the prior elution.
logical mechanism, and chemical affinity. Two types of generators are currently available commer-
cially, known as “wet” and “dry” generators, respectively. The
wet generators have a large reservoir of saline permanently
attached to the inlet. Each elution pulls liquid from the column
2.3 99mTc-Labeled Radiopharmaceuticals by a vacuum in an evacuated vial attached to the outlet of the
chromatography column, this liquid being replaced by an
The 99Mo/99mTc generator makes this transition metal almost equal volume of saline from the reservoir. Thus, the column
ideal for diagnostic nuclear medicine. A radioisotope genera- remains full of saline between two elution procedures. In a dry
tor is a system where a “parent” radionuclide with a rela- generator, elution is performed with a set volume of saline
tively long half-life decays into a “daughter” radionuclide provided each time by attaching a vial to an inlet to the genera-
generally with a shorter half-life. The parent radionuclide is tor. Similarly as with the wet generator, saline is pulled through
generally bound in an insoluble form to the stationary phase the column by a vacuum provided by an empty evacuated vial
of a chromatographic column, while the daughter radionu- attached to the outlet of the chromatography column. The liq-
clide is eluted from the column. There is usually an optimal uid pulled through the column is followed by an additional
interval of time between elutions to allow generation of a volume of air to effectively remove most of the liquid from the
sufficient amount of radioactive daughter. 99Mo/99mTc gen- column. This will not render the column completely free of
erators can be easily transported over long distances to hos- liquid, but the bulk of the saline will be pulled away from the
pitals; the in situ produced 99mTc is available in the chemical column.
form of 99mTc-pertechnetate (99mTcO4−) that can be used for Each 99mTc-labeled radiopharmaceutical is chemically
scintigraphic imaging as such (e.g., for thyroid scintigraphy, defined as a “coordination complex,” i.e., a compound
salivary scintigraphy, detection of ectopic gastric mucosa in formed by a central transition metal (coordination center)
Meckel’s diverticulum) or for instant labeling of different bonding with “coordinate covalent bonds” to other sur-
kits to produce other radiopharmaceuticals. rounding molecules, defined as “ligands or complexing
Most of the 99Mo utilized for the 99Mo/99mTc generators agents.” Technetium is the transition metal, while ligands
had been produced through purification of products from can be single atoms such as chlorine, bromine, oxygen, or
nuclear fission of highly enriched uranium fission products. nitrogen or larger molecules such as ammonia, water, car-
Recently, due to threats from terrorists, alternative approaches bon monoxide, and amino acids. The fundamental require-
to produce 99Mo, through the use of a linear accelerator of ment for a ligand to be labeled with 99mTc is to possess an
low-enriched uranium targets, have gained favor. appropriate set of atoms able to firmly coordinate to the
99
Mo decays with a physical half-life of 2.75 days, mainly metallic center.
with the emission of high-energy β-particles and of γ-rays In a complex, the metal is capable of forming more than
with a wide energy spectrum; the end product of 99Mo decay one bond with as many ligands as the coordination num-
is 99mTc. ber. For example, technetium can form complexes with
The 99Mo/99mTc generator consists of an ion exchange chro- coordination numbers of 4, 5, 6, or 7 (most commonly 5
matographic column. The “parent” radionuclide (99Mo, which and 6).
is insoluble in water or in physiological saline) is firmly bound An important property of the coordination complex is its
in the form of molybdate to a chromatography anion exchange molecular geometry. Each complex can assume a particular
geometric conformation, the transition metal remaining in 2.3.1 99mTc-Sodium Pertechnetate

the center. Octahedral and pyramidal geometries are those
more frequently found in the technetium complexes. 99m
Tc-sodium pertechnetate (Na99mTcO4, present as the
Another important parameter that characterizes a coordi- 99m
TcO4− ion in solution at physiological pH) is obtained
nation complex is the electrical charge, determined by the directly after 99Mo/99mTc generator elution with physiologi-
metal oxidation state. The higher the charge, the greater the cal saline.
hydrophilicity. An uncharged complex is usually the most After i.v. administration, the pertechnetate ions circulate in
lipophilic. the blood partly in their free form and partly bound to plasma
99m
TcO4− is a coordination compound with a very com- proteins; because of their small size, free ions easily leave the
pact structure with tetrahedral geometry and oxidation state vascular compartment and migrate to the interstitial fluids. As
+7. Pertechnetate, the highest possible oxidation state for their concentration in blood diminishes, other 99mTcO4− ions
this metal, is also a very stable chemical species in aqueous are released from their binding to plasma proteins. Therefore,
solution. Thus, in order to label any given molecule with visualization of the vascular space observed in early scinti-
99m
TcO4−, the oxygen atoms bound to the metal must be graphic acquisitions declines gradually over time.
removed through a reducing agent and must be replaced with Pertechnetate ions are concentrated in tissues due the
new ligand-coordinated atoms. During this process, the oxi- activity of anion pumps in the thyroid, stomach, salivary
dation state of technetium is reduced to values lower than +7. glands, intestine, choroid plexus, and kidney. 99mTc-Pertech-
The stannous ion (Sn2+) is generally used as a reducing agent, netate (Fig. 2.4) is commonly used for thyroid scintigraphy,
being added in aqueous solution as a chloride salt (SnCl2) in for the search of ectopic gastric mucosa, and for salivary
such small amounts that it does not alter solubility of the gland scintigraphy.
preparation or cause toxicity to the patient. The oxygen The anion pump responsible for localization in the thyroid
atoms are removed from 99mTcO4− through the formation of is called the Na+/I− symporter (NIS). NIS is a transmembrane
Sn(OH)4 and other similar compounds. The atom of techne- protein consisting of 643 amino acids arranged in 13 domains
tium is now ready to coordinate with a specific binder or a with extracellular amino-terminal and intracellular carboxy-
chelating agent that has been pre-attached to the molecule to terminal (molecular weight ranging between 79 and 90 kDa,
be labeled. The binder must have a high technetium coordi- depending on the degree of glycosylation). The pertechnetate
nating ability, in order to ensure stability of the 99mTc-labeled ion mimics the iodide ion, since 99mTcO4− is similar in terms of
radiopharmaceutical and prevent it from recombining with mass, size, and charge density; however, once inside the thy-
the oxygen atoms in aqueous solution and thus return back to roid cell, 99mTcO4− does not undergo subsequent organification
the anion pertechnetate or form technetium dioxide (TcO2), a as iodide does to participate in the synthesis of thyroid hor-
secondary compound of oxygen that tends to form colloidal mones. NIS is located on the basolateral membrane of thyroid
particles. cells, and it is able to simultaneously transport sodium and
Key Learning Points

• The 99Mo/99mTc generator is an ion exchange chro-
matographic column where the “parent” radionu-
clide (99Mo) with a relatively long half-life
(2.75 days) decays into the “daughter” radionuclide
(99mTc) with a shorter half-life (6 h) and becomes
available for elution in the chemical form of 99mTc-
pertechnetate (99mTcO4−).
• “Wet” and “dry” generators are currently available
commercially.
• 99mTc-Pertechnetate is a transitional metal suitable
for scintigraphic imaging as such or for radiolabel-
ing of different ligands to produce a variety of
“coordination complexes,” in virtue of its tetrahe-
dral geometry and oxidation state +7. Fig. 2.4 Tridimensional chemical structure of the 99mTc-pertechnetate
anion (99mTcO4−). Color codes: yellow = 99mTc; red = O
30 F. Orsini et al.
iodide (with a stoichiometric ratio of 2:1) from the extracellular 2.3.2 99mTc-Bisphosphonates
space into thyroid cells. Expression of this protein is regulated,
among other minor mechanisms, mostly by the thyroid-stimu- 99m
Tc-Bisphosphonates, also called 99mTc-diphosphonates,
lating hormone (TSH, which upregulates the process) and by are employed for bone scintigraphy because of their high
the concentration of iodide in extracellular fluids (which down- uptake in the skeleton and rapid clearance from blood and
regulates the process). Extraction of iodine from plasma and its soft tissue after i.v. injection.
concentration in thyroid cells is a saturable and active process The molecules are called bisphosphonates because they
that takes place against an electrochemical gradient (resulting have two phosphonate (PO(OH)2) groups in their chemical
in an intracellular concentration of iodine 20- to 40-fold higher structure (Fig. 2.5). The P-C-P bond between the two phos-
than in plasma) and requires energy, which is supplied by the phates, which is similar to the P-O-P structure of native pyro-
Na+/K+-dependent ATPase system. phosphate, makes the molecule extremely resistant to
For the search of ectopic gastric mucosa in Meckel’s diver- hydrolysis by the enzyme phosphatase and to acidic condi-
ticulum, 99mTc-pertechnetate is employed because of its physi- tions. Slightly different chemical forms of radiolabeled
ologic secretion by the gastric mucosa. In fact, technetium is a bisphosphonates are available. The most commonly used
member of Group 7A of the periodic table of elements, and it diphosphonates (Fig. 2.6) are methylene diphosphonate
is reasonable to assume that gastric concentration of pertech- (99mTc-medronate or 99mTc-MDP), hydroxymethylene diphos-
netate might be similar to that of chloride (since chlorine is a phonate (99mTc-oxidronate or 99mTc-HMDP), hydroxyethylene
member of Group 7B). This assumption is supported by the diphosphonate (99mTc-HDP), and dicarboxypropane diphos-
correlation observed between acid output and pertechnetate phonate (99mTc-DPD). The agents are substantially equivalent
clearance. It has also been demonstrated that 99mTc-pertechne- regarding lesion detectability; minor differences are reported
tate is secreted also by the mucous epithelial cell [2]. among the different radiopharmaceuticals regarding the rate of
99m
Tc-Pertechnetate is the radiopharmaceutical of choice clearance from the blood, justifying different imaging times
for the scintigraphy of the salivary glands because of its after injection to obtain the best bone-to-soft tissue ratio.
chemical-physical analogy with the anions physiologically 99m
Tc-Bisphosphonates accumulate at sites of bone min-
present in saliva (mainly iodide) excreted by salivary glands. eral rearrangement by chemisorption on the surface of newly
The concentration mechanism for pertechnetate in the human formed hydroxyapatite crystals. Bone remodeling is an inte-
salivary glands is typical of an active ion transport system gral component of the osteoblastic reaction associated with
located in cells of the striated ducts as well as in acinar cells. most tumors metastatic to the bone, as well as primary bone
99m
TcO4− ions do not pass the intact blood-brain barrier (due tumors, infection in the bone, and response to trauma. Some
to the hydrophilicity of the charged molecule). Pertechnetate tumors (e.g., renal cell carcinoma) cause primarily osteolytic
does, however, concentrate in the choroid plexus. 99mTc- lesions. Pure osteolytic lesions may be difficult or impossi-
Pertechnetate has been used in the past for the scintigraphic ble to detect on bone scans.
evaluation of the integrity of the blood-brain barrier (which is Uptake of the radiopharmaceutical in the bone is also
disrupted in brain tumors, meningitis, brain abscess, and vascu- directly correlated with local vascularization, although not in
lar accidents). From a clinical perspective, pertechnetate imag- a linear relationship; in fact, a three- to fourfold increase in
ing to detect the integrity of the blood-brain barrier has been blood flow enhances uptake by only 30–40%. Furthermore, a
replaced by magnetic resonance and CT imaging. reduction of the sympathetic tone has vasodilator effects on
The main radiation dosimetry estimates to patients following capillary blood flow, thus increasing radiopharmaceutical
intravenous administration of 99mTc-pertechnetate are reported
here below, normalized to unit of administered activity [3]: O H O O H O
HO P C P OH HO P C P OH
• Effective dose 0.013 mSv/MBq
• Tissues/organs with the highest values of absorbed dose:
OH OH OH OH H OH
–– Upper large bowel 0.056 mGy/MBq
–– Colonic wall 0.041 mGy/MBq HMDP MDP
–– Stomach 0.026 mGy/MBq
O O
Key Learning Points HO P O P OH

• Tc-Pertechnetate mimics the iodide ion of the
99m
anion pumps (Na+/I− symporter). OH OH

• 99mTc-Pertechnetate as such is commonly used for Pyrophosphate
thyroid scintigraphy, for the search of ectopic gas-
tric mucosa, and for salivary gland scintigraphy. Fig. 2.5 Chemical structure of HMDP, MDP, and native
pyrophosphate
Fig. 2.6 Tridimensional
chemical structure of a
99m
Tc-labeled bisphosphonates
employed for bone
scintigraphy. (a) 99mTc-
HMDP. (b) 99mTc-MDP. Color
codes: yellow = 99mTc;
red = O; white = H;
purple = P; light blue = C
uptake, as observed in the case of reflex sympathetic dystro-

phy, sympathectomy, thrombosis, or hemiplegia.
After i.v. administration, the radiopharmaceutical rapidly
diffuses into the extracellular space, and uptake in bone
begins immediately. About 10% of the initial activity remains
in the circulation 30 min after injection, decreasing to about
1% by 4 h after administration. The mechanism of radiophar-
maceutical clearance from the body is through renal excre-
tion, with about 50% of the intact radiopharmaceutical being
excreted in the first 6 h after injection. The scan is usually
performed 2–3 h after administration, when an optimal tar-
get/background ratio is achieved.
The main radiation dosimetry estimates to patients fol-
lowing intravenous administration of 99mTc-bisphosphonates
derivatives are reported below, normalized to unit of admin-
istered activity:
• Effective dose 0.0049 mSv/MBq Fig. 2.7 Tridimensional chemical structure of 99mTc-DTPA. Color
• Tissues/organs with the highest values of absorbed dose: codes: yellow = 99mTc; red = O; white = H; light blue = C; blue = N
–– Urinary bladder wall 0.047 mGy/MBq
–– Bone surface 0.034 mGy/MBq
–– Kidney 0.0072 mGy/MBq
2.3.3 99mTc-Diethylene Triamine Penta-Acetic
Key Learning Points Acid (99mTc-DTPA)
• Tc-Bisphosphonates (or diphosphonates) accu-
99m
mulate at sites of bone mineral rearrangement by 99m

Tc-DTPA is a 99mTc-chelated compound with a molecular
chemisorption on the surface of newly formed weight of about 500 Da and a negative electrical charge
hydroxyapatite crystals. (Fig. 2.7); this small molecule is therefore able to easily dif-
• These compounds are commonly used for bone fuse in the extracellular fluids. It is excreted unmodified
scintigraphy. through the kidneys through glomerular filtration, and it is
not subject to tubular secretion and/or reabsorption.
32 F. Orsini et al.
Introduced in clinical practice in 1970 [4], 99mTc-DTPA is

still one of the most commonly used radiopharmaceuticals Key Learning Points
for dynamic renal scintigraphy, which provides information • Tc-DTPA is a small chelated compound with
99m
about renal perfusion, glomerular function, and urinary tract negative electrical charge able to easily diffuse in
obstruction. It is readily available, has relative low cost, and, the extracellular fluids.
above all, clinical benefits with good reproducibility. • It is excreted unmodified through the kidneys
Because of a relatively low binding to plasma pro- through glomerular filtration, and it is not subject to
teins (3–5%), accumulation of 99mTc-DTPA in the urine tubular secretion and/or reabsorption, so its main
reflects glomerular filtration. Quantitative analysis of use is for dynamic renal scintigraphy and GFR
the data acquired during dynamic renal scintigraphy calculation.
with 99mTc-DTPA allows calculation of the glomerular • It has also been employed for lung ventilation, brain
filtration rate (GFR), the main parameter of renal func- perfusion, and cysternoscintigraphy or
tion. Nevertheless, despite its very low plasma protein ventriculoscintigraphy.
binding, 99mTc-DTPA tends to yield slightly lower GFR
values than the reference values obtained by measuring
inulin clearance.
Besides its use for the evaluation of renal function and 2.3.4 99mTc-Mercapto-Acetyl-Triglycine
flow through the urinary tract, 99mTc-DTPA has been (99mTc-MAG3)
employed as an inert, low-molecular-weight tracer for
lung ventilation studies. When inhaled as an aerosol, Also known as 99mTc-betiatide or 99mTc-mertiatide, 99mTc-
99m
Tc-DTPA is not absorbed from the lung but remains in MAG3 contains, in addition to a coordinating structure able
the alveolus, reflecting the distribution of ventilation (see to bind 99mTc, an ionic carboxyl group (COOH) and a car-
section on “Lung ventilation radiopharmaceuticals” fur- bonyl group (C=O) (Fig. 2.8) binding to specific receptors
ther below). located primarily in the distal part of proximal tubular cells
Because of its inability to cross the intact blood-brain and to a lesser extent in the early part of the distal tubule. This
barrier, 99mTc-DTPA has also been used for scintigraphic radiopharmaceutical undergoes partial glomerular filtration
evaluation of cerebral perfusion, especially to confirm brain and active tubular secretion; it is therefore characterized by
death (by recording a dynamic planar first-pass acquisition high renal extraction and fast blood clearance. 99mTc-MAG3
in the first 60–120 s after administration of an i.v. bolus con- is commonly employed for dynamic renal scintigraphy [5].
taining up to 740 MBq) as well as the integrity of the blood- Compared to the tubular excretion agent 131I-hippuran (see
brain barrier (static acquisition about 1 h after further below), it has a γ-ray energy emission more favorable
administration). for gamma camera imaging and more favorable radiation
Specific investigations on the cerebrospinal fluid dynam- dosimetry; advantages versus 123I-hippuran include much
ics, such as cysternoscintigraphy or ventriculoscintigraphy, lower cost and wider availability.
are possible after intrathecal administration of 111In-DTPA Following i.v. injection, 99mTc-MAG3 rapidly distributes
(through the lumbar or suboccipital route). These investiga- in the extracellular fluids and is then cleared from the blood
tions aim to identify possible obstructions or pathological mainly by tubular secretion (>90%), without undergoing
leak of the cerebrospinal fluid (e.g., otorrhea or rhinorrhea) tubular reabsorption; the remaining amount of 99mTc-MAG3
and arachnoid or porencephalic cysts, in the case of hydro-
cephalus or brain malformation. When delayed acquisitions
extending up to 3 days are required, DTPA should be labeled
with 111In, which has a longer physical half-life more suitable O
for delayed imaging. N O
The main radiation dosimetry estimates to patients fol- O
lowing intravenous administration of 99mTc-DTPA are Tc N O
reported here below, normalized to unit of administered N S
activity:
OH
O
Fig. 2.8 Chemical and tridimensional structure of 99mTc-MAG3. Color
–– Uterus 0.0079 mGy/MBq
codes: yellow = 99mTc; red = O; white = H; light blue = C; blue = N;
–– Kidney 0.0044 mGy/MBq olive green = S
is cleared through glomerular filtration, despite a high frac- either the luminal membrane or directly by extraction from
tion of plasma protein binding (80%). There is also a minor the blood of the peritubular capillaries via the basocellular
fraction of excretion through the hepatobiliary tract, medi- membrane.
ated by uptake in hepatocytes that is independent from renal Following i.v. injection, 99mTc-DMSA initially distributes
function and mainly attributable to the lipophilic properties in the extracellular fluid, with temporary retention in the
of 99mTc-MAG3. Some of the minimal quantities (1–2%) of liver and spleen. About 90% of the radiopharmaceutical cir-
impurities formed during radiolabeling accumulate in the culates in the blood bound to plasma proteins, which pre-
liver and also undergo hepatobiliary excretion. vents its glomerular filtration (limited to 2–3% of total
Because of its high first-pass renal extraction (around injected activity). Its kidney extraction is 5% at each pas-
60%) due to the active tubular secretion mechanism, 99mTc- sage, and 1 h postinjection, approximately 50% of the admin-
MAG3 is particularly appropriate for use in pediatric-age istered activity is firmly retained in the proximal renal tubular
patients, in the evaluation of the transplanted kidney, and in cells (mainly in the cytoplasm), with a 22:1 cortex/medulla
subjects with impaired renal function. In these cases, the uptake ratio. Maximum renal cortical uptake of the radio-
high extraction fraction results in good images with a high pharmaceutical is reached at about 3 h. Static renal scintigra-
target-to-background ratio with low injected activity. phy acquired 1–3 h after injection provides excellent imaging
Thanks to the properties of this tracer, quantitative analy- of the renal cortex. Washout of 99mTc-DMSA is slow, so that
sis of the data acquired during dynamic renal scintigraphy 6 h after administration about 35% of injected activity is still
allows calculation of the tubular extraction rate (TER), an retained in the renal cortex. Very late images (6–24 h) are
important parameter of renal function. generally acquired only in the event of urinary tract obstruc-
The main radiation dosimetry estimates to patients fol- tion; at this time point, image quality is still suboptimal, due
lowing intravenous administration of 99mTc-MAG3 are especially to the late concentration of 99mTc-DMSA in the
reported here below, normalized to unit of administered liver, with an inverse proportion to renal uptake and function.
activity [3]: About 50% of the injected activity is still bound to the renal
tubules 24 h after administration.
• Effective dose 0.0017 mSv/MBq (urinary bladder emp- The main radiation dosimetry estimates to patients fol-
tied 30 min after administration) lowing intravenous administration of 99mTc-DMSA are
• Tissues/organs with the highest values of absorbed dose: reported here below, normalized to unit of administered
–– Urinary bladder wall 0.11 mGy/MBq activity [3]:
–– Uterus 0.012 mGy/MBq
–– Lower large bowel wall 0.0057 mGy/MBq • Effective dose 0.0088 mSv/MBq
–– Kidney 0.18 mGy/MBq
Key Learning Points
–– Spleen 0.013 mGy/MBq
• Tc-MAG3 (or betiatide or mertiatide) is a chemi-
99m
cal compound undergoing partial glomerular filtra-

tion and active tubular secretion.
• This radiopharmaceutical is characterized by high Key Learning Points
renal extraction, and it is commonly employed for • Tc-DMSA is a chelated radiopharmaceutical
99m
dynamic renal scintigraphy and tubular extraction with high renal extraction and intraparenchymal
rate (TER) calculation. retention commonly employed for static renal
scintigraphy.
• The precise intrarenal handling of 99mTc-DMSA is
2.3.5 99mTc-Dimercaptosuccinic Acid still debated.
(99mTc-DMSA)
99m
Tc-DMSA is a chelated radiopharmaceutical used for
renal parenchymal imaging (static renal scintigraphy), which 2.3.6 99mTc-Radiocolloids
provides information on morphology (malformations, ecto-
pies, agenesis), cortical function, and renal injuries or A “colloid” is defined as a homogeneous mixture in which one
infections. substance of microscopically dispersed insoluble particles is sus-
The precise intrarenal handling of 99mTc-DMSA is still pended throughout another substance (liquid or gas). To qualify
debated. Uptake can occur by tubular reabsorption across as a colloid, the mixture must be one that does not settle or would
34 F. Orsini et al.
take a very long time to settle appreciably. Unlike those in a sus- HSA microcolloid
pension, colloidal particles cannot be separated out by ordinary Rhenium sulfide
filtering or by centrifugation. The dispersed substance alone is Stannous fluoride
called the colloid. Therefore, from the biological point of view, a
Sulfur colloid (unfiltered)
colloid includes all particles ranging from 1 nm to about 4 μm
that are found in the body fluids. These particles are generally Sulfur colloid (prefiltered)
due to debris of microorganisms, fragments of cells produced
HSA nanocolloid
during the physiologic processes of tissue renewal, or products
Antimony sulfide
of intestinal absorption (especially fat). Colloids are generally
removed from circulation by phagocytosis, except for products Dextran
of intestinal absorption that undergo their physiological metabo-
1 10 100 1000 10000
lism in the liver or at other metabolic sites.
Originally introduced to perform hepatosplenic scintigra- Ranges of particle sizes, nm
phy and bone marrow scintigraphy (after systemic i.v. Fig. 2.9 Schematic representation of the ranges of particle sizes (in nm)
administration), their most common current use is for lym- constituting the main radiocolloids employed for lymphoscintigraphy;
phoscintigraphy. Following its intradermal or subcutaneous the logarithmic scale is used to compensate for the wide range in sizes
administration, a colloidal tracer is usually cleared from the (reproduced with permission from Erba PA, Bisogni G, Del Guerra
A, Mariani G. Methodological aspects of lymphoscintigraphy: radio-
injection site by lymphatic drainage, allowing visualization pharmaceuticals and administration. In: Mariani G, Manca G, Orsini
of lymph nodes protecting that tissue. In patients with breast F, Vidal-Sicart S. Valdés Olmos RA, eds. Atlas of Lymphoscintigraphy
cancer, melanoma, or penile tumors, it has become a stan- and Sentinel Node Mapping – A Pictorial Case-Based Approach.
dard of clinical practice to biopsy the first lymph node (called Milan: Springer; 2013: pp 17–26)
the sentinel node) draining the region of the tumor.
Radiopharmaceuticals with colloidal properties include
several types of preparations. The most widely used radiocol- colloidal particles with a diameter ≤80 nm, which easily migrate
loids are 99mTc-nanocolloidal human albumin, 99mTc-sulfur through the lymphatic system but are not totally retained in the
colloid, and 99mTc-antimony trisulfide. Additional radiocol- sentinel node. Lymphoscintigraphy will therefore usually dis-
loids employed for lymphoscintigraphy (depending on local play multiple lymph nodes along a certain lymphatic route, an
availability) include 99mTc-rhenium sulfide nanocolloid, 99mTc- occurrence that can complicate the intraoperative search for the
sulfide nanocolloid, 99mTc-stannous phytate, 99mTc-tin colloid, “true” sentinel lymph node using the gamma probe.
and 99mTc-microaggregated human albumin [6]. Despite 99m
Tc-Sulfur colloid is commonly used in the USA; the
important differences in the size range of the radiolabeled par- range of particle sizes is quite wide (15–5000 nm) depending
ticles (that dictate variable rates of clearance from the site of on the preparation method, with an average size ranging
interstitial injection) (Fig. 2.9), a choice of a certain radiocol- from 305 to 340 nm. The filtered colloidal form of 99mTc-
loid preparation over others is based more on local availability sulfur colloid (particle size ranging from 100 to 220 nm) is
than on differences in diagnostic performance (e.g., the suc- used for lymphoscintigraphy. In this case radiocolloids are
cess rate in sentinel lymph node identification for radioguided retained in the sentinel lymph node more efficiently than
sentinel lymph node biopsy, SLNB). It should be noted that 99m
Tc-nanocolloidal human albumin, although their migra-
smaller-sized particles migrate faster from the site of intersti- tion is quite slow, so that a longer time can be required to
tial injection through the lymphatic channels, a feature that is complete lymphoscintigraphy.
more favorable for investigating the pattern of lymphatic cir- 99m
Tc-Antimony trisulfide, used mostly in Canada and
culation, e.g., in patients with edema of the limbs. However, Australia, has a range of particle size of 3–30 nm. It migrates
small particles are not efficiently retained in the first lymph quite fast from the injection site through the lymphatic sys-
node they encounter along a certain pathway of lymphatic tem; however, it is less efficiently retained in the sentinel
drainage (i.e., the sentinel lymph node), and they progress in node, so that several lymph nodes can typically be visualized
part to visualize other lymph nodes along the same pathway; along a single lymphatic draining channel.
this feature complicates the gamma-probe-guided intraopera- The use of radiocolloids for liver, spleen, and bone mar-
tive search of sentinel lymph node(s). On the other hand, row imaging relies on the fact that, following i.v. administra-
larger particles are retained more efficiently in the sentinel tion, these radiopharmaceuticals are cleared from the
lymph node, but at the expense of a slower migration speed circulation through phagocytosis taking place in macro-
from the site of interstitial injection. phages of the endothelial-lymphatic system, which are par-
99m
Tc-Nanocolloidal human albumin is most frequently used ticularly abundant in the liver, spleen, and reticuloendothelial
in Europe and represents a good compromise between the speed system of the bone marrow.
of migration and lymph node retention. The size of its particles Lymphoscintigraphy is the first phase of radioguided
ranges from 5 to 100 nm, with at least 95% of human albumin SLNB. Along the route of lymphatic drainage from the site
of injection around the tumor, radiocolloids reach lymph • Effective dose 0.0091 mSv/MBq
nodes. The first lymph node is called the sentinel lymph • Tissues/organs with the highest values of absorbed dose:
node; this is the lymphatic station where tumor cells possibly –– Spleen 0.074 mGy/MBq
entering lymph vessels and migrating along the route of lym- –– Liver 0.071 mGy/MBq
phatic drainage encounter the first lymph node (or nodes) of –– Gallbladder wall 0.02 mGy/MBq
the lymphatic basin draining that specific tumor region.
Different parameters such as specific injection site, parti- The main radiation dosimetry estimates to patients fol-
cle size, and pathophysiology of local lymphatic circulation lowing intra-tumoral administration in breast cancer patients
affect the speed of lymphatic drainage of radiocolloids. of 99mTc-labeled small colloids (<100 nm, including nano-
Small particles are drained to the extent of about 40–60%, colloidal albumin and antimony sulfide) 18 h prior to surgery
whereas the larger particles are retained preferentially at the are reported here below, normalized to unit of administered
injection site. For instance, when using 99mTc-nanocolloidal activity [3]:
albumin, lymph node activity reaches a plateau within 2 h
after injection, approximately 3% of the injected activity • Effective dose 0.002 mSv/MBq
being retained in all visualized nodes. • Tissues/organs with the highest values of absorbed dose:
The ideal radiopharmaceutical for sentinel lymph node –– Myocardium 0.0071 mGy/MBq
biopsy should allow rapid visualization of the first lymph –– Remaining breast tissue 0.0064 mGy/MBq
node draining from the tumor/injection site and prolonged –– Lung 0.0063 mGy/MBq
retention in the sentinel node, possibly without further
migration to higher echelon nodes. In principle, radiocol- The main radiation dosimetry estimates to patients fol-
loids with relatively large-sized particles (between 200 and lowing oral administration of solid food labeled with 99mTc-
300 nm) have efficient retention in the first lymph node colloids are reported here below, normalized to unit of
encountered along the migration route, with very little pro- administered activity [3]:
gression to higher echelon nodes. Nevertheless, particles of
this size migrate quite slowly from the injection site. • Effective dose 0.024 mSv/MBq
Since the speed of migration of radiocolloids injected • Tissues/organs with the highest values of absorbed dose:
interstitially is inversely related to particle size, some com- –– Colonic wall 0.10 mGy/MBq
promise must be found between the speed of lymphatic –– Small bowel wall 0.061 mGy/MBq
migration and the degree of phagocytosis of the colloidal –– Gastric wall 0.059 mGy/MBq
particles in lymph nodes. Given the different sizes of radio-
colloids successfully used for lymph node mapping, the
choice is usually based on local availability.
Radiocolloids can also be used as nonabsorbable 99mTc- Key Learning Points
labeled radiopharmaceutical for gastroenteric scintigraphy, • Radiocolloids are very small particles (1 nm to
including esophageal transit and gastric emptying imaging 4 μm) that are found in the body fluids, generally
after oral administration. In this case the radiocolloids are due to debris of microorganisms, fragments of cells,
generally suspended in some liquid, semisolid, or solid food or products of intestinal absorption.
(such as water, jelly, or eggs, respectively). Since radiocol- • Several types of 99mTc-labeled colloids exist, with
loids cannot be absorbed within the gastrointestinal tract, it different origins and particle sizes.
is possible to obtain functional images of the oral, esopha- • Radiocolloids administered intravenously have
geal, and gastric lumen enabling investigation of swallowing been used to perform hepatosplenic scintigraphy
disorders, dysmotilities, or gastroparesis, according to the and bone marrow scintigraphy.
principle of compartmental localization. • As nonabsorbable radiopharmaceutical after oral
An additional use of radiocolloids such as 99mTc-nano- administration, they are employed for gastroenteric
colloidal human albumin has been described, i.e., as inert, scintigraphy (esophageal transit and gastric empty-
small-sized particles for lung ventilation studies, by means ing imaging) and, as inert small-sized particles, for
of a special nebulizer producing a radioactive droplet aero- lung ventilation scintigraphy.
sol [7]. • The most common current use of radiocolloids is
The main radiation dosimetry estimates to patients fol- lymphoscintigraphy after intradermal or subcutane-
lowing intravenous administration of 99mTc-labeled colloids ous administration in the search of the so-called
100–1000 nm in size (including sulfur colloid, tin colloid, sentinel lymph node in patients with different types
microaggregated albumin, and phytate) are reported here of solid cancers.
below, normalized to unit of administered activity [3]:
36 F. Orsini et al.
2.3.7 99mTc-Mannosyl-DTPA-Dextran 2.3.8 99mTc-Macroaggregated Albumin

(99mTc-Tilmanocept) (99mTc-MAA)
99m
Tc-Tilmanocept is a noncolloidal receptor-targeted radio- 99m
Tc-MAA consists of irregular-shaped heat-denatured
pharmaceutical approved by the FDA and EMA for sentinel human albumin particles with varying sizes (range between
lymph node mapping in patients with different types of solid 15 and 100 μm, no particles greater than 150 μm). This
cancers (including breast cancer, cutaneous melanoma, and radiopharmaceutical is employed for perfusion lung scintig-
squamous cell carcinoma of the oral cavity). raphy after i.v. injection of approximately 100,000–400,000
This small-sized macromolecule (average diameter 7 nm) radiolabeled particles (minimum 60,000).
consists of a dextran backbone with multiple units of DTPA After intravenous administration, over 90% of the 99mTc-
(for labeling with 99mTc) and mannose residues, each cova- MAA particles are trapped by microembolization in the pul-
lently attached to the dextran backbone. The uptake mecha- monary capillaries and precapillary arterioles because of their
nism of this radiopharmaceutical in lymph nodes does not large size, with a pattern of anatomic distribution that mirrors
depend on the particle size but on avid binding to the CD206 regional lung perfusion. The particles are then removed from
receptors for mannose expressed on the surface of macro- the pulmonary bed with an effective half-life of 1.5–3 h,
phages and dendritic cells in lymph nodes. through enzymatic degradation and mechanical movement of
The advantages of this novel radiopharmaceutical include the lung; smaller particles produced during degradation enter
rapid clearance from the injection site, high sentinel lymph node the circulation taking on nanocolloidal characteristics, under-
extraction, and high retention in sentinel lymph node, with con- going phagocytosis by macrophages of the reticuloendothe-
sequent low migration to second-echelon lymph nodes. lial system.
The main radiation dosimetry estimates to patients fol- Upon i.v. administration, the 99mTc-MAA particles
lowing peritumoral injection of 99mTc-tilmanocept in patients obstruct only a very small fraction of pulmonary vessels
with breast cancer are reported here below, normalized to (<0.1%), which is negligible from the hemodynamic point of
unit of administered activity: view since there are over 280 billion pulmonary capillaries
and 300 million precapillary lung arterioles. Particles in size
• Effective dose 0.3 mSv/MBq >150 μm would occlude arterioles of greater caliber with
• Tissues/organs with the highest values of absorbed dose: possible undesirable effects.
–– Breast (injection site) 1.65 mGy/MBq In patients with known pulmonary hypertension or right-
–– Ovaries 0.187 mGy/MBq to-left heart shunt, the number of 99mTc-MAA particles
–– Kidney 0.186 mGy/MBq should be reduced to 100,000–200,000. In infants and chil-
dren, the number of administered 99mTc-MAA particles
The main radiation dosimetry estimates to patients with should be further reduced as a function of age and body
melanoma following intradermal injection of 99mTc-tilmano- weight.
cept in patients with cutaneous melanoma are reported here 99m
Tc-MAA is also used for a particular procedure of
below, normalized to unit of administered activity: radioguided surgery, radioguided occult lesion localization
(ROLL). For this specific application, a small amount/vol-
• Effective dose 0.22 mSv/MBq ume of the radiopharmaceutical is injected under imaging
• Tissues/organs with the highest values of absorbed dose: guidance (often ultrasound, rarely CT) directly into the
–– Ovaries 0.299 mGy/MBq lesion to be surgically removed. Because of their size, the
–– Kidneys 0.278 mGy/MBq 99m
Tc-MAA particles remain into the lesion and do not
–– Testes 0.104 mGy/MBq undergo clearance via lymphatic drainage. The lesion is then
located and excised intraoperatively under guidance with a
gamma-detecting probe, according to the same detection
principles as those of radioguided SLNB.
Key Learning Points
99m
Tc-MAA can be injected also during an angiographic
• Tc-Tilmanocept is a small-sized macromolecule
99m session directly into the hepatic artery (or a segmental
with high avidity for binding with the CD206 recep- branch). This procedure is very important for treatment plan-
tors for mannose expressed on the surface of macro- ning prior to trans-arterial radioembolization therapy with
phages and dendritic cells in lymph nodes. 90
Y- or 166Ho-labeled microspheres in patients with primary
• This radiopharmaceutical is used for sentinel lymph or metastatic liver malignancies. In fact, the procedure allows
node mapping in patients with different types of assessment of possible shunt(s) of the intra-arterially injected
solid cancers. particles to the pulmonary and/or gastrointestinal
circulation.
The main radiation dosimetry estimates to patients follow- for characterization of patients with the dementias, 99mTc-
ing intravenous administration of 99mTc-MAA are reported HMPAO still plays an important role for rCBF assessment in
here below, normalized to unit of administered activity [3]: nuclear medicine centers where a PET scanner is not
available.
• Effective dose 0.011 mSv/MBq 99m
Tc-HMPAO has been used also as a brain-specific
• Tissues/organs with the highest values of absorbed dose tracer to perform brain death scintigraphy acquiring both
–– Lung 0.066 mGy/MBq early dynamic flow images and delayed SPECT images to
–– Liver 0.016 mGy/MBq ensure definitive absence of cerebral blood flow.
–– Myocardium 0.0096 mGy/MBq In clinical practice 99mTc-HMPAO is currently used pre-
dominantly for in vitro labeling of autologous leukocytes for
scintigraphic visualization of infectious foci (see Chap. 46 of
this book “Current practical guidelines for the most common
Key Learning Points nuclear medicine procedures” for details on how to radiola-
• Tc-MAAs are macroaggregates of denatured
99m
bel autologous leukocytes with 99mTc-HMAO).
human albumin, with particles ranging in size As a neutral, low-molecular-weight, and lipophilic com-
between 15 and 150 μm. pound, 99mTc-HMPAO easily crosses the blood-brain barrier
• After i.v. injection, they are trapped by microembo- by simple diffusion following its concentration gradient
lization in the pulmonary capillaries and precapil- from blood to the extravascular spaces and further on from
lary arterioles because of their large size and are the extracellular to the intracellular compartment. Because
therefore commonly employed for perfusion lung of this property and also by virtue of the efficient renal and
scintigraphy. hepatobiliary clearance (which remove 40% and 20%,
respectively, of the administered activity), the radiopharma-
ceutical is rapidly cleared from circulating blood after intra-
venous administration.
2.3.9 99mTc-Hexa-Methyl-Propylene- After diffusion across the cell membrane, 99mTc-HMPAO
AmineOxime (99mTc-HMPAO) undergoes the action of the enzyme esterase, which trans-
forms it from a lipophilic to a hydrophilic substance that can-
The molecular structure of 99mTc-HMPAO (also known as not diffuse back from the intracellular to the extracellular
99m
Tc-exametazine) shows Tc+5 to have five coordinate space; the hydrophilic form of 99mTc-HMPAO remains there-
groups, with an oxo-group at the apex and four nitrogen fore trapped within neurons. This process is at the base of the
atoms at the corners of the base of a square pyramid prolonged retention of 99mTc-HMPAO in the brain with a dis-
(Fig. 2.10). tribution pattern that is proportional to rCBF at the time of
This radiopharmaceutical has been developed primarily tracer administration. Uptake in the brain reaches a maxi-
to perform brain perfusion scintigraphy. SPECT acquisition mum of 4–6% of injected activity within 1 min, with very
is mandatory, as it allows the evaluation of changes in little decline of activity over the following 24 h. For this rea-
regional cerebral blood flow (rCBF) that characterize certain son, SPECT acquisition can start over a rather flexible time
clinical conditions such as Alzheimer’s disease and other frame (usually 30–60 min after injection but even at delayed
forms of dementia and cerebrovascular diseases in general. time points if necessary), during which no significant
Although PET radiopharmaceuticals are currently preferred changes in 99mTc-HMPAO concentration occur.
It should be noted that, after reconstitution of the labeling
kit, 99mTc-HMPAO is chemically stable for only 30 min and
it should therefore be administered (or used for leukocyte
H3C CH3 labeling) within this time frame. Therefore, the radiophar-
maceutical should be reconstituted only after the venous
O access has been set in place, in order to avoid delays in
H3C N N CH3
Tc administration that may hamper reliability of the scan.
N N The main radiation dosimetry estimates to patients fol-
H3C
HO O CH3 lowing intravenous administration of 99mTc-HMPAO are
H reported here below, normalized to unit of administered
activity [3]:
Fig. 2.10 Chemical and tridimensional structure of 99mTc-
HMPAO. Color codes: yellow = 99mTc; red = O; white = H; light • Effective dose 0.0093 mSv/MBq
blue = C; blue = N • Tissues/organs with the highest values of absorbed dose:
38 F. Orsini et al.
–– Kidney 0.034 mGy/MBq mechanism of passive diffusion following a concentration

–– Thyroid 0.026 mGy/MBq gradient. After its entry into the cells, 99mTc-ECD is trans-
–– Urinary bladder wall 0.023 mGy/MBq formed into the negatively charged compound 99mTc-ethyl-
enediyl-bis-l-cysteine monoethyl ester complex (or
99m
Tc-ECM) through a pH-dependent esterification enzy-
Key Learning Points
matic process that inactivates the lipophilic component. This
• Tc-HMPAO (or exametazime) is a neutral, low-
99m transformation prevents 99mTc-ECM from crossing the cell
molecular-weight, and lipophilic compound able to membrane back to the extracellular space.
cross the blood-brain barrier and to accumulate in Brain uptake is very rapid, about 5% of injected activity
the intracellular space. concentrating in the brain within the first 2–3 min after
• It is commonly employed to perform brain perfu- administration and only 5% of injected activity remaining in
sion SPECT and for in vitro labeling of autologous the circulation 1 h after administration. Since the removal
leukocytes. from the central nervous system is very slow, brain concen-
tration does not vary for at least 6 h after injection. For best
image quality, acquisitions should be performed 30–60 min
after injection in order to achieve the optimal target-to-back-
2.3.10 99mTc-Ethylenediylbis-Cysteine- ground ratio.
Diethylester (99mTc-ECD) About 50% of injected activity is excreted through the kid-
neys within the first 2 h, while at 24 h urinary excretion is about
99m
Tc-ECD (also known as 99mTc-bicisate) is a neutral lipo- 75%. The wall of the urinary bladder constitutes therefore the
philic complex that crosses the intact blood-brain barrier rap- critical organ; such relatively high radiation burden can signifi-
idly and enters the intracellular space of the brain, where it is cantly be reduced by frequently emptying the bladder.
retained for a long time. Only the stereoisomer l,l-ECD At variance with 99mTc-HMPAO, the 99mTc-ECD complex
undergoes an intracellular enzymatic process that converts it is very stable after reconstitution of the labeling kit; the radio-
into a polar species that is trapped in the human brain. Thus, pharmaceutical can therefore be injected up to about 8 h after
only purified l,l-ECD is used for 99mTc-ECD formulation reconstitution, if stored at a temperature below 25 °C.
(Fig. 2.11). The 99mTc-ECD molecule has the core structure When comparing 99mTc-ECD to 99mTc-HMPAO for brain
of Tc=ON2S2, with a coordination number of 5, consisting of perfusion imaging, some differences should be noted between
two cysteine residues linked by an ethylene bridge. The the two radiopharmaceuticals regarding in vitro stability, reten-
resulting neutral complex has a square pyramidal molecular tion mechanism, and dosimetry. Differences in the retention
geometry, with the Tc=O group at the apex of the pyramid. mechanisms may account for somewhat different patterns of
The only clinical indication for the use of this radiophar- distribution in specific disorders. For instance, in patients with
maceutical is brain perfusion scintigraphy for assessing subacute stroke 99mTc-ECD distribution seems to reflect meta-
regional cerebral perfusion abnormalities and brain death. bolic activity more closely, whereas 99mTc-HMPAO is better
Similarly to 99mTc-HMPAO, 99mTc-ECD crosses the correlated with cerebral perfusion. In conclusion, either of the
blood-brain barrier and neuronal cell membrane with a two tracers can be used, but they are not fully interchangeable.
lowing intravenous administration of 99mTc-ECD are reported
O O
O
here below, normalized to unit of administered activity [3]:
N N
Tc
H3C O O CH3 • Effective dose 0.0077 mSv/MBq
S S • Tissues/organs with the highest values of absorbed dose:
–– Gallbladder wall 0.028 mGy/MBq
–– Colonic wall 0.021 mGy/MBq
Key Learning Points

• Tc-ECD (or bicisate) is a neutral lipophilic com-
99m
plex able to cross the blood-brain barrier and enter

the intracellular space of the brain.
Fig. 2.11 Chemical and tridimensional structure of 99mTc-ECD. Color • This radiopharmaceutical is employed for brain
codes: yellow = 99mTc; red = O; white = H; light blue = C; blue = N; perfusion SPECT.
olive green = S
2.3.11 99mTc-Hexakis-2-Methoxy-2-Isobutyl- Localization of 99mTc-sestamibi in the parathyroid tissue

Isonitrile (99mTc-Sestamibi) is a function of metabolic activity, since this agent accumu-
lates preferentially in mitochondria-rich tissues, as typically
99m
Tc-Sestamibi is a positively charged lipophilic radiophar- is a hyperfunctioning parathyroid tissue [9]. Although
maceutical, consisting of a central atom of technetium linked 99m
Tc-sestamibi accumulates also in the normal thyroid
to six octahedral ligands (Fig. 2.12). parenchyma, its washout rate from the hyperfunctioning
After i.v. administration, 99mTc-sestamibi diffuses rapidly parathyroid cells is much lower than from the thyroid gland.
in the extracellular spaces and then crosses cell membranes This differential retention kinetics seems to be related to
by passive diffusion (because of its lipophilicity) following downregulation in the parathyroid tissue of the
both a concentration gradient and an electrical gradient driven P-glycoprotein system, which is responsible for the efflux of
by the negatively charged intracellular space. After entering various compounds (including 99mTc-sestamibi) from the
into cells, 99mTc-sestamibi rapidly accumulates in the mito- intracellular to the extracellular space. For parathyroid scin-
chondria, which are particularly abundant in cells with tigraphy using 99mTc-sestamibi as a single tracer, images are
enhanced energy demand, such as cardiomyocytes and para- recorded about 15 min postinjection, and at 2–3 h postinjec-
thyroid adenoma cells. Its uptake and retention depend on tion; differential washout rate of the radiopharmaceutical
regional blood flow, cell viability, cell membrane potential, allows detection of hyperfunctioning parathyroid tissue
and mitochondrial density/activity. Since this radiopharma- (Fig. 2.13).
ceutical is retained in mitochondria with a very slow washout Although the precise mechanisms responsible for 99mTc-
rate, no significant redistribution takes place; there is no evi- sestamibi uptake in breast tumors (as in other cancer cells) are
dence of any metabolism of this radiopharmaceutical. still not completely clarified, they most likely depend on
After intravenous injection, 99mTc-sestamibi is rapidly regional blood flow, plasma and mitochondrial membrane
cleared from the blood, with only 8% of the administered activ- potential, angiogenesis, and overall tissue metabolism. For
ity remaining in the circulation 5 min postinjection. This tracer scintimammography, 99mTc-sestamibi imaging was first per-
concentrates predominantly in the muscles (including the myo- formed using standard gamma cameras and imaging tables
cardium, with a biological half-life of approximately 6 h), liver, adapted to allow for prone imaging [10], whereas high-reso-
and kidneys, with a smaller amount accumulating in the sali- lution small-field-of-view gamma cameras specifically
vary glands and thyroid. When 99mTc-sestamibi is injected dur- designed for breast imaging have more recently been
ing an exercise stress test, its uptake in the heart and skeletal introduced.
muscles increases considerably due to the redistribution of car- The main radiation dosimetry estimates to patients fol-
diac output during exercise (decreased perfusion of the splanch- lowing intravenous administration of 99mTc-sestamibi are
nic bed and increased perfusion of the exercising muscles). In reported here below, normalized to unit of administered
the resting state, the main pathway for excretion is via the hepa- activity [3]:
tobiliary system to the gastrointestinal tract (33% at 48 h), with
some additional excretion via the kidneys (27% at 24 h). Most • Effective dose 0.009 mSv/MBq
of the injected activity is excreted within 48 h. • Tissues/organs with the highest values of absorbed dose:
Originally introduced into the clinical routine for myocar- –– Gallbladder wall 0.039 mGy/MBq
dial perfusion scintigraphy [8] 99mTc-sestamibi is used also –– Kidney 0.036 mGy/MBq
for parathyroid scintigraphy and for scintimammography. –– Colonic wall 0.024 mGy/MBq
Fig. 2.12 Chemical and

tridimensional structure of
99m
Tc-sestamibi. Color codes:
yellow = 99mTc; red = O; CH3 H3C O
CH3
white = H; light blue = C; O
blue = N CH3
H3C CH3 CH3
N
O
N N
CH3 CH3
H3C CH3 TC
N N
O
N
CH3 H3C CH3
H3C
O
H3C
O CH3 H C
3
40 F. Orsini et al.
a b
c d
Fig. 2.13 Examples of dual-phase parathyroid scintigraphy with on early imaging (where also the thyroid parenchyma exhibits some
99m
Tc-sestamibi (early image on the left, delayed image on the right). tracer uptake) (c) on delayed imaging (where most of the activity has
(a) No abnormal areas of focally increased 99mTc-sestamibi uptake on cleared from the thyroid parenchyma) (d); the area with slow 99mTc-
both early and delayed imaging. (b) Focal area of increased 99mTc-ses- sestamibi washout corresponds to an adenoma of the right lower para-
tamibi uptake at the lower pole of the right thyroid lobe visualized both thyroid gland
The localization properties of 99mTc-tetrofosmin are simi-

Key Learning Points lar as those of 99mTc-sestamibi; after entering into cells by
• Tc-Sestamibi is a positively charged lipophilic
99m
passive diffusion following a concentration gradient and an
compound, able to enter into cells by passive diffu- electrical gradient, 99mTc-tetrofosmin is retained intracellu-
sion and be retained mostly in the mitochondria. larly mostly bound to cytosol proteins.
• In addition to its routine use for myocardial perfu- Myocardial uptake of 99mTc-tetrofosmin (as well as that of
sion scintigraphy, it is also employed for parathy- 99m
Tc-sestamibi) is proportional to blood flow up to about
roid scintigraphy and for scintimammography. 2 mL/min per gram, which is twice the resting level, reach-
ing a plateau at higher flow rates.
Following systemic intravenous administration, less than
5% of the administered activity remains in the circulation
2.3.12 99mTc-6,9-Bis(2-Ethoxyethyl)-3,12- 10 min postinjection. About 65% of the administered activity
Dioxa-6,9-Diphospha-Tetradecane is excreted within 48 h, 40% in the urine and 25% through
(99mTc-Tetrofosmin) the hepatobiliary tract. The hepatic clearance of 99mTc-tetro-
fosmin is more rapid than that of 99mTc-sestamibi, thus allow-
99m
Tc-Tetrofosmin is a positively charged lipophilic techne- ing earlier myocardial imaging.
tium phosphine whose molecular structure is characterized 99m
Tc-Tetrofosmin is employed almost exclusively for
by a core of four atoms of phosphor-binding 99mTc, stabilized myocardial perfusion scintigraphy [11], whereas, its use for
by two oxygen atoms (O=Tc=O) (Fig. 2.14). parathyroid scintigraphy is much less frequent than that of
Fig. 2.14 Chemical and

tridimensional structure of
99m
Tc-tetrofosmin. Color CH3 CH3
codes: yellow = 99mTc;
red = O; white = H; light O O
blue = C; purple = P
H3C O O CH3
P O P
Tc
P O P
H3C O O CH3
O O
CH3 CH3
99m
Tc-sestamibi, mostly due to the fact that its kinetics of clinical practice, mainly 99mTc-mebrofenin (i.e., N-(3-bromo-
washout from the thyroid parenchyma is not so much differ- 2,4,6-trimethylphenylcarbamoylmethyl)-iminodiacetic acid)
ent from the washout rate from hyperfunctioning parathyroid and 99mTc-diisopropyl-IDA are used.
tissue. These derivatives are used for hepatobiliary function
The main radiation dosimetry estimates to patients fol- imaging, as they allow scintigraphic evaluation of overall
lowing intravenous administration of 99mTc-tetrofosmin are hepatocyte function, conjugation capability, and kinetics of
reported here below, normalized to unit of administered biliary excretion. Following i.v. injection, hepatocytes rap-
activity [3]: idly extract these radiopharmaceuticals from circulating
blood against an electrochemical and concentration gradient,
• Effective dose 0.008 mSv/MBq thanks to non-sodium-dependent carriers located on the
• Tissues/organs with the highest values of absorbed dose: bloodstream side of the cell membrane; this mechanism is
–– Gallbladder wall 0.036 mGy/MBq similar to the mechanism that bilirubin, bile acids, and other
–– Colonic wall 0.024 mGy/MBq anions metabolized in the liver undergo. Binding of 99mTc-
–– Urinary bladder wall 0.017 mGy/MBq IDA derivatives with the bilirubin receptors is about 15-fold
greater than that for bilirubin. Therefore, it is possible to per-
form the examination in the presence of high bilirubin blood
Key Learning Points values. However, based on our experience, when the biliru-
• Tc-Tetrofosmin is a positively charged lipophilic
99m bin is >10 mg/dL, hepatobiliary visualization is poor. In
compound able to enter into cells by passive diffu- hepatocytes these compounds are rendered water-soluble by
sion and be retained mostly bound to cytosol conjugation with glucuronic acid and are then excreted into
proteins. the bile through the hepatocyte bile ducts, ending up ulti-
• It is mainly employed for myocardial perfusion mately into the intestine.
scintigraphy. About 8–17% of the injected activity of 99mTc-mebrofenin
remains in the circulation 30 min after injection.
Approximately 1–9% of the administered activity is excreted
in the urine over the first 2 h after injection. In fasting indi-
2.3.13 99mTc-Labeled Iminodiacetic Acid (IDA) viduals, the maximum liver uptake occurs by 10 min postin-
Derivatives (99mTc-IDA) jection and peak gallbladder activity by 30–60 min after
injection. Only a minor fraction of the injected activity is
Radiopharmaceuticals belonging to this group are analogs of eliminated in the urine (<5%), because of their binding to
lidocaine consisting of lipophilic anions with very low affin- plasma proteins and low water solubility. In cases where the
ity for plasma proteins. Based on variable N-substitution of gallbladder is visualized but no activity has entered the
IDA, they are named according to the type of substitute. The bowel, additional pharmacologic intervention, with chole-
main compounds of this group are diethyl-IDA, diisopropyl- cystokinin, is required. In cases where the gallbladder is not
IDA, p-isopropyl-IDA, p-butyl-IDA, and mebrofenin. In visualized within 1 h of administration of 99mTc-mebrofenin,
42 F. Orsini et al.
morphine administration may be helpful to cause contraction anti-murine antibodies—HAMA) and that Fab fragments of
of the sphincter of Oddi, resulting in filling of the gallbladder antibodies have much faster clearance kinetics than the intact
(in the absence of cholecystitis). IgG molecules. Furthermore, the free thiol groups present in
The main radiation dosimetry estimates to patients following the Fab′ fragment following chemical reduction allow direct
intravenous administration of 99mTc-IDA derivatives are reported radiolabeling with 99mTc. Following i.v. injection, concentra-
here below, normalized to unit of administered activity [3]: tion of 99mTc-sulesomab in the blood at 1 h postinjection is
34% of the initial level, declining to 17% at 4 h and 7% at
• Effective dose 0.016 mSv/MBq 24 h. The route of excretion is predominantly renal, 41% of
• Tissues/organs with the highest values of absorbed dose: the radiolabel being excreted in the urine over the first 24 h
–– Gallbladder wall 0.11 mGy/MBq after administration.
–– Colonic wall 0.072 mGy/MBq Imaging with 99mTc-sulesomab should be acquired 1–8 h
–– Small intestine wall 0.043 mGy/MBq postinjection, without relevant differences in the detection
rate of osteomyelitis between the 1–2 h time point and the
5–8 h time point. Although planar imaging is performed first
Key Learning Points in all views necessary to adequately visualize the affected
• Tc-labeled iminodiacetic acid derivatives
99m area(s), SPECT/CT imaging is crucial for distinguishing
(99mTc-IDA) are analogs of lidocaine consisting of pure osteomyelitis from infection of soft tissues only.
lipophilic anions with very low affinity for plasma The main radiation dosimetry estimates to patients fol-
proteins. lowing intravenous administration of 99mTc-sulesomab are
• They are used for hepatobiliary functional imaging, reported here below, normalized to unit of administered
as they are extracted by hepatocytes and then activity [3]:
excreted into the bile.
–– Kidneys 0.0449 mGy/MBq
2.3.14 99mTc-Sulesomab –– Urinary bladder wall 0.0221 mGy/MBq
–– Spleen 0.0115 mGy/MBq
The murine monoclonal antibody IMMU-MN3 was origi-
nally raised with the aim of targeting the tumor-associated
carcinoembryonic antigen (CEA)—to be used as a radiola- Key Learning Point
beled agent for immunoscintigraphy and possibly radioim- • 99m
Tc-Sulesomab is a murine monoclonal antibody
munotherapy of CEA-expressing tumors. Although the used for the identification and localization of
results of CEA-targeting studies were disappointing, it infection.
turned out that the antibody efficiently bound an antigen
abundantly expressed on the cell membrane of granulocytes,
the so-called nonspecific cross-reacting antigen NCA90. The
use of 99mTc-sulesomab (a mixture of Fab′-SH (Fab2), heavy 2.3.15 99mTc-EDDA/HYNIC-[Tyr3]-Octreotide
and light chains, and other fragments of the IMMU-MN3 (TOC) and Other Somatostatin Analogs
antibody) for imaging sites of infection/inflammation is
based on this property. 99m
Tc-EDDA/HYNIC-octreotide is a somatostatin (SST)
The main indication for scintigraphy with 99mTc-sulesomab analog labeled with 99mTc. SST is a 14-amino acid peptide
is for the identification and localization of infection, particu- produced endogenously mainly in the central nervous sys-
larly osteomyelitis of appendicular bones (e.g., osteomyelitis tem. Physiologic functions of SST include inhibition of
complicating diabetic foot ulcers). In certain clinical condi- secretion of various hormones, such as insulin and gluca-
tions, it constitutes therefore a valid alternative to scintigraphy gon. Receptors for somatostatin are overexpressed in most
with radiolabeled autologous leukocytes, which is burdened of neuroendocrine tumors, and receptor activation inhibits
with some limitations, e.g., requirement for a quite large blood function cell growth, therefore also tumor growth. For this
sample (from patients who are frequently critically ill) and reason, SST analogs are used in oncology as antitumor
potential hazard of infection for the nuclear medicine person- drugs in patients with some neuroendocrine tumors. Five
nel handling leukocytes from high-risk patients. different types of receptors for SST have been identified,
The choice of fragments rather than whole intact immu- named SSTR1-5 (type 2 being divided into subtypes 2a
noglobulins is based on the notion that these preparations are and 2b). Native SST has a very short biological half-life
less immunogenic (thus reducing the risk of raising human (3 min), which limits its use both as an oncological thera-
Fig. 2.15 Tridimensional (1-Nal)

structure of the somatostatin
(Tyr)
analogs of major interest in
nuclear medicine and of the
chelating agents used for
labeling with different
radionuclides (HYNIC, HYNIC: 99mTc
DTPA, DOTA). Color codes:
red = O; white = H; light
blue = C; blue = N; olive
green = S
DTPA 111In
(Thr) Octreotide (OC)
[ Tyr3 ]-Octreotide (TOC)

[ Tyr 3,Thr0 ]-Octreotide (TATE)
[1-NaI3 ]-Octreotide (NOC)
DOTA 111In, 68Ga, 90Y, 177Lu
peutic drug as such and, upon radiolabeling, as a radio- Mainly because of the favorable physical properties and
pharmaceutical for either diagnostic or therapeutic wide availability of 99mTc, 99mTc-EDDA/HYNIC-octreotide
application. offers several advantages over 111In-pentetreotide, such as
Several SST analogs have therefore been developed, with ready availability (because of the possibility to label the
longer biological half-life and physiological biodistribution and lyophilized kit “on demand” when needed), possibility to
biological activity similar to native SST (Fig. 2.15), as follows: administer higher amounts of activity (because of the more
favorable dosimetry parameters), faster imaging (completed
• Octreotide (OT), with medium/high affinity for SSTR2 within 4–6 h versus up to 24 h), better image quality and
and medium affinity for SSTR3 and SSTR5. It is employed detection capability (with higher target/nontarget ratios
for imaging in conventional nuclear medicine, labeled because of more favorable counting statistics), and overall
with 111I through the DTPA chelator (111In-DTPA- better cost-effectiveness.
octreotide or 111In-pentetreotide). After intravenous administration, 99mTc-EDDA/
• [Tyr3]-Octreotide (TOC), with high affinity for SSTR2, HYNIC-TOC is rapidly cleared from the blood, although
medium/high affinity for SSTR5, and low affinity for SSTR3. with some residual binding to blood cells (<5% of
It is employed for imaging in conventional nuclear medicine, injected activity throughout several hour postinjection).
labeled with 99mTc through the HYNIC/EDDA chelator As early as 10 min postinjection, accumulation of 99mTc-
(99mTc-EDDA/HYNIC-octreotide). Through the DOTA1 che- EDDA/HYNIC-TOC is detected in the main sites of
lator, it can be labeled with 68Ga for PET imaging (68Ga-DOTA- physiologic uptake/excretion (i.e., the liver, spleen, and
TOC) or with 90Y for therapy (90Y-DOTA-TOC). kidneys), as well as in the tumor lesions expressing SST
• [Tyr3,Thr0]-Octreotide (TATE), with very high affinity for receptors.
SSTR2 (six- to ninefold greater than TOC), medium Maximal tumor/background ratios are observed at 4 h
affinity for SSTR5, but no affinity for SSTR3. Through after injection, cancer lesions remaining still detectable at
the DOTA chelator, it can be labeled with 177Lu for ther- 24 h. Minor excretion through the gastrointestinal tract is
apy (177Lu-DOTA-TATE). observed in delayed images. 99mTc-EDDA/HYNIC-TOC is
• [1,NaI3]-Octreotide (NOC), with high affinity for SSTR2 excreted mainly by the renal route, with only a small fraction
and SSTR5 and medium/high affinity for SSTR3. Through undergoing hepatobiliary clearance.
the DOTA chelator, it can be labeled with 68Ga for PET The main radiation dosimetry estimates to patients fol-
imaging (68Ga-DOTA-NOC). lowing intravenous administration of 99mTc-EDDA/HYNIC-
TOC are reported here below, normalized to unit of
DOTA = tetraazacyclododecane-1,4,7,10-tetraacetic acid
1
administered activity:
44 F. Orsini et al.
• Effective dose 0.005 mSv/MBq lic radionuclides with the DOTA chelator). This cyclotron-
• Tissues/organs with the highest values of absorbed dose: produced radionuclide has a quite long physical half-life
–– Spleen 0.037 mGy/MBq (2.8 days) and emits γ-rays with energy peaks of 171 keV
–– Kidney 0.02 mGy/MBq (90%) and 245 keV (94%), with additional X-ray emission of
–– Urinary bladder wall 0.012 mGy/MBq 23–26 keV. Due to these energy emission levels, scintigraphy
with 111In-labeled agents is generally performed with gamma
There are also other 99mTc-labeled SST analogs, such as cameras equipped with medium-energy collimators.
99m
Tc-HYNIC-octreotide, 99mTc-vapreotide, 99mTc-depreo- Scintigraphy with 111In-pentetreotide (that has medium/
tide, and 99mTc-demotate. 99mTc-HYNIC-Octreotate differs high affinity for SSTR2 and medium affinity for SSTR3 and
from 99mTc-HYNIC-octreotide for the presence of the more SSTR5) is primarily employed to investigate patients with
hydrophilic tyrosine in the third position and of the terminal neuroendocrine tumors expressing SST receptors (especially
threonine causing higher receptor affinity and better internal- gastroenteropancreatic and carcinoid tumors) for the purposes
ization properties. of staging, restaging, and evaluation of response to treatments.
99m
Tc-Depreotide had been approved for discriminating Increased expression of SST receptors occurs also in tumors
malignant from benign solitary pulmonary nodules. This originating from the neural crest, such as paragangliomas,
synthetic 10-amino acid peptide (which can be easily labeled neuroblastomas, pheochromocytomas, and medullary thyroid
with 99mTc due to the presence of a particular amino acid carcinomas. Furthermore, varying degrees of SST receptor
sequence) was originally developed as a potent SST analog expression are also observed in non-neuroendocrine tumors,
exhibiting good affinity for SSTR2, SSTR3, and SSTR5. such as breast cancer and some lymphomas (both Hodgkin’s
After initial studies aimed at targeting neuroendocrine and non-Hodgkin’s lymphomas). Finally, increased SST
tumors, strong binding of 99mTc-depreotide to SST receptors receptor expression occurs even in non-neoplastic disorders,
expressed in both small cell and non-small cell lung cancer especially in subacute and chronic inflammation (mainly of
was observed. However, this radiopharmaceutical is no lon- the granulomatous type, as in sarcoidosis).
ger available commercially, because of the growing use of Following i.v. injection, 111In-pentetreotide is rapidly cleared
[18F]FDG PET for characterizing solitary pulmonary nodules from the plasma; only one third of the injected activity remains
(and lung cancers in general). in the blood pool at 10 min after administration. Plasma levels
continue to decline steadily, so that by 20 h postinjection, about
1% of the activity is found in the blood pool.
111
In-Pentetreotide is mainly excreted by the kidneys,
Key Learning Points
50% of injected activity being recovered in the urine within
• Receptors for somatostatin are overexpressed in the
6 h and up to 85–90% at 24 h after administration. A smaller
majority of neuroendocrine tumors, and their acti-
fraction of radiopharmaceutical is excreted through the hep-
vation inhibits function cell growth.
atobiliary system, corresponding to about 2% of the injected
• These targets are used in oncology for cancer treat-
activity after 3 days. Scintigraphy shows physiologic uptake
ment, with the purpose of inhibiting tumor growth.
in the spleen (about 2.5% of injected activity at 24 h), liver
• Several exogenous somatostatin analogs have been
(2%), and kidneys. In most patients, the thyroid, pituitary,
developed, labeled with different radionuclides for
and intestine are also visualized during scintigraphy with
both therapeutic and imaging purposes; 99mTc- 111
In-pentetreotide. The standard injected activity into an
EDDA/HYNIC-octreotide is the most commonly
adult patient is 110–220 MBq.
used radiopharmaceutical labeled with 99mTc to
image neuroendocrine tumors.
Key Learning Points

2.4 Radiopharmaceuticals Labeled • In-Pentetreotide is a somatostatin analog that
111
with Indium-111 avidly binds to somatostatin receptors on cell mem-

branes, generally overexpressed in neuroendocrine
2.4.1 111In-DTPA-Octreotide tumors.
(111In-Pentetreotide) • It is the most commonly 111In-labeled radiopharma-
ceutical used in conventional nuclear medicine and
The 8-amino acid 111In-pentetreotide is an analog of the active is employed to investigate patients with neuroendo-
part of SST (Fig. 2.15), linked to a chelating group constituted crine tumors, especially gastroenteropancreatic and
by DTPA that makes radiolabeling with 111In possible carcinoid tumors.
(although such binding is not as stable as the binding of metal-
The main radiation dosimetry estimates to patients following 2.5 Radioiodinated Imaging Agents
intravenous administration of 111In-pentetreotide are reported
here below, normalized to unit of administered activity: The radioisotopes of iodine used for conventional gamma
camera imaging are 123I (for diagnosis) and 131I (mostly for
• Effective dose 0.054 mSv/MBq therapy), whereas 124I is used for PET imaging. In addition to
• Tissues/organs with the highest values of absorbed dose: its use in the simple chemical form of iodide anion for imag-
–– Spleen 0.57 mGy/MBq ing the thyroid parenchyma and follicular cell-derived thy-
–– Kidney 0.41 mGy/MBq roid cancers, radioiodine can replace some functional groups
–– Urinary bladder wall 0.2 mGy/MBq in larger molecules, such as the –CH3− group (whose radius
is similar to the atomic radius of iodine) or the hydroxylic
group –OH− through a process of nucleophilic/electrophilic
2.4.2 111In-DTPA substitution. Both the –CH3− group and the –OH− radical are
abundant in amino acids (typically in tyrosine) that are
The primary use of 111In-DTPA is for intrathecal radiophar- widely present and frequently exposed to the surface of mac-
maceutical administration for cisternoscintigraphy, deter- romolecules present in biological systems—being therefore
mining CSF shunt patency, and assessing the dynamics of available for a substitution reaction with (radio)iodine.
the cerebrospinal fluid (CSF). Tracers administered through 131
I is historically one of the first radionuclides made
this route undergo partial absorption in the subarachnoid available for medical applications and still remains a mile-
space, some of the CSF flowing to the basal cisterns within stone for therapeutic applications, due to its emission of
2–4 h and then progressing to the Sylvian cisterns, to the β-particles. Because of its concomitant γ-emission, 131I
interhemispheric cisterns, and over the cerebral convexities. (which has an 8-day physical half-life) is also still used for
In normal individuals, the radiopharmaceutical ascends to some specific diagnostic applications, typically in the chemi-
the parasagittal region within 24 h, clearing partially or com- cal form of sodium iodide for imaging patients with differen-
pletely from the basal cisterns and Sylvian regions. tiated thyroid cancer. Because of the main γ-emission with a
Due to reabsorption and passage of 111In-DTPA into the relatively high energy (364 keV), image acquisition with
systemic circulation, about 65% of the administered activity 131
I-labeled radiopharmaceuticals requires the use of gamma
is excreted in the urine within 24 h after intrathecal injection, cameras equipped with medium-high-energy collimators,
increasing to 85% at 72 h. The recommended activity of which results in suboptimal image quality. Moreover, the
111
In-DTPA to be injected intrathecally for radionuclide cis- β-particle emission associated with its decay results in a rela-
ternography is 18.5 MBq in a standard adult individual with tively high radiation burden to patients.
70-kg body weight. 123
I does not suffer from these drawbacks, since it has a
The main radiation dosimetry estimates to patients fol- much shorter physical half-life (13.2 h) than 131I and does not
lowing intrathecal administration of 111In-DTPA are reported emit β- particles, thus resulting in a more favorable radiation
here below, normalized to unit of administered activity burden to patients. Moreover, the energy of its γ-emission
according to the route of injection—lumbar or cisternal: (159 keV) is optimal for gamma camera imaging and allows
for the use of high-resolution collimators.
• Effective dose 0.14 mSv/MBq (lumbar) and 0.12 mSv/
MBq (cisternal)
• Tissues/organs with the highest values of absorbed dose: 2.5.1 123I/131I-Sodium Iodide
–– Spinal cord 0.95 mGy/MBq (lumbar) and 0.57 mGy/
MBq (cisternal) 123
I-Iodide and 131I-iodide share exactly the same pharmaco-
–– Red marrow 0.24 mGy/MBq (lumbar) and 0.14 mGy/ kinetic and uptake characteristics. As mentioned above, the
MBq (cisternal) use of 123I-iodide is to be preferred over 131I-iodide for diag-
–– Urinary bladder wall 0.20 mGy/MBq (lumbar) and nostic imaging (especially in children) because of its lower
0.18 mGy/MBq (cisternal) radiation burden to patient and better image quality; the
only practical drawback is the much higher cost of
123
I-iodide, associated also with a more complex distribution
Key Learning Points logistics.
• In-DTPA is a small chelated compound with neg-
111
Nevertheless, there are some specific diagnostic applica-
ative electrical charge. tions where 131I-iodide is still largely employed in the clinical
• Radiolabeling with 111In permits to perform cister- practice. In particular, this radiopharmaceutical is used for
noscintigraphy after intrathecal administration, measuring radioiodine uptake prior to radiometabolic ther-
with delayed imaging—even up to 72 h apy of hyperthyroid patients and for whole-body scintigra-
post-administration. phy (131I-WBS) in patients with differentiated thyroid cancer.
For these two applications, the administered activities are very
46 F. Orsini et al.
different: only a trace amount of radioactivity (0.37–1.85 MBq) • Effective dose 0.15 mSv/MBq

to measure thyroid uptake but up to 185 MBq for 131I-WBS • Tissues/organs with the highest values of absorbed dose:
in thyroid cancer patients. In both cases, the radiopharma- –– Thyroid 2.7 mGy/MBq
ceutical is generally administered orally as a capsule, while –– Urinary bladder wall 0.077 mGy/MBq
in pediatric patients or in patients with swallowing problems –– Gastric wall 0.064 mGy/MBq
a liquid formulation is also available, for either oral or intra-
venous administration. The main radiation dosimetry estimates to patients fol-
After administration, 123I- and 131I-iodide have a pattern of lowing oral administration of 123I-iodide are reported here
distribution and metabolism exactly as the native 127I-iodide below, normalized to unit of administered activity [3]:
present in the circulation and in interstitial fluid. In particular,
iodine is actively transported into the thyroid cells by the Na+/ • Effective dose 0.15 mSv/MBq
I− symporter (NIS) against a concentration gradient, intracel- • Tissues/organs with the highest values of absorbed dose:
lular concentration being 20- to 40-fold higher than extracel- –– Thyroid 2.5 mGy/MBq
lular concentration. Once inside the thyroid cells, iodine is –– Gastric wall 0.12 mGy/MBq
oxidized by the thyroid-specific enzyme peroxidase; this reac- –– Urinary bladder wall 0.066 mGy/MBq
tion starts the organification process that leads to incorpora-
tion of iodine into tyrosine, an amino acid contained in the
thyroglobulin molecules produced by the endoplasmic reticu- Key Learning Points
lum and Golgi apparatus. Molecules of monoiodotyrosine • The radioisotopes of iodine used for conventional
(MIT) and diiodotyrosine (DIT) are thus formed. These two gamma camera imaging are 123I and 131I.
chemical species are then combined to produce the thyroid • The use of the simple chemical form of iodide anion
hormones thyroxine (T4) and triiodothyronine (T3). Thyroid intends for imaging the thyroid parenchyma and
hormones stored in the follicle-filling colloid are finally follicular cell-derived thyroid cancers.
secreted into the circulation. The fundamental steps of this • Through a process of nucleophilic/electrophilic
complex metabolic pathway (i.e., NIS expression/activity and substitution, 123I and 131I can be used to label further
secretion of thyroid hormones by pinocytosis) are stimulated molecules with biological targets.
by TSH.
In addition to the thyroid, iodine is also concentrated by
the salivary glands, sweat glands, stomach, mammary glands, 2.5.2 123I-Ortho-Iodo-Hippuric Acid
and placenta. (123I-Hippuran)
123
I-Iodide is generally used for thyroid scintigraphy in
patients with benign thyroid disease, both for diagnostic 123
I-Hippuran (or 123I-OIH) is an anionic structural analog of
imaging and for measuring radioiodine uptake. 123I-Iodide is para-aminohippurate, a natural substance that is normally
in principle to be preferred over 131I-iodide for whole-body excreted in the urine by both tubular secretion (80%) and
dosimetric purposes in patients with thyroid malignancies glomerular filtration (20%). Despite a relatively high binding
and also to avoid the thyroid stunning effect prior to high- to plasma proteins (approximately 60% of circulating activ-
dose therapy with 131I-iodide. The concept of thyroid stun- ity), 123I-hippuran (Fig. 2.16) has a high extraction fraction
ning relies on the hypothesis which implies that the radiation (about 85% at each passage through the kidneys), because of
effects from the relatively low activity of the diagnostic
β-emitter 131I-iodide activities may impair the ability of the
O
remnant thyroid tissue or metastases to concentrate iodine,
thus jeopardizing the therapeutic outcomes. O
NH
123
I-Iodide is generally administered orally, although an
OH
i.v. formulation is also available. The standard activity 131I
administered for thyroid scintigraphy in a 70-kg adult indi- 131l-Ortho-lodo-Hippurate

vidual varies from 2 MBq (for measuring radioiodine uptake)
to 111 MBq (for thyroid scintigraphy). Higher 123I activities O
are required for whole-body scans (40–200 MBq) and even O
higher for SPECT acquisitions, images being acquired 24 h NH
after administration. OH
The main radiation dosimetry estimates to patients fol- H2N
Para-Amino-Hippurate
lowing intravenous administration of 123I-iodide are
reported here below, normalized to unit of administered Fig. 2.16 Chemical structure of para-aminohippurate (bottom) and of
activity [3]: its radioiodinated analog labeled with 131I
an active secretion process that results in progressively rising 2.5.3 Radioiodinated Meta-Iodo-Benzyl-
concentrations along the proximal renal tubules. Similarly as Guanidine (MIBG)
described for 99mTc-MAG3, active trans-tubular transport of
123
I-hippuran is mediated by transmembrane exchangers In the MIBG molecule (also known as iobenguane), the ben-
located in the basolateral membrane of tubular cells, only zyl group of bretylium is combined with the guanidine group
indirectly coupled with Na+/K+ ATPase. of guanethidine. Both bretylium and guanethidine are potent
After i.v. injection, 123I-hippuran is rapidly cleared from neuron-blocking agents that act selectively on adrenergic
the blood and enters different extracellular compartments. In nerves [13]. A combination of the benzyl portion of brety-
addition to the predominant renal excretion (about 70% of lium with the guanidine group of guanethidine results in the
injected activity within 30 min postinjection), there is also a production of a variety of substituted aralkyl-guanidines
minor fraction of hepatobiliary excretion. with even greater anti-adrenergic potency [14]. MIBG is one
The radiopharmaceutical is supplied as a sterile, pyrogen- of these derivatives that can be radioiodinated for imaging
free solution ready for administration, which is performed by both the adrenal medulla and distribution of adrenergic
rapid i.v. bolus injection. Although 123I-hippuran has optimal innervation in the myocardium [15].
characteristics for imaging associated with a low radiation As a structural analog of catecholamines, MIBG is taken
burden to patients (but is burdened by high cost and complex up by chromaffin cells via the same metabolic pathway of nor-
distribution logistics), the less expensive 131I-hippuran is also adrenaline, through a sodium- and ATPase-dependent active
available, as also is 125I-hippuran; the latter tracer is used for process. Once in the cytoplasm, MIBG is actively stored in
measurements of renal function based only on blood sam- cytoplasmic secretion granules based on the vesicular mono-
pling and in vitro counting with a well-type γ-counter— amine transporter (VMAT1 and VMAT2). Thus, it physiologi-
without gamma camera imaging. cally accumulates in the neuronal secretory vesicles located in
Quantitative analysis of the data acquired during dynamic sympathetic system ganglia, in the medulla of adrenal glands,
renal scintigraphy with radiolabeled hippuran allows calcula- and in all other human tissues with high adrenergic innerva-
tion of the “effective” renal plasma flow (ERPF), an overall tion—such as myocardiocytes. The secretory vesicles release
parameter of renal function that can be associated with measure- their contents by a process of exocytosis stimulated by mem-
ment of GFR in order to calculate the renal filtration fraction. brane depolarization linked to calcium flow.
The main radiation dosimetry estimates to patients follow- MIBG can be labeled with either 123I (for diagnostic use
ing intravenous administration of 123I-hippuran are reported only) or 131I (for both diagnostic and therapeutic use).
here below, normalized to unit of administered activity [12]: 123
I-MIBG is to be considered the diagnostic radiopharma-
ceutical of choice, especially for use in children, because of
• Effective dose 0.015 mSv/MBq its more favorable radiation dosimetry and better image
• Tissues/organs with the highest values of absorbed dose: quality being also suitable for hybrid SPECT/CT imaging.
–– Urinary bladder wall 0.2 mGy/MBq On the other hand, 131I-MIBG is used for therapy.
–– Uterus 0.017 mGy/MBq MIBG scintigraphy is indicated to image tumors of neu-
–– Lower large bowel 0.0075 mGy/MBq roendocrine origin, particularly those of neuroectodermal
nature, such as pheochromocytomas, paragangliomas, and
The main radiation dosimetry estimates to patients fol- neuroblastomas. Nevertheless, other neuroendocrine tumors
lowing intravenous administration of 131I-Hippuran are (e.g., carcinoids, medullary thyroid carcinomas) can also be
reported here below [12], normalized to unit of administered visualized (Fig. 2.17). It is an important diagnostic tool for
activity: staging, for evaluating response to treatment, and in patients
who are candidate to 131I-MIBG therapy.
• Effective dose 0.066 mSv/MBq In addition, MIBG scintigraphy can be employed to
• Tissues/organs with the highest values of absorbed dose: investigate disorders of sympathetic innervation, for exam-
–– Urinary bladder wall 0.96 mGy/MBq ple, in ischemic and nonischemic cardiomyopathy, as well as
–– Uterus 0.035 mGy/MBq in patients with movement disorders (to distinguish idio-
–– Kidney 0.03 mGy/MBq pathic Parkinson’s disease from multisystem atrophy) and
for functional studies of the adrenal medulla.
Following i.v. injection, about 50% of the administered
Key Learning Points
activity is excreted in the urine within the first 24 h, 70–90%
• I-hippuran is an anionic structural analog of para-
123
of the activity being cumulatively excreted within 48 h. Since
aminohippurate, a natural substance that is nor-
>95% of injected activity is excreted through the kidneys, the
mally excreted in the urine by both tubular secretion
bladder and the excretory urinary tract can be visualized on
and glomerular filtration.
the scan; about 3% of MIBG activity is eliminated via the
• It is used for dynamic renal scintigraphy and “effec-
gastrointestinal tract. The normal adrenal glands are visual-
tive” renal plasma flow (ERPF) calculation.
ized in 15% of the cases at 48–72 h after injection.
48 F. Orsini et al.
a b
Fig. 2.17 Examples of planar whole-body imaging with 123I-MIBG (b) Abnormally increased focal uptake in the right adrenal gland in a
(anterior view on the left, posterior view on the right). (a) Pattern of patient with suspected pheochromocytoma (optimal blocking of thyroi-
normal physiologic distribution, without abnormal areas with focally dal uptake of radioiodide released during metabolic degradation of
increased tracer uptake (no thyroid-blocking protocol was employed). 123
I-MIBG)
MIBG does not bind to the postsynaptic adrenergic receptors (see also Chap. 46 of this book “Current practical guidelines
and is not degraded by the enzymes normally degrading cate- for the most common nuclear medicine procedures”).
cholamines, monoamine oxidase (MAO), and catecholamine- Furthermore, radioiodide released during degradation of the
O-methyl transferase (COMT). Background activity visible in radiopharmaceutical can accumulate in the gastrointestinal
scans acquired early postinjection is related to early binding to tract as well as in the thyroid. Therefore, a thyroid-blocking
circulating platelets and to nonspecific accumulation of the protocol is recommended, especially in children, by admin-
radiopharmaceutical (mainly by passive diffusion) in tissues istering supersaturated potassium iodide solution or Lugol’s
such as the liver, spleen, lungs, salivary glands, thyroid gland, solution starting 24–48 h before radiopharmaceutical
nasal mucosa, colon, skeletal muscles, and uterus. Since there is administration and continuing for at least 3 days, according
no physiological MIBG uptake in the bone, any focal or diffuse to protocols published in international guidelines.
uptake in the bones are to be considered as skeletal metastasis The main radiation dosimetry estimates to patients fol-
and/or pathological involvement of the bone marrow. lowing intravenous administration of 123I-MIBG are reported
Due to a certain “mass” effect of radiolabeled MIBG (espe- here below, normalized to unit of administered activity [12]:
cially for low-specific activity 131I-MIBG), administration may
induce the release into circulation of catecholamines previ- • Effective dose 0.018 mSv/MBq
ously stored in neurosecretory granules, with ensuing possible • Tissues/organs with the highest values of absorbed dose:
pharmacologic effects such as tachycardia and an arterial –– Liver 0.071 mGy/MBq
hypertensive crisis in patients with pheochromocytoma. Slow –– Urinary bladder wall 0.07 mGy/MBq
i.v. infusion of the radiopharmaceutical (over 1–5 min for a –– Spleen 0.02 mGy/MBq
diagnostic procedure) is therefore recommended.
Moreover, since many drugs can interfere with MIBG The main radiation dosimetry estimates to patients fol-
uptake, the patient’s medication list should be reviewed and lowing intravenous administration of 131I-MIBG are reported
interfering drugs discontinued before MIBG administration here below, normalized to unit of administered activity [12]:
• Effective dose 0.2 mSv/MBq mine receptors in the presynaptic endings of dopaminergic

• Tissues/organs with the highest values of absorbed dose: neurons in the substantia nigra whose ending axons connect
–– Liver 0.83 mGy/MBq with the postsynaptic neurons in the putamen and caudate
–– Urinary bladder wall 0.59 mGy/MBq (striatum nucleus) using dopamine as the neurotransmitter.
–– Spleen 0.49 mGy/MBq Dopamine is produced in the presynaptic nerve endings and
stored into vesicles by the vesicular monoamine transporter
2 (VMAT-2, an integral membrane protein that transports
Key Learning Points neurotransmitters such as dopamine from the cytosol into
• Radioiodinated MIBG (or iobenguane) is a struc- vesicles). Upon excitation, dopamine is released from these
tural analog of catecholamines. vesicles into the synapse and binds to the predominantly
• MIBG can be labeled with either 123I (for diagnostic postsynaptic dopamine receptors. On the presynaptic side,
use only) or 131I (for both diagnostic and therapeutic the DAT system removes dopamine out of the synaptic cleft
uses) for imaging the adrenal medulla and tumors and back into the nigrostriatal axon endings for either stor-
of neuroectodermal nature, such as pheochromocy- age or degradation.
tomas, paragangliomas, and neuroblastomas. 123
I-FP-CIT scintigraphy is performed to evaluate integ-
• 123I-MIBG is also commonly used to investigate the rity of the dopaminergic nigrostriatal pathway. The main
distribution of adrenergic innervation in the myo- clinical indication of brain scintigraphy with 123I-FP-CIT is
cardium, in patients with ischemic cardiomyopathy, for the identification of patients with essential tremor (or
as well as in patients with movement disorders. some other conditions such as drug-induced Parkinsonism
and psychogenic Parkinsonism) versus those with presynap-
tic Parkinsonian syndromes (including idiopathic Parkinson’s
disease and atypical Parkinsonisms such as multistage atro-
2.5.4 123I-N-ω-FluoroPropyl- phy, progressive supranuclear palsy, or cortico-basal degen-
2βCarbomethoxy-3β-(4-Iodophenyl)- eration). In patients with presynaptic Parkinsonism,
nor-Tropane (123I-FP-CIT) scintigraphy neuronal degeneration in the substantia nigra
causes reduced uptake of 123I-FP-CIT in the putamen and
This radiopharmaceutical (also known as 123I-ioflupane) is a caudate, whereas tracer uptake is normal in essential tremor
radioiodinated cocaine analog in which the external bond and in the other forms of Parkinsonism.
between the aromatic ring and the group tropane is replaced An additional indication for scintigraphy with 123I-FP-
with a direct C=C bond (Fig. 2.18). CIT is in the differential diagnosis between dementia with
As an antagonist of the dopamine transporter (DAT), Lewy bodies (where tracer uptake is reduced because of
123
I-FP-CIT binds with high affinity and selectivity to dopa- nigrostriatal degeneration) and Alzheimer’s disease (where
integrity of the nigrostriatal system is preserved).
F
The standard injected activity of 123I-FP-CIT is 111–
O CH3
185 MBq, and SPECT acquisition is performed within 3–6 h
O after administration. Due to the release of 123I-iodide during
N
tracer catabolism, a thyroid-blocking protocol is recom-
mended, by administering supersaturated potassium iodide
123 solution or Lugol’s solution.
l
Uptake of 123I-FP-CIT in the brain is very rapid: about
7% of the injected activity is retained in the cerebral tissue
as early as 10 min after i.v. administration (30% of this
amount being taken up in the striatum). At 5 h postinjection,
the brain-retained activity has already decreased to 3%.
Blood clearance of 123I-FP-CIT is rapid, so that 5 min after
injection, only 5% of the radiopharmaceutical is still present
in circulating blood. Excretion occurs predominantly
through the kidneys (>60% at 48 h), while approximately
15% of the injected activity is excreted through the gastro-
intestinal tract.
Although very rarely, some side effects have been
Fig. 2.18 Chemical and tridimensional structure of 123I-FP-CIT. Color
codes: magenta = 123I; red = O; white = H; light blue = C; blue = N; reported, such as headache, tingling, dizziness, increased
gray = F appetite, and pain at the intravenous injection site.
50 F. Orsini et al.
Since uptake of 123I-FP-CIT is receptor-mediated, atten- pathic Parkinson’s disease (with generally preserved—or
tion must be paid to possible interference by other medica- even upregulated D2 postsynaptic receptor activity) and
tions/drugs involved in the dopaminergic system that may other Parkinsonisms (with degeneration of the D2 postsyn-
alter the specific tracer binding; if possible, such medications aptic structures). Additional indications include the evalua-
should be discontinued for at least five half-lives. The main tion of D2 receptor blockade by neuroleptics, suspected
substances that may interfere with the results of the scan Wilson’s disease and Huntington’s disease, and pituitary
with 123I-FP-CIT are amphetamines, benzatropine, bupro- adenoma (the presence of D2 receptors may have implica-
pion, cocaine, and sertraline. tions for the therapeutic strategy).
The main radiation dosimetry estimates to patients follow- Furthermore, scintigraphy with 123I-IBZM has also been
ing intravenous administration of 123I-FP-CIT are reported performed in patients with extra central nervous system con-
here below, normalized to unit of administered activity: ditions, i.e., to image melanoma lesions. The exact mecha-
nism thereby 123I-IBZM is taken up by melanoma cells is not
• Effective dose 0.025 mSv/MBq clear yet. In addition to the increased arterial flow typical of
• Tissues/organs with the highest values of absorbed dose: tumor neoangiogenesis, 123I-IBZM might bind to D2 recep-
–– Liver 0.085 mGy/MBq tors expressed on the melanoma cell membrane, and/or it
–– Colonic wall 0.059 mGy/MBq might interact specifically with some intracellular structures,
–– Spleen and gallbladder wall 0.044 mGy/MBq such as melanin and its precursors.
Following i.v. injection, about 75% of 123I-IBZM circu-
lating in the blood is bound to plasma proteins; only about
Key Learning Points 3–4% of the injected activity crosses the blood-brain bar-
• I-FP-CIT (or ioflupane) is a radioiodinated
123 rier due to its lipophilic characteristics, neutral charge, and
cocaine analog able to bind with high affinity and small size. The majority of circulating 123I-IBZM is in fact
selectivity to dopamine transporter (DAT) in the quickly converted into two hydrophilic metabolites that are
presynaptic endings of basal ganglia. therefore no longer able to diffuse through the blood-brain
• It is commonly used for brain SPECT when investi- barrier. The fraction of 123I-IBZM entering the brain is
gating patients with movement disorders and/or taken up by the postsynaptic D2 receptors present in the
neurodegenerative diseases. striatum dopaminergic system. Optimal target/background
ratio is reached about 40 min after injection, plateauing up
to over 3 h.
The radiopharmaceutical undergoes predominant urinary
2.5.5 123I-Iodobenzamide (123I-IBZM) excretion (about 40% at 24 h and 60% at 48 h after adminis-
tration). As with other 123I-labeled radiopharmaceuticals,
I-IBZM (also known as 123I-iolopride; see Fig. 2.19) tar-
123
adequate patient’s preparation is recommended to prevent
gets the D2 postsynaptic receptors in the striatum nucleus thyroid uptake of radioiodide released during the metabolic
(putamen and caudate). degradation of the tracer.
The main clinical indication for the use of this radiophar- The main radiation dosimetry estimates to patients fol-
maceutical is for the differential diagnosis between idio- lowing intravenous administration of 123I-IBZM are reported
here below, normalized to unit of administered activity [12]:

H
–– Thyroid 0.16 mGy/MBq
N
O NH CH3 –– Lower large bowel wall 0.064 mGy/MBq
HO O
CH3
Key Learning Points
123I
• I-IBZM (or iolopride) targets the D2 postsynaptic
123
receptors of the basal ganglia.

• It is used for brain SPECT when investigating
Fig. 2.19 Chemical and tridimensional structure of 123I-IBZM. Color patients with movement disorders.
codes: magenta = 123I; red = O; white = H; light blue = C; blue = N
2.6 201Tl-Chloride quality of gamma camera imaging with 201Tl is suboptimal

due to the low energy of the X-ray emission that implies sig-
The monovalent cation Tl+ has biochemical characteristics nificant scatter and attenuation by soft tissues. Furthermore,
similar to those of the K+ ion. Following i.v. injection, 201Tl there is a relatively high radiation burden to patients, which
is rapidly cleared from the blood because of diffuse uptake limits the amount of activity that can be administered, there-
into cells of all organs and tissues, with a pattern of distribu- fore also counting statistics, which leads to low signal-to-
tion that is largely determined by the magnitude of regional noise ratios—especially in obese patients.
blood flow. Therefore, in excitable cells (typically neurons The main radiation dosimetry estimates to patients fol-
and muscle cells), 201Tl undergoes the continuous trans- lowing intravenous administration of 201Tl-chloride are
membrane flux that physiologically occurs along an electro- reported here below, normalized to unit of administered
potential gradient during depolarization (passive transport) activity [12]:
and subsequent repolarization (an energy-dependent pro-
cess linked to the Na+/K+ pump). These sustained trans- • Effective dose 0.14 mSv/MBq
membrane exchanges explain the high intracellular • Tissues/organs with the highest values of absorbed dose:
concentration of Tl+ ions, mimicking the distribution of the –– Kidneys 0.48 mGy/MBq
K+ ions. –– Bone surface 0.38 mGy/MBq
Because of the typical high accumulation in myocardial –– Large bowel wall 0.25 mGy/MBq
cells (that physiologically undergo continual, sequential
depolarization, and repolarization), 201Tl has been used to
investigate myocardial perfusion; in fact, myocardial uptake
Key Learning Points
increases proportionally with perfusion, and prolonged
• The monovalent cation 201Tl+ of 201Tl-chloride has
retention depends on the integrity of cell membrane, hence
biochemical characteristics similar to those of the
on myocardiocyte viability. Since 201Tl is not trapped in myo-
K+ ion.
cytes or in other tissues, redistribution in rest condition of
• In addition to its main use for myocardial perfusion
201
Tl occurs over several hour following administration at the
scintigraphy, 201Tl-chloride has been used for tumor
peak of a stress test. This redistribution process renders the
imaging as an indicator of cell viability, especially
Tl ion available for possible myocardial accumulation at
201 +
in patients with brain tumors, and also for parathy-
rest in regions that were ischemic when 201Tl was injected at
roid scintigraphy.
peak stress, thus allowing redistribution images to be
acquired that are fairly independent of perfusion and mainly
reflect viability.
Besides its uptake in myocardial tissue, the 201Tl+ ions
tend to accumulate in all metabolically active cells, including 2.7 67Ga-Citrate
cancer cells. 201Tl-Chloride has therefore been used for tumor
imaging as an indicator of cell viability, particularly for dis- The cyclotron-produced radionuclide 67Ga (a metal element)
criminating viable tumor tissue from fibrous scar tissue after is obtained by proton irradiation of enriched zinc. It has a
surgery and/or radiation therapy (especially in patients with physical half-life of 3.26 days and decays by electron capture
brain gliomas) and also for parathyroid scintigraphy (where to stable zinc (67Zn) emitting γ-rays with multiple energy
this agent has however been completely replaced with peaks, principally 93 keV (39%), 185 keV (21%), 300 keV
99m
Tc-sestamibi). (17%), and 394 keV (7%).
Following i.v. administration, about 85% of 201Tl+ ions are Following i.v. injection of 67Ga-citrate, 67Ga2+ ions form at
cleared from the blood during a single circulation, so that physiological pH; the metabolic pathways in the body of
about 2 min after administration <5% of the injected activity these metallic ions are similar to those of the Fe2+ ions. About
is still circulating. About 5% of the injected activity localizes 90% of injected activity binds to plasma proteins, especially
in the myocardium in proportion to regional perfusion and transferrin, with ensuing slow blood clearance. During the
viability. The tracer has a biological half-life of approxi- first 24 h, up to 25% of the injected activity is excreted
mately 10 days in the body. through the kidneys, while at later time points, the main
201
Tl is a cyclotron-produced radionuclide that decays by excretion route becomes the bowel. By 48 h postinjection,
electron capture with a relatively long physical half-life about 75% of the injected activity remains in the body and is
(73.1 h) and emission of γ-photons of 135 and 167 keV (12% equally distributed among the liver, bone and bone marrow,
abundance) and X-rays of energy 67–82 keV (88% abun- and soft tissues. Less intense physiologic uptake is usually
dance, which is the emission on which gamma camera imag- observed in the nasopharynx, lacrimal glands, thymus,
ing is based). Because of these physical characteristics, the breasts, and spleen.
52 F. Orsini et al.
Although not yet fully clarified, most of this process is

due to intracellular accumulation mediated by the transfer- Key Learning Points
rin receptors. The main mechanism of accumulation in • Ga-Citrate is a historical radiopharmaceutical
67
tumor and inflammation sites appears to be linked to the with a quite complex mechanism of accumulation
complex formed by 67Ga2+ with transferrin, for which mem- that has had a role both in benign conditions (char-
brane receptors are present ubiquitously but are overex- acterization and localization of infection/inflamma-
pressed especially on membranes of cells undergoing active tion and chronic granulomatous diseases such as
proliferation and metabolic activation, such as tumor and sarcoidosis) and in neoplastic conditions (especially
inflammatory cells. lymphomas and sarcoma).
An additional mechanism of 67Ga accumulation is medi- • Currently, the traditional applications of scintigra-
ated by its binding to lactoferrin, generally located in the phy with 67Ga-citrate have mainly been replaced
extracellular inflammatory spaces. Thus, 67Ga reaching with [18F]FDG PET imaging.
inflammation sites bound to transferrin may migrate to lacto-
ferrin, which in turn binds to specific receptors present on
macrophages’ surface. Enhanced expression of lactoferrin 2.8 Lung Ventilation
receptors has been shown also in lymphoma cells, a factor Radiopharmaceuticals
that seems to contribute for 10–25% of total uptake to 67Ga
accumulation in lymphomas. Lung ventilation scintigraphy is indicated to assess regional
Following its introduction into the clinical practice in distribution of ventilation in different clinical conditions,
the late 1960s [16], 67Ga-citrate has had a well-defined role ranging from acute pulmonary embolism to chronic obstruc-
both in benign conditions (localization of infection/inflam- tive pulmonary disease. The radiopharmaceuticals that have
mation sites, characterization of chronic granulomatous been used or are currently used for lung ventilation scintigra-
diseases such as sarcoidosis) and in neoplastic conditions phy can be divided into two classes:
(especially lymphomas and sarcoma, for staging, progno-
sis, and evaluation of residual disease after treatment). • Radioaerosols, i.e., a two-phase system consisting of par-
However, the traditional applications of scintigraphy with ticles that may be liquid, solid, or a combination sus-
67
Ga-citrate have currently been replaced with [18F]FDG pended in gas, mapping distribution of the ventilated lung
PET wherever a PET scanner is available. volume (usually with static imaging at equilibrium)
Images should be ideally acquired 48 h after i.v. admin- • Radioactive gases that allow dynamic single-breath imag-
istration, when a satisfactory tumor-to-background ratio is ing either at equilibrium or during the washout phase
reached; sometimes images are also recorded at 72 and
96 h to determine if subtle regions of activity become more Special nebulizers are employed to produce radioaerosols
prominent above background activity. SPECT or SPECT/ using 99mTc-DTPA, 99mTc-Technegas, or 99mTc-nanocolloidal
CT imaging may enhance reliability of the scan, although albumin as the suspended phase.
the γ-emission energies of this radionuclide are subopti-
mal for imaging and the high radiation burden limits the
maximum activity that can safely be administered to 2.8.1 99mTc-DTPA Aerosol
patients.
The standard activity administered to an adult individual This radiopharmaceutical is the most common agent
is 150–220 MBq but up to 330 MBq in large-sized patients. employed to produce radiolabeled liquid aerosols, using a
The main radiation dosimetry estimates to patients follow- nebulizer to produce particles (ideal size 0.1–0.2 μm, maxi-
ing intravenous administration of 67Ga-citrate are reported mum diameter <2 μm) capable of being readily inhaled. Air
here below, normalized to unit of administered activity [3]: or oxygen is forced through the nebulizer at 30–50 psi to pro-
duce aerosol droplets that are then inhaled by the patient
• Effective dose 0.1 mSv/MBq through a mouthpiece. Exhaled air from the patient is trapped
• Tissues/organs with the highest values of absorbed dose: in the filter attached to the aerosol unit, thus preventing any
–– Bone surface 0.63 mGy/MBq contamination of the surrounding area with radioactive
–– Red bone marrow 0.21 mGy/MBq materials. About 10% of the activity is deposited in the lungs
–– Large bowel wall 0.16 mGy/MBq of a subject with normal respiratory function, while the
remainder remains airborne and is exhaled. The pattern of 2.8.3 Xenon-133

deposition of the radioactive droplets within the lung depends
on their size, shape, density, and electrical charge. The larger Xenon-133 (133Xe) is a noble (i.e., chemically inert), lipo-
particles tend to settle in the central area, while the smaller soluble radioactive gas, which is historically the first agent
particles deposit in the peripheral areas. that was used for lung ventilation studies [18]. Because of its
The biological half-life in the lung is 55–108 min in lipophilicity, this gas diffuses rapidly also through the blood-
healthy individuals and 15–33 min in healthy smokers. brain barrier and has therefore been used in the past to mea-
Clearance of radioactivity deposited in the alveolar region sure in a quantitative manner rCBF (expressed as milliliter/
occurs by transepithelial diffusion; 99mTc-DTPA enters then minute per unit mass) and also myocardial blood flow (fol-
the general circulation and is excreted by urinary clearance. lowing intracoronary artery injection).
Even though the activity of 99mTc-DTPA dispensed in the 133
Xe has a physical half-life of 5.3 days and decays through
nebulizer is approximately 900–1300 MBq, the activity actually emission of γ-rays, associated however with some β- emission.
reaching the lungs is generally <100 MBq. This amount is much The energy of its γ-emission photopeak (81 keV) is lower than
lower than the activity of 99mTc-MAA normally injected i.v. for that of 99mTc (140 keV); therefore, when performing a ventila-
a lung perfusion scan; therefore, when planning a ventilation/ tion/perfusion scan in patients with suspected acute pulmo-
perfusion scan in patients with suspected acute pulmonary nary embolism, lung perfusion scintigraphy with 99mTc-MAA
embolism, the 99mTc-DTPA ventilation scan (with multi-view is acquired after the 133Xe ventilation study.
planar imaging or with SPECT) should be performed first. Upon inhalation in an airtight closed system, 133Xe dif-
The main radiation dosimetry estimates to patients following fuses through the alveolar wall and enters the pulmonary
administration of 99mTc-DTPA as a radioaerosol are reported venous circulation via the capillaries to reach peripheral
here below, normalized to unit of administered activity [12]: organs/tissues through arterial circulation. When it returns to
the lung via venous blood, 133Xe is exhaled after a single
• Effective dose 0.1 mSv/MBq pass, with a very short biological half-life (30 s).
• Tissue/organ with the highest value of absorbed dose: 133
Xe single deep-breath inspiration, equilibrium, and
–– Urinary bladder wall 0.047 mGy/MBq washout images are acquired using a standard gamma cam-
era. The standard activity administered for a lung ventilation
study is 200–750 MBq.
2.8.2 99mTc-Technegas For radiation safety and in order to avoid dispersion of
the radioactive gas in the environment, the room where the
The aerosol 99mTc-technegas, mostly used in Europe, con- scan is performed must be equipped with a special closed
sists of 99mTc-labeled solid graphite particles (diameter system for inhaling 133Xe, and the air conditioning system
0.005–0.2 μm) generated at high temperature [17]. These must include an activated charcoal filter for trapping any
“ultrafine” particles tend to grow by aggregation; thus, 99mTc- radioactive gas before release of the filtered air in the
technegas should be used within 10 min from generation. environment.
The patient (who is sitting with his/her back lying against The main radiation dosimetry estimates to patients fol-
the gamma camera collimator) inhales the radioaerosol lowing inhalation of 133Xe are reported here below, normal-
through closed, single-use circuit breathing tubes, until a ized to unit of administered activity:
counting rate of 2000 counts/s is attained. Inhaled 99mTc-
technegas particles adhere to the walls of the alveoli and are • Effective dose 0.00071 mSv/MBq
slowly cleared by reabsorption. • Tissue/organ with the highest value of absorbed dose:
The images obtained with 99mTc-technegas have superior –– Lung parenchyma 0.0041 mGy/MBq
quality than those obtained with 99mTc-DPTA aerosol or
133
Xe (see further below); moreover, its use has minimized
the problem of particle clumping in patients with chronic Key Learning Points
obstructive pulmonary disease, which is instead observed • Radiopharmaceuticals for lung ventilation scintig-
when using 99mTc-DTPA aerosol. The administered activity raphy are administered by inhalation.
to the lung is 20–30 MBq. • Radioaerosols such as 99mTc-DTPA, 99mTc-techne-
The main radiation dosimetry estimates to patients following gas, or 99mTc-nanocolloidal albumin are produced
administration of 99mTc-technegas as a radioaerosol are reported by special nebulizers allowing inhalation of the
here below, normalized to unit of administered activity [3]: suspended phase containing the radioactive
particles.
• Effective dose 0.015 mSv/MBq • The radioactive gas historically employed for lung
• Tissue/organ with the highest value of absorbed dose: ventilation is 133Xe.
–– Lung parenchyma 0.011 mGy/MBq
54 F. Orsini et al.
2.9 Other Radiopharmaceuticals formation of thrombi; its main indication was therefore for
the detection of acute deep vein thrombosis in lower
Over the years, many other radiopharmaceuticals have been extremities.
developed in addition to those described in the previous sec- 99m
Tc-Glucoeptonate has been used for static renal scin-
tions. Some of these radiopharmaceuticals are no longer tigraphy, while 51Cr-EDTA and 125I-iothalamate can still be
employed in the clinical practice, mostly because they have used for measuring the glomerular filtration rate through
been replaced with others, e.g., with more specific binding or techniques based on in vitro measurements (i.e., using only
with more favorable radiation exposure characteristics, or blood sampling, without scintigraphic imaging).
are not commercially available in Europe and/or the Finally, several novel peptide-based radiopharmaceuti-
USA. Other single-photon-emitting radiopharmaceuticals cals are being evaluated for the visualization and character-
have not yet proceeded through all phases of clinical trial, ization of neuroendocrine tumors. In addition to
also because most of the attention is currently devoted to the 99m
Tc-depreotide, new radiolabeled SST analogs with more
development of PET radiopharmaceuticals. favorable characteristics than 111In-pentetreotide or 99mTc-
Among radiopharmaceuticals for nuclear cardiology, EDDA/HYNIC-TOC are under investigation. Besides
mention should be made of tracers developed for assessing ligands for the SST receptors, other receptor/ligand systems
myocardial perfusion (such as 99mTc-teboroxime) or myocar- are also under intensive investigations, such as those based
dial metabolism (123I-labeled fatty acids, used predominantly on, e.g., bombesin, bioactive fragments of cholecystokinin,
in Japan), radiopharmaceuticals localizing in acute myocar- and vascular-active intestinal peptide.
dial infarction (such as 99mTc-pyrophosphate, 111In-antimyosin
antibodies, 99mTc-glucarate, 99mTc-HYNIC-annexin-V), and
radiopharmaceuticals with preferential accumulation in 2.9.1 P
erspectives in Molecular Imaging
hypoxic tissues (used also in the cerebral or tumor areas). Based on Single-Photon-Emitting
Radiolabeled analogs of amphetamine diffusing through Radiopharmaceuticals
the blood-brain barrier (e.g., 123I-IMP and 123I-HIPDM) have
been used for assessing rCBF, while 123I-iomazenil displays Advancing knowledge in the pathophysiology of tissues/
high affinity for central-type benzodiazepine receptors— organs opens new opportunities for the development of novel
constituting therefore a useful marker of neuronal viability. radiopharmaceuticals, most of which are not approved yet
131
I-6-iodomethyl-norcholesterol and 75Se-norcholesterol for commercial use. Intensive search is ongoing especially
are radiolabeled analogs of cholesterol that have been used for tumor-targeting agents with better specificity and sensi-
for scintigraphy of the adrenal cortex. Because of this prop- tivity. The features that tumor cells exhibit regarding expres-
erty, indications to their use included evaluation of patients sion of intra- and extracellular molecules are often different
with hypercortisolemia (Cushing’s disease), hyperaldoste- from those exhibited by normal, nonmalignant transformed
ronism, and hyperandrogenism and in the case of adrenal cells. These attributes constitute the basis for developing new
incidentaloma (to evaluate the nature of the lesion), as well imaging targets for diagnostic applications as well as for
as identification of any ectopic adrenal tissue or adrenal developing new anticancer drugs and for assessing tumor
residual tissue after surgery. However, both tracers (espe- response to treatment. The successful choice of a target mol-
cially 75Se-norcholesterol) suffer from important limitations ecule potentially leads to the development not only of a
in terms of both emission energy and, above all, radiodosi- molecular imaging probe but also of a therapeutic agent
metric burden to the patient. capable to inhibit the disease process, an approach that is
The radiolabeled monoclonal antibody Fab′ fragment now defined as “theragnostics.”
99m
Tc-arcitumomab is directed against the carcinoembryonic Tumor cells often display altered energy metabolism, as
antigen (CEA), typically overexpressed by colorectal cancer reflected, for example, in increased glucose uptake and
cells (but also other tumors, both of the gastrointestinal tract shifted balances in metabolic products. Thus, at the preclini-
and of various organs and/or tissues). This radiopharmaceu- cal level, a variety of single-photon-emitting tracers are
tical is no more employed in the clinical practice, as it has under evaluation for use as markers for (neo)angiogenesis,
been replaced with [18F]FDG for PET/CT imaging with apoptosis, hypoxia, acidosis, metabolic activity, and proteo-
clearly superior diagnostic performance. The main indica- lytic activity.
tion for 99mTc-arcitumomab scintigraphy was restaging of Angiogenesis represents an interesting molecular target
colorectal cancer patients carcinoma with rising serum CEA not only for imaging but also for targeted forms of therapy.
levels posttreatment. It consists of formation of new vessels as an essential pro-
99m
Tc-Apcitide has been proposed to visualize recent cess in the growth of solid tumors. Examples for target
venous thrombosis. It binds to the GP IIb/IIIa receptors on anti-angiogenic therapies currently used in the clinical
activated platelets that are responsible for aggregation and practice are cilengitide (that inhibits integrin receptors
αvβ3 and αvβ5) and bevacizumab, an antibody targeting the human malignancies and is often associated with an aggres-
vascular endothelial growth factor (VEGF). Different sive tumor phenotype and poor prognosis. “Cold” (i.e., non-
potential molecular targets to monitor angiogenesis are radiolabeled) monoclonal antibodies against this receptor
potentially available for imaging purposes. The most suit- have been developed for therapeutic purposes, including
able candidates for the development of new radiotracers cetuximab and trastuzumab; other small molecules (such as
are currently represented by integrin antagonists, extracel- gefitinib and erlotinib) that act as tyrosine kinase inhibitors
lular matrix protein inhibitors, or matrix metalloprotein- (TKIs) target the receptor catalytic domain of EGFR. Several
ase, as well as by ligands binding to tyrosine kinases or single-photon-emitting radiopharmaceuticals have been
growth factor receptors. developed for preclinical trials, including 99mTc-cetuximab
Integrins are heterodimeric membrane receptors consti- or 111In-DTPA-pertuzumab.
tuted by α- and β-subunits that mediate the interactions Apoptosis is a process of regulated (or programmed) cell
between cells and the extracellular matrix and soluble mol- death that, in its classic form, does not cause inflammation.
ecules (such as growth factors). So far 18 different Upon a variety of molecular signals that activate apoptosis,
α-subunits and 8 different β-subunits have been identified, the lipid composition of the outer and inner leaflets of the
corresponding to 24 different integrin receptors. Integrin plasma membrane is altered, so that some molecules such as
αvβ3 is being intensively investigated in oncology, because phosphatidylserine and phosphatidylethanolamine (that are
it is highly expressed on the cell surface of activated endo- normally retained on the intracellular side of the cell mem-
thelial cells in newly formed blood vessels. In the preclini- brane) are exposed on the outer side of the cell membrane.
cal setting, a large variety of imaging strategies have been The most promising single-photon-emitting radiopharma-
successfully employed for imaging integrin expression. All ceuticals currently under investigation that target phosphati-
the tracers that are used for imaging of integrin expression dylserine and phosphatidylethanolamine are represented by
are based on the tripeptide sequence arginine-glycine- radiolabeled annexin-V (99mTc-HYNIC-annexin-V) and by
aspartic acid (or RGD, in the single letter code). RGD binds duramycin (99mTc-duramycin).
to the integrin containing the αv subunit, a physiologic inte-
grin-binding ligand that is abundant in proteins of the extra-
cellular matrix. Accordingly, a variety of radiolabeled Key Learning Points
RGD-based peptides have been developed for noninvasive • Over the years, many other radiopharmaceuticals
imaging of integrin αvβ3 expression with either single-pho- have been developed in addition to those described
ton imaging or, in most of the cases, PET. 99mTc-labeled in the previous sections.
RGD peptides have been the subject of few investigations, • Some of these radiopharmaceuticals are no longer
some of these peptides having proceeded to the stage of employed in the clinical practice mostly because
clinical investigation, such as 99mTc-aP2, 99mTc-NC100692 they have been replaced with others with more spe-
(also known as 99mTc-maraciclatide), 99mTc-3PRGD2, cific binding or with more favorable radiation expo-
99m
Tc-RGD-BBN, and 99mTc-labeled RGD dimeric peptides sure characteristics.
with PEG4 and Gly3. • On the other hand, the current efficacy of the molec-
Among the extracellular matrix proteins, the polymorphic ular targeting of new several anticancer drugs has
matrix protein fibronectin belongs to the family of universal increased interest on the development of novel
cell adhesion molecules involved in vascular proliferation radiopharmaceuticals not only for diagnostic appli-
and widely expressed in neoplastic tissues. Fibronectin is the cations but also for therapeutic purpose. This
target of the experimental tracer 99mTc-AP39. approach is now defined as “theragnostics.”
Also matrix metalloproteinases are potential targets for
molecular imaging. They are involved in the degradation of
the basement membrane and of the extracellular matrix,
thus facilitating the migration of endothelial cells and for- References
mation of new vessels, both processes being typically
enhanced in malignant tissues. Overexpression of matrix 1. Lassmann MS, Treves T, For the EANM/SNMMI Paediatric
Dosage Harmonization Working Group. Paediatric radiophar-
metalloproteinases correlates with tumor aggressiveness maceutical administration: harmonization of the 2007 EANM
and metastatic potential; in this regard, hydroxamate-based paediatric dosage card (version 1.5.2008) and the 2010 North
tracers labeled with 111In, 99mTc, or 123I display promising American consensus guidelines. Eur J Nucl Med Mol Imaging.
binding properties. 2014;41:1636. Accessible at: http://www.eanm.org/publications/
dosage-card/.
The epidermal growth factor receptor (EGFR) is involved 2. Williams JG. Pertechnetate and the stomach. A continuing contro-
in the regulation of cell growth, as well as in the differentia- versy. J Nucl Med. 1983;24:63–6.
tion and survival of cells. It is overexpressed in several
56 F. Orsini et al.
3. ICRP. Radiation dose to patients from radiopharmaceuticals: a 17. Kropp J, Buhr W, Bockisch A, Grünwald F, Ruhlmann J, Hotze
compendium of current information related to frequently used sub- A, Biersack HJ. Inhalation scintigraphy of the lung using the new
stances. ICRP publication 128. Ann ICRP. 2015;44(2S):7. ultrafine aerosol Technegas. Nuklearmedizin. 1989;28:46.
4. Eckelman W, Richards P. Instant 99mTc-DTPA. J Nucl Med. 18. Knipping HW, Bolt W, Venrath H, Valentin H, Ludes H, Endler P. A
1970;11:761. new method of heart and lung function testing, the regional func-
5. Fritzberg AR, Kasina S, Eshima D, Johnson DL. Synthesis and bio- tional analysis in the lung and heart clinic by the radioactive noble
logical evaluation of technetium-99m MAG3 as a hippuran replace- gas xenon 133 (isotope thoracography). Dtsch Med Wochenschr.
ment. J Nucl Med. 1986;27:111–6. 1955;80:1146–7.
6. Alazraki N, Aarsvold JN, Audisio RA, Glass E, Grant SF,
Kunikowska J, et al. The EANM and SNMMI practice guideline for
lymphoscintigraphy and sentinel node localization in breast cancer.
Eur J Nucl Med Mol Imaging. 2014;41:1259–60.
7. Fazzi P, Borsò E, Albertelli R, Mariani G, Giuntini C. Comparative
Suggested Readings
performance of two inhaler systems to assess distribution of
convective ventilation by 99mTc-labeled aerosol scintigraphy in Ell PJ, Gambhir SS, editors. Nuclear medicine in clinical diagnosis and
patients with airway obstruction. Q J Nucl Med Mol Imaging. treatment. 3rd ed. New York, NY: Churchill Livingston; 2004.
2009;53:428–36. Herbert JC, Eckelman WC, Neumann RD, editors. Nuclear medicine –
8. Abrams MJ, Davison A, Jones AG, et al. Synthesis and character- diagnosis and therapy. New York, NY: Thieme Medical Publishers;
ization of hexakis (alkylisocyanide) and hexakis (arylisocyanide 1996.
complexes of Technetium (I)). Inorg Chem. 1983;22:2798. IAEA. Good practice for introducing radiopharmaceuticals for clini-
9. Coakley AJ, Kettle AG, Wells CP, O’Doherty MJ, Collins RE. cal use. Vienna: International Atomic Energy Agency (IAEA; 2015.
99m
Tc-Sestamibi – a new agent for parathyroid imaging. Nucl Med IAEA. Operational guidance on hospital radiopharmacy. Vienna:
Commun. 1989;10:791–4. International Atomic Energy Agency (IAEA); 2008.
10. Khalkhali I, Mena I, Jouanne E, et al. Prone scintimammography in IAEA. Radiopharmaceuticals for sentinel lymph node detection: status
patients with suspicion of carcinoma of the breast. J Am Coll Surg. and trends. Vienna: International Atomic Energy Agency (IAEA);
1994;178:49–7. 2015.
11. Kelly JD, Forster AM, Higley B, et al. Technetium-99m-tetrofosmin IAEA. Technetium-99m radiopharmaceuticals: status and trends.
as a new radiopharmaceutical for myocardial perfusion imaging. J Vienna: International Atomic Energy Agency (IAEA); 2010.
Nucl Med. 1993;34:222. Kowalsky RJ, Falen SW, editors. Radiopharmaceuticals in nuclear
12. ICRP. Radiation dose to patients from radiopharmaceuticals. ICRP pharmacy and nuclear medicine. 3rd ed. Washington, DC: American
publication 53. Ann ICRP. 1988;18(1-4):149. Pharmacists Association; 2011.
13. Nickers NM, Collier B. Drugs inhibiting adrenergic nerves and Owunwanne A, Patel M, Sadek S, editors. The handbook of radiophar-
structures innervated by them. In: Goodman LS, Gilman A, editors. maceuticals. New York, NY: Springer; 1995.
The pharmacological basis of therapeutics. 5th ed. New York, NY: Schwochau K, editor. Technetium: chemistry and radiopharmaceuti-
Macmillan Publishing Co; 1975. p. 553–64. cals. Hoboken, NJ: Wiley; 2000.
14. Short JH, Darby TD. Sympathetic nervous system blocking agents. Subramanian G, Rhodes BA, Cooper JF, Sodd VJ. Radiopharmaceuticals.
III. Derivatives of benzylguanidine. J Med Chem. 1967;10:833–40. New York, NY: Society of Nuclear Medicine; 1975.
15. Wieland DM, Wu J, Brown LE, Mangner TJ, Swanson DP,
Theobald T, editor. Sampson’s textbook of radiopharmacy. 4th ed.
Beierwaltes WH. Radiolabeled adrenergic neuron-blocking agents: London: Pharmaceutical Press; 2010.
adrenomedullary imaging with[131I]Iodobenzylguanidine. J Nucl Vallabhajosula S. Molecular imaging – radiopharmaceuticals for PET
Med. 1980;21:349–53. and SPECT. Berlin: Springer; 2009.
16. Edwards CL, Hayes RL. Tumor scanning with 67Ga citrate. J Nucl Welch MJ, Redvanly CS, editors. Handbook of radiopharmaceuticals:
Med. 1969;10:103–5. radiochemistry and applications. Hoboken, NJ: Wiley; 2003.
Positron-Emitting
Radiopharmaceuticals 3
Piero A. Salvadori, Elena Filidei, and Assuero Giorgetti
Contents
3.1 Section I: Positron-Emitting Radionuclides and Labeling Strategies 58
3.1.1 Positron-Emitting Radionuclides and Basic Chemistry 58
3.1.2 Substrate Characteristics and Radiolabeling Strategies 60
3.1.3 Radiopharmaceuticals and Biological Systems 61
3.1.4 Production Flowchart and Related Issues 66
3.1.5 Basic of Automation and Synthesis Devices 67
3.1.6 18F-Labeled Radiopharmaceuticals 69
3.1.7 11C-Labeled Radiopharmaceuticals 73
3.1.8 68Ga-Labeled and Radiometal-Based Radiopharmaceuticals 75
3.1.9 Miscellanea 78
3.1.10 What Might Be Next 79
3.2 Section II: Positron-Emitting Radiopharmaceuticals for Clinical Use 79
3.2.1 Sodium 18F-Fluoride 79
3.2.2 Flow/Perfusion Agents 80
3.2.3 Energy Substrates 81
3.2.4 Substrates for Phospholipid Synthesis 85
3.2.5 Substrates for Protein Synthesis 86
3.2.6 Substrates for DNA Synthesis 89
3.2.7 Positron-Emitting Agents Based on Ligand-Acceptor Interaction 90
3.2.8 Amyloid Imaging Agents 92
3.2.9 Tissue Hypoxia Imaging Agents 93
3.2.10 Neoangiogenesis Imaging Agents 94
3.2.11 Apoptosis Imaging Agents 95
Further Reading 96
Learning Objectives • Describe the impact of molecular size, polarity, and func-
• Provide an overview of radionuclides with relevant appli- tional groups on distribution.
cations in PET. • Summarize key issues on handling positron-emitting
• Define structure-activity relationships and how radiola- radionuclides (shielding) and radiochemical synthesis.
beling can induce molecular/biological perturbation. • Explain main advantages and issues regarding 11C-labeled
• Describe routes of administration and their impact on radiopharmaceuticals.
biodistribution. • Provide basic information on 68Ga-labeled radiopharma-
ceuticals and the 68Ge/68Ga generator.
P. A. Salvadori (*) • Provide basic information on the synthesis and use of
PET/Cyclotron Unit, CNR Institute of Clinical Physiology, 124
I-labeled radiopharmaceuticals.
Pisa, Italy • Provide information on the rationale for using positron-
e-mail: salvador@ifc.cnr.it
emitting radiopharmaceuticals in different pathophysio-
E. Filidei · A. Giorgetti logic conditions according to the biologic/metabolic
Nuclear Medicine Unit, Fondazione CNR/Regione Toscana
event targeted by different classes of imaging agents.
Gabriele Monasterio, Pisa, Italy

58 P. A. Salvadori et al.
3.1 ection I: Positron-Emitting

S tomultiplier detectors were not invented until 1947) limited
Radionuclides and Labeling their clinical use.
Strategies Meanwhile, other radionuclides, such as 3H, 14C, and 131I
(and later on 99mTc), became more widely available, thus
Imaging with positron emission tomography (PET) is based on facilitating the use of radioactive compounds to trace bio-
the detection of the annihilation radiation. Annihilation radia- chemical processes in vivo. These tracer studies demon-
tion occurs when matter is converted to energy. This phenom- strated that many biochemical processes occur in vivo at a
enon occurs when a negatively charged electron (negatron) and very fast rate, so that the limitation of short physical half-life
a positively charged electron (positron, emitted in the course of might not be as crucial as expected. In parallel, more effi-
decay of a positron-emitting radionuclide) come in contact. cient scintillation detectors and coincidence detection elec-
The resulting positron/negatron (matter) disappears and is tronics were developed, which improved detection sensitivity
replaced by two gamma photons (energy). While electrons are and spatial resolution.
basic constituents of our physical world, positrons (or antielec-
trons) are very rare. They are produced by the decay of neu-
tron-deficient nuclei, which are not naturally occurring. 3.1.1 Positron-Emitting Radionuclides
The development of the cyclotron in the 1930s demon- and Basic Chemistry
strated that it is possible to generate new elements from a
target with a selected composition of nuclei. The cyclotron is PET radiochemistry entails production of the β+-emitting
the equivalent of the alchemists’ “philosophical stone,” since radionuclides and purification and incorporation of the β+-
it can transmute one element into a different one, thus facili- emitting radionuclide in diagnostic radiopharmaceutical.
tating separation of the new nucleus from the target matrix. Three major features of the radionuclide depend on the
It also allows the production of neutron-deficient nuclei, i.e., methodology used: (1) radionuclide chemical form(s), (2)
those prone to positron decay. specific radioactivity, and (3) radionuclidic and/or radio-
In the early 1940s, 11C, 13N, and 18F (i.e., the major mod- chemical impurities. It is now possible to rely on a number of
ern players for PET imaging) (Table 3.1) were the first radio- radiochemical building blocks, based on the “physiological”
nuclides produced with a cyclotron in the form of simple radionuclides (Table 3.2). Several radionuclides (Table 3.3)
compounds to be used as in vivo tracers. However, their short are currently being considered for further developments to
physical half-life and the poor detection capability for the address innovations in theragnostics (diagnostic and thera-
high-energy 511 keV annihilation photons (scintillation/pho- peutic use of the same construct).
Table 3.1 The “physiological” positron-emitting radionuclides

Specific radioactivity (GBq/μmol) Positron
Radionuclide Half-life (min) Decay Theoretical Practical Emax (keV) Range (H2O) mm
Oxygen-15 2.05 β+ 3.39 × 106 n.a. 1720 8.2
100%
Nitrogen-13 10.0 β+ 6.96 × 105 <15 1190 5.4
100%
Carbon-11 20.4 β+ 3.43 × 105 110–2600 970 4
99%
Fluorine-18 109.7 β+ 6.34 × 104 18–740 635 2.4
97%
Table 3.2 Nuclear reactions and most common chemical forms of positron-emitting radionuclides
Radionuclide Nuclear reaction Target composition Particle energy (MeV) Main chemical form(s)
Oxygen-15 15
N(p,n)15O 15
N2 (+O2 trace) 10 15
O2
14
N(d,n)15O 15
N2 (+O2 1–3%) 8
15
N2 (+CO2 1–3%) 8 C15O2
Nitrogen-13 16
O(p,α) N
13
H216O 11–18 13
NO2 + 13NO3
H216O + ethanol 11–18 13
NH3
Carbon-11 14
N (p,α)11C 14
N2 (+O2 trace) 10–13 11
CO2 + 11CO
14
N2 (+2–5% H2) 10–13 11
CH4
Fluorine-18 18
O(p,n)18F H218O 7–18 H18F
20
Ne(d,α)18F 20
Ne + 0.5–1% 19F2 8–14 18
F2 (+19F2)
Formalism adopted to describe nuclear reactions: A(x,y)B, where A, target nuclide; B, produced nuclide; x, bombarding particle; y, emitted
particle(s)
3 Positron-Emitting Radiopharmaceuticals 59
Table 3.3 Radionuclides “of interest” for PET other than 11C, 13N, 15O, and 18F
Theoretical specific Principal emissions (keV)

Radionuclide Half-life β decay mode radioactivity (GBq/μmol) Emax β+ EC main peaks Source
Rubidium-82 1.3 min β+ 96% 5.4 × 106 3350 776 Generator
2580 82
Sr/82Rb
Copper-62 9.8 min β+ 98% 7.1 × 105 2930 1172 Generator
62
Zn/62Cu
Technetium-94m 52 min β+ 70% 1.3 × 105 2500 871 Cyclotron
94
Mo(p,n)94mTc
Gallium-68 68 min β+ 89% 1.0 × 105 1900 1077 Generator
68
Ge/68Ga
Iodine-120 1.4 h β+ 56% 8.3 × 104 4000 1523 Cyclotron
640 120
Te(p,n)120I
Copper-61 3.3 h β+ 62% 3.5 × 104 1150 940 Cyclotron
1200 560 61
Ni(p,n)61Cu
Copper-64 12.7 h β+ 18% 9.1 × 103 657 1346 Cyclotron
β− 37% 64
Ni(p,n)64Cu
Yttrium-86 14.7 h β+ 34% 7.9 × 103 1220 1077 Cyclotron
1578 628–443 86
Sr(p,n) 86Y
Zirconium-89 78.4 h β+ 23% 1.5 × 103 902 909 Cyclotron
89
Y(p,n)89Zr
Iodine-124 4.2 d β+ 23% 1.2 × 103 1532 603 Cyclotron
2135 1691 124
Te(p,n)124I
β+, positron; EC, electron capture; d, days; h, hours; min, minutes
Ga opened the way for PET generator products

68
By definition, radionuclides maintain the chemical fea-
( Ge/68Ga generators). In addition to other properties of
68
ture of the element they belong to. Therefore, while the
radionuclides used for PET, Table 3.1 reports for each radiochemistry of C, O, N, and F is based on the formation of
nuclide their theoretical and practical specific activity (SA), covalent bonds, alternative bonds may be considered for
i.e., the activity associated to the number of atoms or mole- other elements. For instance, metals can form nonstoichio-
cules of interest; SA is expressed as Bq/mol, more com- metric compounds called coordination complexes, in which
monly as GBq/μmol for practical reasons. SA is primarily they constitute the core of the complex and stably bind suit-
related to the isotopic dilution of the radionuclide by iso- able chelating groups (ligands) in their coordination sphere.
topes of the same element, and it is influenced by chemical These coordinate covalent bonds are the basis for technetium
factors (e.g., abundance, natural occurrence, and distribution chemistry as well as for β+-emitting metals, such as 68Ga,
of other isotopes) associated both with radionuclide produc- 89
Zr, and 64Cu. In order to form stable coordination com-
tion and with their handling. SA may refer to a radiolabeled plexes, a strict relationship must exist between the element
precursor as well as to a final product, and it declines with (ionic radius, electronic configuration) and the ligand(s)
time elapsed after production—due to radionuclide decay. (size of the molecule, number and nature of the coordinating
Any radiolabeled product used in vivo should have a spe- groups in the molecule).
cific activity suitable to the intended scope, i.e., consistent
with the target concentration in vivo; due consideration
should also be paid to toxicity or biological activity of the Key Learning Points
radiolabeled compound, especially concerning high-potency • Positron-emitting radionuclides are employed for
agonists or antagonists. diagnostic imaging applications based on the
A few radionuclides can be employed clinically in their emission of two 511 keV monochromatic photons
elemental form, such as sodium [18F]fluoride, sodium [124I] following the positron-negatron annihilation
iodide, or [15O]oxygen. For most applications they are bound event.
to a molecular frame that confers to the radiopharmaceutical • Cyclotrons are used to produce chemical forms
the intended function in vivo. The chemical form of the containing positron-emitting radionuclides.
radionuclide and its features as a chemical element influence • Radiolabeled chemical forms (radioactive precur-
the chemical reactions through which this “precursor” is sors) can be used to prepare more complex radiola-
incorporated into any given radiopharmaceutical. For beled molecules.
instance, 18F can be produced either as fluoride salt ([18F] • Specific activity indicates the radioactivity per
NaF) or as elemental fluorine ([18F]F2). These two chemical number of atoms or molecules of the element or
forms, or precursors, have completely different chemistries radiolabeled compound under consideration.
(see further below in this chapter).
3.1.2 Substrate Characteristics • To select fast reactions and introduce the short-lived
and Radiolabeling Strategies radionuclide as late as possible in the synthetic pathway
(possibly as the last step), in order to limit loss of activity
The clinical development of PET scanners began in earnest in due to physical decay.
1976 with the demonstration that the glucose analog 2-[18F] • To design the substrate (cold precursor) in such a way so
fluoro-2-deoxy-d-glucose ([18F]FDG) was safe and effective as to promote selectivity and region-specificity (i.e.,
for imaging glucose utilization in the human brain. [18F]FDG directing the reaction toward the desired product and at
was conceived as an in vivo extension of Sokoloff’s inspiring the same time avoiding/limiting side products).
experiments on measurement of glucose metabolic rate using • To avoid the presence of interfering/competing chemical
2-deoxy-d-[14C]glucose and tissue autoradiography. groups that may inhibit or affect the radiolabeling reac-
The introduction of the 18F atom in position 2 of the glu- tion. Usually, protective groups are used to eliminate
cose molecule, replacing the naturally occurring hydroxyl interferences and to direct the radionuclide toward a spe-
group, is the key point for the synthesis of [18F]FDG (Fig. 3.1). cific target portion in the molecule.
To achieve this a synthesis pathway was designed, which uti-
lized a labeling-prone substrate (or “cold precursor”). Some labeling approaches have become “classic solu-
The design of a radiolabeling strategy (Fig. 3.2) goes tions,” such as the introduction of suitable “leaving groups” in
beyond the chemistry issues and must consider the intended substrates conceived for fluorine nucleophilic substitutions or
use of the tracer. Fundamental requirements in the radio- the use of alkylating agents able to carry 11C or 18F atoms.
chemistry process for the synthesis of PET radiotracers are: The introduction of “foreign” elements, groups, and
molecular fragments into a biologically active molecule
may affect its biological behavior and acceptance. [18F]
OH OH FDG was a clear example of how substituting a fluoride
atom for a hydroxyl group in the glucose molecule trans-
O O formed a fuel molecule into a cell-trapped substrate. This is
OH OH OH OH still true for many other fluorinated analogs of bioactive
HO HO molecules.
18 The quest for new PET radiotracers is continuously grow-
OH F
ing and extending to many different substrates, in particular
D-glucose 2-[18F]fluoro-2-deoxy-
proteinaceous substrates. Therefore, molecular design must
D-glucose
consider conformational analysis of ad hoc functionalized
Fig. 3.1 A priori molecular design in radiotracer development: 2-[18F] peptides, proteins, and nanostructured and composite materi-
fluoro-2-deoxy-d-glucose ([18F]FDG) als prior to start a radiolabeling project.
Cyclotron Process Chemical “Cold”

Nuclear Available Available hot
energy & methodology reactions precursors
reactions radionuclides precursors
particles
Labeling
strategy
Expected β+ energy &

In vivo Specific Biochemical
Image other Dosimetry Half-life
intended use activity pathway
quality emissions
Fig. 3.2 Key points to be analyzed during the design of a radiolabeling strategy
Fig. 3.3 Chelating cores

used with positron-emitting R
radiometals. One of the CO2H
N N HO2C N N CO2H
groups external to the ring
may be used as a linker to a
protein or peptide
HO2C N N
HO2C N N
CO2H
R
R=H TE2A DOTA

R = CH2CO2H TETA
R CO2H
HO2C N N CO2H
N
HO2C N N CO2H HO2C N N

CONHR
R=H NOTA DOTA – peptide linking

R = (CH2)2CO2H NODAGA R = Tyr3-octreotide DOTATOC
In general, complexes or chelates of positron-emitting PET radiopharmaceuticals are most frequently administered
radiometals cannot be easily bound directly to a biologically intravenously, while less frequently performed procedures
active molecule. Therefore, introduction of a suitable chelat- are based on inhalation of radioactive gases, such as [11C]
ing core in the molecular is required for labeling a biologi- CO, [11C]CO2, [15O]CO, and [15O]O2.
cally active substrate with a radiometal (Fig. 3.3). The blood is both the vehicle distributing the radiophar-
These bridging groups, usually called prosthetic groups, maceutical throughout the body and the first biochemical
are “bifunctional,” i.e., they have chemical affinity for the environment the radiotracer is exposed to. The molecular/
substrate on one side and for the radionuclide on the other biochemical characteristics of the radiopharmaceutical
side. They are also used for nonmetallic radionuclides when determine its residence in circulating blood as well as its
chemistry hurdles have to be overcome between radioactive metabolic fate (Fig. 3.4). Interactions with the blood compo-
precursor chemistry and substrate. nents (proteins and blood cells) must also be considered to
ensure that the radiopharmaceutical is not degraded or trans-
formed. The vascular phase is important in data analysis, as
the delivery of the tracer (input function) is a key parameter
Key Learning Points
in data analysis.
• Fast reactions and ad hoc precursors are required to
Molecular size, polarity, and the presence of active
prepare PET radiotracers.
permeation mechanisms affect the ability of a radiophar-
• The nature and position of the radionuclide (label)
maceutical to cross the vascular endothelium and diffuse
may influence the biochemical acceptance and met-
into the interstitial space. [18F]FDG enters the brain by a
abolic fate of the radiolabeled tracer.
facilitated diffusion process mediated by glucose trans-
• Bifunctional reagents may help in achieving bind-
porter-1 (GLUT-1). Many organs have highly specialized
ing of the radiolabel to the biologically active
endothelia (such as the blood-brain barrier); therefore,
vector.
each tracer must comply with specific requirements that
should be considered a priori when designing the
molecule.
3.1.3 Radiopharmaceuticals and Biological Classic parameters commonly used in drug design (LogP,
Systems LogS, PSA, etc.) are considered during the design phase of
the tracer as necessary (although not always sufficient) char-
As with single-photon radiopharmaceuticals (see Chap. 2), acteristics (Table 3.4). PET tracers generally undergo spe-
there are multiple routes of administration, including oral, cific molecular interactions to reach a specific target tissue/
intravenous, intra-arterial, subcutaneous, and inhalation. organ. When quantifying the radioactivity distribution, it is
Fig. 3.4 Diagrammatic representation

of the molecular characteristics of the
tracer that influence its kinetic and Drug Drug
metabolic fate in vivo absorption injection
Circulating
drug
elimination
Urine
Distribution Other
elimination
routes
Organs
• Trapped metabolite
Metabolism or irreversible binding
or binding • Reversible metabolite
or reversible binding
• Other metabolites
Tissues
Table 3.4 Molecular characteristics and early tests for their assessment
Molecular properties Characteristics Description
Intrinsic molecular MW Molecular size/weight MW influences diffusion coefficient in physiological membranes
properties clog P Lipophilicitya Calculatedb log P
log S Solubility Predicted solubility (calculated from molecular parameters)
PSA Polar surface area Metric (Å2) used to predict the ability of a drug to permeate cellular barriers
Ha H-bond acceptors Number of groups that are H-bond acceptor
Hd H-bond donors Number of groups that are H-bond donor
Experimental properties log P Lipophilicity Logarithm of the molecule partition coefficient between an organic
(physicochemical) solventc and water (neutral species)
Δ log P Hydrogen bonding (N-octanol/water log P)-(hexane/water log P)
log DpH Lipophilicity/ logarithm of the distribution coefficient between an organic solventc and
ionization a buffer (usually pH = 7.4)
S Solubility Speciation done in water and/or biological fluids
pKa Acid-base Negative logarithm of the ionization constant in water
CHI Chromatographic Parameter derived from analytical tests (HPLC) on the candidate
hydrophobicity index substance
Experimental properties Plasma Metabolic stability Plasma and blood contain hydrolases, in particular proteases and
(in vitro biological) stability esterases, with hydrolytic activity on ester and/or amide bonds
Blood
stability
F% Free drug fraction Binding to plasma protein may deeply influence (bio)availability of the
drug at the target
Hepatocytes Biotransformation Hepatocytes contain all drug-metabolizing enzymes and closely mimic
Liver metabolic stability the in vivo situation. Liver microsomal preparation contains all CYP
microsomes isozymes’ other metabolic enzymes, such as FMO and UGT Oxidative
and conjugation metabolism
EC50/IC50 Binding affinity EC50 = concentration of a drug that gives half-maximal response
IC50 = concentration of an inhibitor where the response (or binding) is
reduced by half
Caco-2 Biomembrane Cell model, derived from cells of human colon adenocarcinoma,
permeability expressing efflux transporters (and variable influx transporters)
MDCK Biomembrane Madin-Darby canine kidney cell model. Polarized cells with low
permeability expression of ABC transporters
a
Lipophilicity or hydrophobicity affects drug absorption, bioavailability, hydrophobic drug interactions
b
Calculation based on molecular structure and chemical groups present in the molecule
c
Octanol is typically used; other solvents may be used to explore particular hydrogen-bonding (neutral/donor/acceptor, e.g., hexane/chloroform/
propylen glycol dipelargonate) characteristics of the molecule
important to distinguish the chemical form carrying the ing to the expected pathway. An example of the first type of
radionuclide. Transformations that may occur from blood to substrate is [11C]-l-deprenyl-D2, used to investigate astrocy-
final excretion must be known. This is mandatory when tosis in neurodegenerative diseases. This ligand binds selec-
employing a modeling approach for quantitative estimates of tively and irreversibly to monoamine oxidase-B (MAO-B) in
the regional concentration of a selected radiolabeled species. the brain and acts as a suicide inhibitor through the forma-
Blood sampling and radiochemical analysis may be neces- tion of a covalent bond with the enzyme moiety, thus allow-
sary to identify metabolites derived from the original ing a regional mapping of enzyme density sites.
radiopharmaceutical. [18F]FDG is a paradigmatic example of the second type of
On the other hand, in vivo biotransformation per se can be substrate/enzyme interaction. [18F]FDG is a substrate of the
the target of the investigation with a PET tracer. For instance, enzyme hexokinase that transforms it into the phosphory-
tissue ischemia leads to alteration of the local redox capacity lated product, [18F]FDG-6-P. In turn, [18F]FDG-6-P, which is
and pH; certain PET tracers may respond to these alterations not fit to be a substrate of the enzyme glucose-6-phosphate
by changing their initial chemical nature and thus causing isomerase, remains trapped into the cell. Indeed, the phos-
local accumulation of radioactivity. Hypoxic myocardium or phorylation rate and the transport/extraction are not the same
hypoxic tumors are the typical targets for this imaging for [18F]FDG and glucose, and adequate corrections must be
approach. Tracers that accumulate in hypoxic tissues are employed to measure local glucose metabolic rates using
based on redox-sensitive elements (such as copper in PET-derived data on [18F]FDG extraction rates. To this pur-
67
Cu-PTSM) or on redox-sensitive molecular groups (such as pose, a “lumped constant” has been calculated from head-to-
the 2-nitroimidazole moiety of 18F-FMISO). Tracers whose head experiments, e.g., comparing [18F]FDG with [3H]
stability is pH-dependent, such as 68Ga-citrate or pHLIPs glucose or [14C]glucose (biochemically identical to native
(pH (low) insertion peptides), are labeled with either 18F or glucose). In this regard, it should be considered that the
68
Ga. Nevertheless, the mechanism of local accumulation of lumped constant may vary among different tissues, among
both redox- and pH-sensitive tracers is not fully elucidated. different regions of the brain, and also within the same tis-
The majority of PET tracers are designed to either (1) fit sue/organ during disease versus physiologic conditions. The
a specific binding pocket of a target molecule or (2) enter a example of [18F]FDG as a metabolically trapped substrate
biochemical pathway. Therefore, molecular structure and has opened the way to the development of other PET tracers
impact of the labeling moiety (radionuclide and prosthetic (Fig. 3.5).
group if present) must be carefully evaluated for designing An additional target of PET tracers is the assessment of
suitable tracers. molecular transport and pool measurements. Cellular
Receptor and enzymes, the target of several PET ligands/ homeostasis and function are largely based on mechanisms
substrates, constitute the most typical applications pursued regulating the transit of molecules across membranes and
to exploit the selectivity and specificity of the lock-and-key their intracellular concentration. Solute carrier proteins, ion
mechanism. The receptor ligand (or the enzyme substrate) channel-linked receptors, G-protein-linked receptors, and
fits a specific binding site and, doing so, determines the bio- enzyme-linked receptors have been exploited to develop
chemical response. A wide variety of receptor systems have PET tracers for molecular imaging. For example, many
been assessed with PET tracers (Table 3.5). transport systems for amino acids (AAs) (e.g., system L,
PET ligands can facilitate the early development phases system A/N, etc.) are enhanced in neoplastic cells. [11C]
of a drug compound to evaluate pharmacokinetics in humans. Methyl-l-methionine has been used to image-enhanced AA
Localization of these radiolabeled agents demonstrates the transport and protein metabolism in brain tumors, thus over-
proof-of-principle, saving time in the development of new coming the limitations of the [18F]FDG PET scan due to the
drugs. The high-specific radioactivity of PET tracers plays a high background associated with the physiologically high
key role in this application, as they can be administered and uptake of [18F]FDG in the brain. The development of
traced back even at the low concentrations of ligands at their 18
F-labeled AAs, with improved logistic and scan time, has
signaling sites. In the specific case of receptor ligands, considerably expanded the clinical applications of this
reversibility of binding and affinity of the PET ligand with approach, targeting different tumors via the associated AA
the target must be considered, to make sure the tracer will transport systems.
reflect the true behavior of the receptor under physiological Transporters also regulate the efflux of xenobiotics and
conditions and the intended use (e.g., for mapping the total toxic substances from the intracellular to the extracellular
receptor population, mapping receptor modulation, etc.). space. The ATP-binding cassette (ABC) family, and particu-
There are two main strategies to assess an enzymatic pro- larly glycoprotein-P (PgP) in its isoform I (MDR1/ABC1 or
cess: (1) suicide substrate, thereby the tracer interacts with multidrug resistance pump), can be targeted with PET trac-
the binding site in an irreversible way, and (2) “normal” sub- ers for selecting patients for chemotherapy and for treatment
strates, thereby the tracer interacts with the enzyme accord- adjustment. PET tracers for either PgP substrates (e.g., [11C]
Table 3.5 Radiotracers for neurotransporter and receptor systems

Radiotracer Traced system Clinical applicationa
18
F-FDOPA Dopamine synthesis Dopamine Parkinson’s disease (DA)
Insulinomas (DA)
18
F-fluoromethyltyrosine Brain tumors (DA, RT)
(18F-FMT)
[11C]SCH23390 D1 receptor antagonist Parkinsonian syndromes (RT) Schizophrenia (RT)
[11C]NNC-112 CNS drug development
[11C]N-methylspiperone D2/3 and 5HT2A
receptor densities
[11C]raclopride D2/D3 receptors
(antagonist)
[11C]-(+)-PHNO D2/D3 receptors
(agonist)
18
F-fallypride Extrapyramidal D2 Psychosis in AD (RT)
receptor (reversible) Impulse control disorders in PD
CNS drug development
[11C]WIN-35428 Dopamine transporter Narcolepsy (RT)
[11C]cocaine Drug abuse (RT)
18
F-FECNT Parkinson’s disease (DA)
18
F-FP-β-CIT Psychiatric symptoms in PD (RT)
124
I-nor-β-CIT Dementia (Lewy bodies) (RT)
[11C]hydroxytriptan Serotonin synthesis Serotonin Neuroendocrine tumors (DA)
([11C]5-HTP) precursor Dysphoria and social anxiety disorders (RT)
[11C]WAY-100635 5HT1A receptor Depression
antagonist Functional recovery poststroke (RT)
18
F-altanserin 5HT2A receptor
antagonist
18
F-setoperone 5HT2A receptor
antagonist
4-18F-ADAM Serotonin transporter Behavioral disorders (RT)
[11C]AFM Depression (RT)
[11C]MADAM CNS drug development
[11C]dihydrotetrabenazine Vesicular monoamine Monoamine Depression (CA, RT)
([11C]DTBZ) transporter Therapy monitoring (RT)
[11C]l-deprenyl-D2 Irreversible inhibitor of
pF-[18F]deprenyl MAO-B
[11C]chlogyline Irreversible inhibitor of
MAO-A
[11C]carfentanil μ-OR agonist Opioid receptor (OR) Schizophrenia (RT)
[11C]LY-2795050 κ-OR antagonist Dependencies (RT)
18
F-(-)cyclofoxy μ/κ-OR antagonist Expectation and reward brain systems (RT)
Mood transitions (RT)
[11C]methylnaltrindole δ-OR antagonist
[11C] or 18F-diprenorphine μ,κ,δ-OR antagonist
18
F-FE-PEO OR full agonist
[11C]buprenorphine OR mixed agonist/
antagonist
18
F-FPEB mGluR subtype 5 Metabotropic glutamate Alcohol addiction (RT)
[11C]ABP688 (mGluR5) receptor (mGluR) Focal cortical dysplasia (RT)
[11C]ITMM mGluR subtype 1 Neuropsychiatric disorders (schizophrenia, bipolar
18
F-FIMX (mGluR1) disorder, addiction) (RT)
Principal use: DA, clinical use as diagnostic agent; RT, clinical research tool; AD, Alzheimer’s disease; PD, Parkinson’s disease
a
daunorubicin) or PgP inhibitors (e.g., [11C]verapamil) have in response to brain injuries and inflammation due to microg-
been tested to assess MDR expression and quantification in lia activation; TSPO substrates have been radiolabeled for
patients with malignancies. evaluating neuroinflammation as well as some cancers and
The 18 kDa translocator protein (TSPO) is minimally peripheral inflammation. Examples of the tracers depicting
expressed in normal healthy glial cells but is overexpressed molecular transport are summarized in Table 3.6.
Extracellular pool Intracellular pool Metabolic pool
HK
Glucose Glucose Glucose-6P
Pyruvate
G6Pase
HK
[18F]FDG H2O+ CO2
[18F]FDG [18F]FDG
-6P
G6Pase
Thymidine-diP
TK1 Thymidine-triP
Thymidine-
Thymidine Thymidine
monoP
TYPase
DNA
TK1
18F-FLT-
18 18
F-FLT F-FLT monoP
18
TYPase F-FLT-triP
18
F-FLT-diP
LC-FACS
CAT
Beta
Palmitate Palmitate
oxidation
?
LCFACS H2O+ CO2
CAT
18
18 18F-FTHA F-FTHA
F-FTHA
metabolite
?
HK = hexokinase TYPase = thymidine phosphatase

G6Pase = glucose-6P phosphatase CAT = carnitine-acyl-CoA transferase
TK1 = thymidine kinase 1 LC-FACS = Long chain fatty acyl-CoA synthetase
Fig. 3.5 Examples of PET tracers that exploit the metabolic trapping principle (FTHA also enters triglyceride synthesis)
Table 3.6 Examples of molecular characteristics and interactions used to investigate biological processes in vivo
Radiotracer Mechanism of uptake Biochemical process Clinical applications
[15O]water Freely diffusible substrate Myocardial blood flow Regional perfusion
and perfusion (cardiology, oncology)
[13N]ammonia Facilitated diffusion; intracellularly trapped as glutamine Myocardial perfusion, BBB
transport
18
F-flurpiridaz NADH-ubiquinone oxidoreductase inhibitor of Myocardial perfusion
mitochondrial complex-1 (MC-1)
82
Rb-chloride Internalized as monovalent cation and substrate of Myocardial perfusion
Na/K-ATPase
[18F]FDG GLUT-facilitated transport. Trapped intracellularly, after Regional metabolism Oncology
hexokinase phosphorylation, with direct relationship to Neurology
glucose metabolic rate Cardiology
Infection
Inflammation
[11C]choline Upregulated choline transport and phosphorylation in Prostate cancer
tumors. Substrate of choline kinase, the phosphorylated Cholinergic systema
[11C]choline is essentially trapped within cells
18
F-FTHA Facilitated transport by monocarboxylate transporter Oxidative metabolism
[11C]palmitate family. Substrates of intracellular acetyl-CoA Nonalcoholic steatohepatitis
[11C]acetate synthetases. FTHA remains intracellularly trapped due and fatty liver disease
to β-oxidation interruption by sulfur-for-methylene
substitution along the chain
(continued)
Table 3.6 (continued)
Radiotracer Mechanism of uptake Biochemical process Clinical applications
18
F-FMISO All are based on the 2-nitroimidazole moiety. The tracer Regional hypoxia Tumor oxygenation in
18
F-FAZA passively diffuses through membranes and then is oncology (pre-radiation
18
F-FETNIM reduced by nitroreductase and forms covalent bonds therapy)
intracellularly under reduced O2 levels
64
Cu/62Cu-ATSM The tracer passively diffuses through membranes; the Tumor oxygenation in
complex undergoes thiol-mediated intracellular oncology
reduction. Cu(I) separates from complex and binds to Myocardial ischemia/
intracellular proteins (Cu pathway) hypoxia
68
Ga-PSMA (and Targeting glutamate carboxypeptidase II or prostate- Upregulated protein Prostate cancer
18
F-PSMA) specific membrane antigen (PSMA) function and expression
[11C]methyl-l- System L (Na+-independent transport of neutral AA) Enhanced amino acid Brain tumors
methionine ([11C]MET) transport
[11C]α-1- Biomarker of epileptogenic
methyltryptophan ([11C] process
AMT)
18
F-fluoroethyltyrosine Prostate cancer
(18F-FET)
18
F-fluciclovine ASCT2 (Na+-dependent exchangers for Prostate cancer
(18F-FACBC) Ala-Ser-Cys-Asp-Gln)
18
F-cis-4-fluoro-l- System A (Na+-dependent transport of neutral AA) and Preclinical: brain
proline B0,+ (Na+-independent transport of cystine and neutral neurodegeneration and
and dibasic AA) radionecrosis
18
F-propyl-l-glutamate System Xc- (cystine/glutamate antiporter) Preclinical: glutathione-
based drug resistance
[11C]PK11195 Candidate ligands to 18 kDa translocator protein Upregulated protein Neuroinflammation
[11C]DPA-713 Function and (Microglia activation)
18
F-GE-180 expression
[11C]verapamil P-glycoprotein inhibitor Multidrug resistance
[11C]loperamide P-glycoprotein substrate
18
F-fluoropaclitaxel
[11C]-6-bromo- Multidrug resistance protein substrate
methylpurine
[11C]dantrolene Breast cancer resistance protein
Most of the tracers described here are still under development. The molecular level of interaction requires due consideration of inter-individual
differences, disease genotype, and phenotypic manifestation
a
High specific radioactivity required
Besides transmembrane transport, it is also important to

investigate what is happening inside the cells, including • Radiolabeled compounds may be transformed (in
activity of specific genes. Reporter genes can be a substrate the blood or tissues/organs) and thus give rise to
(transformed into a product that is metabolically trapped radiolabeled metabolites, which may confound
intracellularly or acting as a suicide substrate) of an enzyme image interpretation.
encoded by the reported gene. Alternatively, a radiolabeled • Radiotracer molecular structures are specifically
ligand acting as the reporter probe can be used to bind to the designed to interact with specific tissue targets (e.g.,
receptor encoded by the reporter gene. The major limitation receptor or enzymatic binding sites) or characteris-
of this approach is immunogenicity associated with the use tics (e.g., pH, redox activity).
of the therapeutic transgene, in particular those based on • Key cellular functions can be explored with radio-
viral and nonhuman sources. tracers (e.g., transmembrane transport, gene or pro-
tein expression).
Key Learning Points

• Bloodstream drives the distribution of radiotracers 3.1.4 Production Flowchart and Related Issues
throughout the body and may influence pharmaco-
kinetic parameters. Successful production of a suitable PET tracer moves
through a number of steps:
• Production (at a cyclotron or a generator) of the radionu-

clide in the necessary amount, suitable chemical form and • Radiotracers are to be considered as medicinal
quality products, and appropriate quality standards must be
• Availability (cost, stability, quality) of the cold precursor(s) maintained during the product’s entire life cycle.
• Reproducibility of the radiolabeling reaction(s)
• Efficient purification procedure
• Formulation for administration and dispensing of the final
product (vials, syringes) 3.1.5 B
asic of Automation and Synthesis
• Effective quality controls (e.g., chemical, radiochemical, Devices
radionuclidic purity, sterility, and bacterial endotoxins)
Automation plays a major role in radiopharmaceutical chem-
Full compliance of the process with the pharmaceutical stan- istry. In the late 1970s, most radiochemistry expertise was
dards required for human administration may require the adop- focused on protein radioiodination and on technetium radio-
tion of challenging solutions to cope with the short time frame chemistry/radiopharmacy. This technology was adapted to
and with the need for adequate shielding of the high-energy start remote processing of cyclotron-produced radiogases
radiation associated with positron emission. Pharmaceutical and, by using flow-through chemistry, the production of the
grade standards must be maintained throughout the process for earliest PET radiolabeling precursors (Table 3.7).
equipment and environment (see Chap. 40 of this book The passage from complicated switching boards, manu-
“Radiopharmacy/radiochemistry for positron-emitting radio- ally controlled by the operator who had to visually control
pharmaceuticals, including Quality Assurance and process vali- metering instruments and devices inside the hot cell via a
dation principles”). lead glass, to sequencers and computers has been driving the
“Closed systems” based on automation of radiochemistry and process. Automation technology progress has been paral-
radiopharmacy procedures successfully address this challenge. leled by miniaturization of the components (valves, switches,
sensors, and metering instruments) and electronics for input/
output interfaces.
Key Learning Points
• PET tracer production must be well planned and 3.1.5.1 Automation of Radiopharmaceutical
optimized in all steps to cope with the radionuclide Synthesis
short half-lives. A key point in the synthesis of a PET tracer is to keep the
whole process under control. The operator must know how
Table 3.7 Earliest radiolabeling precursors used for positron-emitting radiopharmaceuticals

Target material Nuclear reaction Product Conversion method Final product Use of final product
14
N2 (+O2 trace) 14
N(p,α)11C [11C]CO2 Oxidation (CuO/750 °C) [11C]CO2 PET tracera
+ Labeling precursor
[11C]CO Reduction (Zn/400 °C) [11C]CO PET tracerb
Labeling precursor
14
N2 (+1% H2) H[11C]CN In-target production H[11C]CN Labeling precursor
14
N2 (+5% H2) [11C]CH4 Catalytic (NH3 + Pt/1000 °C) H[11C]CN
In-target production [11C]CH4
H218O 18
O(p,n)18F H18F CsCO3/calcination Cs18F
20
Ne (+0.5–1% 19F2) 20
Ne(d,α) 18F [18F]F2 CH3CO2NH3 CH3CO218F
XeF2 [18F]XeF2
14
N2 (+2% O2) 14
N(d,n)15O [15O]O2 In-target production [15O]O2 PET tracerc
Reduction (H2 + Pd/150 °C) [15O]H2O PET tracerd
15
N2 (+1% H2) 15
N(p,n)15O [15O]H2O In-target production
14
N2 (+0.1% CO) 14
N(d,n)15O [15O]CO2 Oxidation (MnO2-CuO) [15O]CO PET tracere
+ Purification (charcoal/soda lime) [15O]CO2 PET tracerf
[15O]CO
a
Clearance rate, gas exchange, blood flow
b
Blood volume, splenic red blood cell mass
c
Oxyidative metabolism; oxygen extraction ratio (in combination with [15O]H2O or [15O]CO2 and [15O]CO)
d
Blood flow/perfusion
e
Blood pool
f
Blood flow/perfusion
the process is progressing, and have the possibility to interact loading reagents and mounting the components) are avoided.
with the system should something go wrong. To achieve this, Cassettes are available as kits in which all reagents and puri-
the level of radioactivity (at start, in-process, and in the final fication supports and ancillary components (e.g., sterilizing
product collect vial) is measured and monitored as a sensi- filters) are provided and certified by the manufacturer.
tive and meaningful indicator of process performance. Many Besides facilitating the quick setup of the system, this is a
additional process parameters, such as reactor temperature, convenient simplification in the phase of suppliers/raw
pressure, pH, etc., may need continuous monitoring to ensure materials validation and acceptance, as key parts of process
a timely and a well-controlled process. validation and radiopharmaceutical quality systems accord-
The fluidics of the system must be suitable to handle the ing to good manufacturing practice. A further advantage is
very small volumes used in radiopharmaceutical chemistry that the cassettes are produced and certified as sterile and
(usually microliter to milliliter) and the extremely low con- pyrogen-free.
centrations (often at picomoles) of high-specific activity PET In general, the cassette kit is specific for the production of
products. Efficient transfer of the product(s), as a solution or a selected radiopharmaceutical and is replaced after the pro-
in a gas phase, from a collect vial on to a reactor, a purifica- cess is completed before starting a new synthesis. This step
tion column, or to the final product vial requires novel is sometimes critical, due to residual activity in fluidics and
devices. components. Specific technical solutions in the design of hot
In current systems the controlling device (electronics, cells (shielded repository), remote-controlled manipulators,
valve actuators, syringe motors) is separated from the fluid- and cassette self-cleaning are often adopted to circumvent
ics, the latter being gathered into a cassette to be loaded onto this drawback.
the synthesizer (Fig. 3.6). Advantages of this approach In cassette-based chemistry, microfluidics is emerging as
include (1) cleaning of fluidics is not necessary because a a new approach, since we deal with high-specific radioactiv-
new cassette is used for each synthesis, (2) mis-connections ity products, which requires small volumes of reagents. Most
are avoided, and (2) errors by the operator (especially in of the microfluidic radiochemistry technology derives from
Fig. 3.6 Modern approach to automatic synthesizers based on cassette-like fluidics (tailored on the radiotracer process) and exploiting a separate,
multipurpose hardware
microelectronics manufacturing developments that have per- point during the radionuclide production process [18F]F2,
mitted miniaturization, at the micron scale. The result is a requires the addition of F2 as a carrier, to allow its efficient
chip-like “microcassette” in which microliter scale processes recovery from the target. Therefore, carrier-added [18F]F2 has
can be performed. a lower specific radioactivity than the no-carrier-added [18F]
Furthermore, microfluidic technology involves higher fluoride. Moreover, the [18F]fluoride targetry is simpler, and
surface-to-volume ratios of flow-chemistry microreactors the nuclear reaction yields higher than those of [18F]F2.
than “normal” vessel radiochemistry. In these microreactors
with tens of micron diameter and length of few centimeters, 3.1.6.1 Electrophilic Fluorinations
reagents come quicker and more efficiently in contact, while [18F]F2 was the first radionuclidic chemical for cyclotron-
temperature (energy) is precisely and efficiently transferred produced 18F. Therefore, the earliest reaction pathways using
to the reactants. Furthermore, flows of reactants can be accu- fluorine-18 for labeling molecules of biological interest were
rately controlled, from laminar to highly turbulent, and used the electrophilic reactions based formally on F+ as the reac-
as an additional condition in the process setup. tive intermediate. In this regard, it should be noted that the
In general, higher yields and faster reaction rates (times- F–F complex has a positive charge; therefore, it is electron-
cale can be changed from minutes to seconds) than common avid (i.e., electrophilic). Since the F–F bond is rather weak,
vessel chemistry are achieved, and it is possible to repeat the elemental fluorine tends to resume its electronegative charge
reaction several times, as one single batch of reagents and by avidly attracting electrons. Pure F2 reacts so explosively
precursors in the milliliter scale is sufficient for many runs. In with many substrates that it needs to be diluted with inert
principle, it should be possible to replace the bulky produc- noble gases and used at low temperature. Even in these con-
tion of a single radiopharmaceutical (to be dispensed to sev- ditions, it exhibits very high chemical reactivity, thus inter-
eral patients) with a “dose-on-demand” approach (selecting acting very quickly and in a rather nonselective manner with
on the spot from a list of available radiopharmaceuticals). the electron-rich portion(s) of the molecule yielding electro-
philic substitutions and additions (Fig. 3.7). Notably, the
reaction of [18F]F2 (diluted to 0.1% in neon gas) with O-tetra-
Key Learning Points acethylglucal as cold precursor was the initial way to synthe-
• Automation and miniaturization are key enabling size [18F]FDG.
technologies in radiopharmaceutical chemistry. In general, F2 reactivity is so strong that it is very difficult
• Remote-controlled radiochemical platforms inside to limit side products/reactions and avoid substrate oxidative
shielded hoods (hot cells) are used to prepare and degradation. A number of “tamed” derivatives of [18F]F2
process the radiopharmaceuticals. have therefore been prepared in the quest for better electro-
• Radiotracer-specific pre-loaded cassettes (hardware philic radiolabeling agents (Table 3.8). These derivatives are
and reagents as a kit) are used on general purpose still used with electron-rich substrates via electrophilic sub-
radiochemical platforms. stitution, in particular demetalation of organometallic
reagents (e.g., organotin and organomercury derivatives).
The most common application of electrophilic 18F-fluorination
is the synthesis of 18F-FDOPA, through regioselective elec-
3.1.6 18F-Labeled Radiopharmaceuticals trophilic demetalation of an organotin substrate (although a
nucleophilic alternative is being developed).
The majority of PET radiopharmaceuticals employed in clin-
ical practice are based on 18F. The short positron range in 3.1.6.2 Nucleophilic Fluorinations
tissues results in better image quality than with other PET The 18O(p,n)18F nuclear reaction is well known. Nevertheless,
radionuclides, and its physical half-life allows not only more early production strategies produced 18F-fluoride ion aque-
complex synthetic strategies but also the delivery of radioac- ous solutions (in which the nucleophilic reactivity is
tive products to off-site satellite PET centers. quenched), limiting initial diffusion of nucleophilic fluorina-
The stability of the C–F bond prevents fluorine leakage tion in PET radiochemistry laboratories. The development of
from the labeled molecule, and the small ionic radius pro- cryptands and phase-transfer reagents, able to bind ionic spe-
duces minor molecular alteration, much less pronounced cies in the aqueous phase and transfer them into an organic
than those produced by other halogens such as iodine. phase (the preferred medium for most of the organic reac-
Fluorine-18 can be produced as molecular fluorine ([18F] tions), paved the way to a revolution in radiofluorination.
F2) and aqueous solution of fluoride salts (e.g., [18F]KF). The relevance of this approach emerged with the synthe-
These two molecular forms usually originate from different sis of [18F]FDG, by a nucleophilic route, based on the dis-
nuclear reactions and have completely different chemistry. placement of a triflate leaving group by an “activated”
The extremely high reactivity of molecular fluorine, at some nucleophilic fluoride. Activation was obtained with a crown
Fig. 3.7 General principles X F

of electrophilic substitution
δ-
(upper row) and electrophilic
Z F
addition (lower row) +
δ δ-
+ F Y + XY
F
+
δ δ- F
δ- + F Y
Y
Y
Y = F; CH3C(O)O-; CI3O-
X = -H, -Sn(butyl)3,-Si(methyl)3, -HgOC(O)CF3
Z = electron-donating substituents
Table 3.8 Direct and derived radiolabeling agents for electrophilic fluorination (examples)
Cyclotron-produced
precursor Conversion New radiolabeling reagent Applications
[18F]F2 No conversion (direct reaction) n.a. [18F]FDG (electrophilic addition),
18
F-FDOPA (electrophilic demetallation),
4-18F-fluoromethyltyrosine
(fluorodemercuriation),
2-18F-fluoro-l-tyrosine
(fluorodestannylation)
[18F]F2 (gas) on solid ammonium CH3CO218F [18F]FDG
acetate 5-[18F]fluorouracil
4-18F-fluoroantipyrine
18
F-RGD-peptides
[18F]F2 on Xe at high temperature [18F]XeF2 [18F]FDG
[18F]F2 on 1,4-disubstituted [18F]selectfluor bis(triflate) Preparation of fluoroaromatics and
1,4-diazabicyclo[2.2.2]octane difluoroalkylarenes
[18F]F2 on N-substituted pyridines [18F]N-fluoro-pyridine (or Fluorination of electron-rich “cold
pyridinium) substrates” (organometallic)
[18F]F2 on sulfonamides 18
F-labeled
N-fluoro-alkylsulfonamides
18
F-labeled
N-fluoro-benzenesulfonamides
[18F]F2 on KClO3 [18F]FClO3 Preparation of 2-deoxy-2-fluoro-d-hexoses
by fluorination in water
H18F (anhydrous) Exchange labeling in freon [18F]diethylaminosulfur Conversion of alcohols into fluorides
trifluoride ([18F]DAST)
[18F]fluoride (Post-target method) microwave [18F]F2 High specific activity molecular [18F]fluorine
discharge on [18F]methyl fluoride/F2
[18F]fluoride/crown ether complex in F-labeled Pd(IV) fluoride/
18
Fluorination of arene moiety
solution aryl complex
Arenes = substituted benzene rings (aromatic rings)
amino polyether (Kryptofix® or K2.2.2) having the ability to new method was faster (30 min versus 120 min) and more
complex the potassium, used as counterion to the fluoride efficient than the older electrophilic route, in which 50% of
and thus leaving the fluoride “naked” and more prone to the activity was lost due to the reaction stoichiometry that
reaction. At the same time, the bulky organic complex included the separation of two fluorinated isomers (Fig. 3.8).
between the crown ether and the potassium had enough Thus, the use of activated 18F-fluoride and nucleophilic
“organic” character to carry along the fluoride in organic substitution (Fig. 3.9) has quickly become a common syn-
phase ([18F]KF/K2.2.2). The synthesis of [18F]FDG with this thesis strategy to prepare a variety of new tracers. The cold
Fig. 3.8 (a) The original a OAc OAc OH

[18F]FDG synthesis based on O
the electrophilic route. (b) O OAc 18F 18F
O
The improved version [18F]-F2 (g) AcO 1. hydrolysis
exploiting the nucleophilic OAc OH OH

Freon-11 OAc 2. purification
reaction with fluoride AcO HO
O
OAc 18F
18F
AcO
18F
b OAc OH
OAc
O <K+18F-> K[2.2.2] O
1. hydrolysis
CH3CN O
OAc OTf OAc OAc OAc OH OH
AcO 2. purification
AcO 18F HO
18F
Yield Synthesis time Specific activity

method a) 8-10% 120 min carrier-added
method b) >50% 30 min no carrier-added
Fig. 3.9 General principle of X1 X1

a reaction proceeding via
nucleophilic substitution Breaking bond
Forming bond
Nu- C L Nu C L–
X2 X3 X2 X3
Nu = nucleophile Xi = substituents L = leaving group
precursor to be labeled could be designed, often adopting Classic biologically active, organic molecules (e.g.,
complex chemistry to make radiolabeling straightforward, in Lipinski-like) are not the sole targets for radiofluorination.
general completed in no more than two steps. Neither [18F]F2 nor activated 18F-fluoride have favorable
chances to directly interact with these substrates; however,
3.1.6.3 Prosthetic Groups and Large Molecule these molecules can be radiolabeled with ad hoc synthons
Radiofluorination and ad hoc “prosthetic” chemical groups (Fig. 3.10).
The efficiency of the nucleophilic radiofluorination triggered Once the prosthetic group is introduced in the molecular
a generation of “second-line” reagents. These are radiola- frame by direct functionalization of the peptide final product
beled molecules exhibiting a peculiar reactivity, different or during its synthesis, radiolabeling can be performed keep-
from fluorine/fluoride, that have the function to carry on ing in mind that with any biological molecules (large mole-
board the radiolabel and are used as a radiolabeled reagent cules in particular) reaction conditions may play a key role in
(or synthon) in the synthesis with a cold precursor (Table 3.9). maintaining the integrity of the biological activity. For exam-
Usually the entire labeled molecule is incorporated into the ple, the use of organic solvents and/or excessive tempera-
final product. A classic example is the 18F-fluoromethylation tures may alter the biological properties of the peptide. A
of a nor-methyl substrate. In practice, with these reagents, new strategy potentially circumventing this limitation
radiolabeling with fluorine-18 is limited not only to the for- encompasses the use of orthogonal chemistry. This approach
mation of a C–F bond but can take advantage of a much exploits chemistries and groups that are not encountered in
wider chemistry. biomolecules or involved in biological processes and there-
Table 3.9 Fluorinated reagents and prosthetic groups

Cyclotron- Fluoride
produced activation Typical labeling target(s) in the
precursor strategy Synthon substrate End-product
[18F]fluoride K2CO3/K2.2.2 [18F]fluorobromomethane Amines, thiols, alcoholsa Fluoromethylation
(or TBAH) [18F]fluoroethyltosylate Fluoro-ethylation
p-[18F]fluoroiodobenzene Alkynes, aromatic rings Pd-mediated cross-coupling
p-[18F] reactions
fluoroaryl-tributylstannane
p-[18F]fluoroarylboronic acids
p-[18F]fluorobenzaldehyde N-hydroxylamine or Hydrazine Oximesc or hydrazonesc
2-[18F]fluoro-2-deoxyglucoseb derivatives
4-([18F]fluoromethyl)phenyl Aminesa (e.g., terminal amino Amide bond formation with
isothiacyanate groups and ε-NH2 of lysines)a protein, peptides, and antibodies
4-[18F]fluorobenzoic acid
N-succinimidyl 4-([18F]
fluoromethyl) benzoate
N-(4-[18F]fluorophenyl) Thiolsa (e.g., cysteine) Protein, peptides, and antibodies
maleimide
[18F]fluorophenacyl bromide
ω-[18F]fluoroalkyne Azide-containing substratesd Triazolec formation
ω-[18F]fluoroazides Alkyne-containing substratesd
Synthon = molecular fragment that will be incorporated into the final compound
a
Off-target multiple labeling is possible. Protective groups may be needed
b
As open-chain anomer (aldehyde)
c
Bio-orthogonal bond. This bond is not degraded in vivo by common biochemical processes
d
“Click chemistry”: cycloaddition azide-alkyne. May be mediated by catalysts (e.g., copper) or facilitated by intramolecular strains in a substrate
(e.g., cyclooctyne)
Fig. 3.10 Radiofluorination
of protein substrates SH
Protein NH2 + 18F –
N3
18F
O
18F
N N Protein
O
O H O
SH N N
H
O O
F
18F F 18F
O
O
F H
F
Protein NH2 N Protein
O
18F Protein
O
H n 18F
N O
N3 n
N=N
fore can be applied without interference or competition by 3.1.7 11C-Labeled Radiopharmaceuticals

existing groups and withstand biodegradation.
Carbon-11 is regarded as an ideal radionuclide for labeling
molecules of biological interest but with the major disadvan-
Key Learning Points tage of its 20.3-min physical half-life. From the chemical
• The majority of PET radiopharmaceuticals are point of view, substitution of native 12C with the positron
based on 18F. emitter 11C leaves the biological characteristics unchanged;
• Two alternatives chemical methods are available to nevertheless, the short time window for PET imaging per-
prepare radiofluorinated compounds: nucleophilic mits only those applications in which the PET study addresses
route (based on 18F-fluoride as precursor) or electro- a fast biokinetic process. In these conditions, the short physi-
philic route (based on [18F]F2 as precursor). cal half-life may even be an advantage, as repeat tests can be
• Nucleophilic labeling is simpler, more efficient, and performed in the same subject (e.g., test/retest applications).
widely used. The production of carbon-11 in a gas target takes advan-
• Nucleophilic-based radiofluorinations can achieve tage of hot atom chemistry and target gas composition to
better specific activity than the electrophilic direct the product of the nuclear reaction toward different
reaction. chemical forms, mostly [11C]CO/CO2 and [11C]CH4. These
• Depending on the chemistry of the process, the are currently the building blocks that are available to synthe-
radiofluorinated precursor can be converted into size 11C-labeled compounds.
derived reagents having better reactivity with Direct use of [11C]CO2 in carboxylation reactions was the
selected substrates (cold precursors). simplest approach to produce organic compounds, such as
• Selected functional groups (prosthetic groups) may carboxylic acids and their derivatives. However, carbon-11
be introduced into the molecular frame to allow radiochemistry has expanded to a high number of labeling
direct radiolabeling with radioactive fluorine. agents, thus increasing the efficacy and flexibility of
11
C-labeling reactions (Table 3.10). Although the list of
Table 3.10 Basic 11C-synthons and carbon chemistry employed per product categories
Cyclotron-
produced Synthons derived
precursor from precursor Typical use End-product(s)
[11C]CO2 Neata Formation of carboxylates by direct reaction with Grignard reagents and Carboxylic acids
organolithiums; these compounds can be functionalized to Amides
carboxyamides and amines (by reduction of amides) Carbamates
[11C]formaldehyde Electrophilic aromatic substitutions and ring-closure reactions, Mannich- Oxazolidinones
[11C]phosgene like alkylation reactionsb Substituted ureas
Alkyl carbonates
[11C]methyl iodide Alkylations N-, O-,
[ C]CH4
11
[11C]methyl iodide S-methylderivatives;
[11C]methyl triflate
[11C]cyanidric acid Nucleophilic additions and substitutions Amides,
and [11C]cyanate Amines,
Amino acids, Hydantoins,
nucleosides
Ureas
[11C]CO Neatc Carbonylation reactions ketones, -amides, -ureas, -carboxylic acids, and Ketones,
-esters. Selenium-mediated carbonylation; transition metal catalyzed Alcohols,
insertion and reductive elimination of C–C, C–N, and C–O bonds Carboxylic amides,
Aminocarbonylations
a
Specific radioactivity may vary
b
Complementary to alkylations obtained with methyl iodide
c
CO trapping efficiency and solubilization in the reaction media are challenging
Fig. 3.11 (a) Precursors for

labeling with 11C and (b–d) a cyclotron b 11
NH2 11 N CH3
general application in carbon [molecule] CH3OTf [molecule] H
chemistry NHR 11or
11
N CH3
[molecule] CH3I [molecule] R
11 11 11
CO CO2 CH4 SH 11
S CH3
[molecule] [molecule]
11
11 11 11 OH O CH3
COCl2 CH3OH CH3I H11CN [molecule] [molecule]
11 11 11
CH3COCI CH2O CH3Li
c 11
O
CO2
R-CH2 MgX+ (or Li+) R-CH2- 11COH
O O O
d
11 11 11
C C C
11
CO R R’ RO NR’R’’ R2N NR2
11
COCI2 ketones carbamates ureas
11 O O
H CN O
11 11 11
C C C
R H R NR’R’’ R OR’
aldehydes amides esters
11
C-labeled “building-blocks” may appear limited palladium-mediated reactions even if the transition to a pro-
(Fig. 3.11a), the variety of biological molecules that can be duction-sustainable procedure still depends on the palladium
labeled is extremely ample. complex.
3.1.7.1 11
C-Methylation Reactions 3.1.7.2 Emerging 11C-Labeling Reactions
Alkylation reactions proceed quickly and efficiently, with very Carboxylation reactions of organometallic reagents
good selectivity. 11C-Methyl iodide has been an early target in (Fig. 3.11c), currently included under the more general term
carbon-11 radiochemistry; later, this methylation agent has of CO2‐fixation reactions, have constituted the first approach
been joined by the cognate compound 11C-methyl triflate. Both to 11C-labeling, as they made use of the [11C]CO2 directly
compounds can currently rely on very efficient and dedicated coming from the cyclotron target. Strong anionic-like
synthesizers, in which the most crucial point is keeping high reagents, such as Grignard reagents or organolithiums, must
the specific radioactivity and the conversion of the precursor be used to overcome the chemical inertness of CO2. A full
radiogas into the desired methylation agent. Methylation is a spectrum of carboxylic acids and fatty acids (FA) have been
one-step reaction that is performed directly on the cold precur- prepared following this route; in particular, [11C]acetate and
sor (Fig. 3.11b). To this purpose, the molecule is designed in 1-[11C]palmitate are widely used as tracers to assess oxida-
such a way that it carries an alkylation-prone group (including tive metabolism and FA consumption/storage in the heart
either an N, O, or S heteroatom) that is usually the desmethyl and liver. The synthesis is very flexible, and many other FAs
(nor-methyl) analog of the final, target product. (oleate, arachidonate, heptadecanoate, docohexaenoate, ace-
The substrate is always present in much larger concentra- toacetate, etc.) have been labeled in different positions and
tion than the labeling agent, thus increasing both the reaction with different purposes (e.g., neurotransmission in the brain)
rate and reaction selectivity, in practice eliminating the pos- (Fig. 3.12). These reactions require extremely dry conditions
sibility of multiple additions to the same molecule. The reac- and are generally carried out under an inert atmosphere (N2
tion is quite frequently carried out in a liquid phase, but it has or Ar). Furthermore, careful attention must be paid to avoid
also been developed on solid supports and even in microflu- contamination by atmospheric CO2 (which is readily
idic apparatuses. In general, the alkylating agent is prepared absorbed by the very reactive reagents) in order to maintain
and distilled directly into the vessel containing the substrate high and reliable specific radioactivity.
to be labeled in a suitable organic solvent (e.g., DMSO, Direct formation of 11C‐labeled compounds by [11C]CO2‐
DMF, acetone), which has to be tested for in the final prod- fixation has been more recently carried out using strong
uct. More recently, the use of 11C-methyl iodide has been organic bases, as in the synthesis of 11C-labeled carbamates
expanded to the formation of carbon–carbon bonds, by using that, like urea, are stable in vivo.
a p-phenylisothiocyanate-DOTA
route 2
HO2C CO2H HO2C CO2H HO2C CO2H
NCS
N N N N N N
NCS
N N N N N N
HO2C CO2H HO2C CO2H HO2C CO2H
DOTA p-phenylisothiocyanate-DOTA
route 1
b SCN
O DOTA O H H
NH2 N N
peptide peptide C
R R S DOTA
O H H O H H
N N <68Ga> N N
peptide C peptide C
S DOTA R S [DOTA68Ga]
R
Fig. 3.12 Example of a prosthetic group (a) used to label peptides with radiometals (b)
Other radioactive precursors have been used for the prep-

aration of 11C-carbonyl compounds, such as [11C]phosgene Key Learning Points
and [11C]isocyanates, but seldom have found their way in the • Simple 11C-radiolabeled chemical species obtained
preparation of products applied for clinical use. at the cyclotron can be used as building blocks to
The precursor [11C]KCN has been more frequently used prepare more complex molecules.
in nucleophilic substitutions of aliphatic substrates (such as • Most of the 11C-labeled radiotracers with clinical
halogenated compounds and tosylates) to yield 11C-labeled applications are obtained via a methylation reaction
amides, carboxylic acids, and amino acids, such as [11C]glu- (onto O, N, or S heteroatoms in the molecule) of
tamine used for metabolic imaging of tumors and [11C]leu- selected desmethyl precursors.
cine used for tracing amino acid transport. • Other organic chemistry reactions are being devel-
Finally, direct introduction of [11C]CO—easily avail- oped to extend the number of 11C-radiolabeled com-
able from the carbon-11 target at the cyclotron—into sub- pounds (e.g., carboxylation, carbonylation).
strates has also been pursued. The main limitation of this • Short half-life limits the use of 11C-radiotracers to
approach is the very poor solubility of carbon monoxide in-house applications, where clinical investigation
and the difficulties in trapping efficiently the precursor. prevails over routine use.
Recent developments in the rhodium- and palladium-
mediated CO insertion reaction may provide more viable
solutions. 3.1.8 68Ga-Labeled and Radiometal-Based
All the new radiolabeling pathways mentioned above Radiopharmaceuticals
contribute to the development of novel 11C-labeled radio-
pharmaceuticals (Fig. 3.11d), although mostly for research Gallium-68 can be obtained from a generator or also directly
purposes rather than for clinical applications, targeting rele- from cyclotron bombardment. The 68Ge/68Ga generator has
vant enzymatic (e.g., FAAH, LAT1, MAO-A, and -B) and been known since 1964, but drawbacks in generator perfor-
receptor systems (κ-opioid, D3, and 5-HT1A). mance (such as low elution yield, 68Ge breakthrough, and
interference from metallic impurities) have limited its appli- facilitate the insertion of modified amino acids or amino
cation. In aqueous solution, gallium has a 3+ oxidation state acid-like intermediates carrying the chelating group.
and can exist as free hydrated ion only at acidic pH. The Whenever necessary, e.g., when active conformation exists
development of suitable gallium chemistry was not an easy for the molecule, in silico calculations may be performed to
task: mimicking Fe3+, 68Ga3+ produces a very stable complex evaluate any structural perturbation following prosthetic
with transferrin in vivo. Therefore, the synthesis of ligands group insertion.
ensuring the formation of Ga-complexes exhibiting high Commercial availability of more efficient and reliable
thermodynamic and kinetic stability was a necessary step to 68
Ge/68Ga generators has further promoted the use of gal-
overcome the Ga-transferrin complex formation in vivo, lium-68 in nuclear medicine departments (especially those
both directly (larger stability constant) or via ligand exchange without a cyclotron), and many 68Ga-labeled radiopharma-
(higher kinetic stability). ceuticals are currently available for either clinical use or
A crucial step forward was the observation that macrocy- research purposes (Table 3.11).
clic chelators (DOTA, NOTA, etc.) (Fig. 3.3) produce stable The most notable family of novel PET
complexes with metals in the 3+ oxidation state, including 68
Ga-radiopharmaceuticals is a group of ligands that bind
gallium. A major breakthrough was the elaboration of these with strong affinity and avidity to the prostate-specific mem-
macrocycles as prosthetic groups linked to peptides of bio- brane antigen (PSMA), a transmembrane glycoprotein with
logical interest (Fig. 3.12). enzymatic activity (glutamate carboxypeptidase II) that is
highly specific for either normal prostate tissue and (espe-
3.1.8.1 68
Ga-Labeling cially) of prostate carcinoma. This property makes PSMA an
The “lead” compound of 68Ga-radiopharmaceuticals is enticing target for the imaging and therapy of prostate can-
68
Ga-DOTA-TOC (Fig. 3.3). This compound has high affin- cers. Several small molecular weight 68Ga-PSMA ligands
ity for the somatostatin (SST) receptors (particularly subtype have been produced, including 68
Ga-PSMA-11,
2) and is clinically employed for PET imaging of neuroendo- 68
Ga-PSMA-617, and Ga-PSMA-I&T that appear as the
68
crine tumors (NET). 68Ga-DOTA-TOC PET yields much bet- most promising for imaging prostatic cancer. The current
ter imaging results than 111In-DTPA-TOC SPECT, therefore trend is to consider these radiopharmaceuticals not only for
becoming a reliable PET screening agent to select patients their high diagnostic potential in detecting early recurrences
for therapy with 90Y-DOTA-SST analogs. The promising of prostate cancer but also within an integrated scenario of
results obtained in early clinical trials constitute an impor- theranostics thereby their uptake in cancel lesions can pre-
tant impulse toward developing new SST analogs by follow- dict the efficacy of therapy with, e.g., 177Lu-PSMA-ligands.
ing two directions: (1) finding new and better chelators and
(2) developing potent and selective peptide vectors to which 3.1.8.2 Other Radiometals
a suitable chelating prosthetic group can be linked. In par- Success of the new methodologies developed for labeling
ticular, total peptide chemistry (manual or automatic) could bioactive peptides with 68Ga has triggered more attention on
Table 3.11 Ga-based radiotracers obtained via direct or bifunctional labeling

68
Ga-labeling strategy
68
Acronym Short description Target use
Direct labeling Ga-EDTA
68
Ga-ethylenediaminetetraacetic acid
68
BBB integrity
Renal function
68
Ga-Cit 68
Ga-citrate Hypoxia
Inflammation
68
Ga-MAA 68
Ga-human serum albumin Perfusion
macroaggregates
Labeling with 68
Ga-DOTA-TOC 68
Ga-DOTA-Tyr3-octreotide Somatostatin analogs used in neuroendocrine
bifunctional agents 68
Ga-DOTA-TATE 68
Ga-DOTA-Tyr3,Thr8-octreotide tumor detection (NET)
68
Ga-DOTA-NOC 68
Ga-DOTA-NaI-octreotide
68
Ga-NOTA-RGD 68
Ga-NOTA-[arginyl-glycyl-aspartic RGD motif targeting αvβ3 integrins
seq.]
68
Ga-NOTA-GSA 68
Ga-NOTA-glycosyl human serum Liver function
albumin
68
Ga-DOTA-hEGF 68
Ga-Human epidermal GF EGFR expression
68
Ga-DOTA-MG0 68
Ga-DOTA-minigastrin Cholecystokinin-2 gastrin receptor (in NET)
68
Ga-DOTA-BZH3 68
Ga-DOTA-bombesin derivative Gastrin-releasing peptide receptor
68
Ga-NOTA-BBN2 68
Ga-NOTA-bombesin derivative
other radiometals as candidates for labeling both existing Table 3.12 Examples of radiolabeling of biological molecules with
and new molecules, so to expand the range of available trac- radiometals
ers and their use can. Radionuclide Biological vector Biochemical target
The relatively short physical half-life of 68Ga (about Zirconium-89 Anti-PSMA Prostate cancer
68 min) results in a short useful imaging window for radio- Cetuximab EGFR expression
Rituximab CD20 expression
pharmaceuticals labeled with this radionuclide.
Trastuzumab HER2 expression
Unfortunately, there are few radionuclides having an accept-
Nanocolloidal Sentinel lymph node
able half-life and decay mode with reasonable positron emis- albumin
sion abundance. Furthermore, nuclear chemistry constraints Bevacizumab Ovarian cancer
(targetry technology and production protocols) and nuclide- Copper-64 Trastuzumab HER2 expression
specific chemistry play important roles in selecting a radio- DOTA-NT-Cy5.5 Neurotensin receptor
nuclide for labeling compounds of biological interest. DOTA-alendronate Mammary
microcalcifications
Copper-64 (half-life 12.7 h) is considered an acceptable,
DOTA-patritumab HER3/ERBB3 solid tumors
although not optimal, candidate; in fact, its positron energy/
CBTE2A-galectin-3 Breast carcinoma
range is similar to that of 18F, but its positron abundance is peptide
only 19%. 64Cu-ATSM has been used mainly for detecting DOTA-AE105 Urokinase plasminogen
hypoxic areas, such as those present in many tumors or myo- activator receptor
cardial ischemia. On the other hand, the radionuclide per se
in the simple chemical form 64Cu-chloride is emerging not
only because of its potential for imaging in oncology but also low this route. Table 3.12 summarizes some examples of
because of its potential for antitumor therapy due to its emis- radiometal-labeled products that are under development,
sion of β− particles—within the scenario of theragnostics. particularly for oncology applications.
Furthermore, its ability to form stable complexes and its Direct radiolabeling of large molecules (such as antibod-
12.7 h half-life have renewed attention on the possibility to ies or other proteins) is complicated by the variety of groups
transfer the experience gained with 68Ga-labeling to that may be present in such macromolecules and by the high
64
Cu-labeling, for instance, using peptides functionalized risk of altering the active conformation or the active site(s).
with a prosthetic group containing a TETA or TE2A chelat- Usually, exposed functional groups (such as ε-amino groups
ing core. Thus, 64Cu has been chelated to different peptides of lysine or thiol groups of cysteine residues) are targeted by
and proteins, such as the integrin-targeting peptide contain- bifunctional agents with chemical selectivity/affinity for
ing the RGD sequence (or Arg-Gly-Asp) and antibodies these groups. Nevertheless, newer reactions (such as orthog-
(e.g., 64Cu-DOTA-Cetuximab). Theranostics is the common onal chemistry ligation via click chemistry) are being tested
denominator in this rapidly expanding field, thereby imaging to expand the radiolabeling perspectives. An interesting fea-
and therapy using the same antibody construct as the tumor- ture of click chemistry is the possibility to pretreat the pro-
targeting carrier molecule (e.g., an anti-EGFR antibody), tein with the click-prone reagent and to perform the click
labeled with 64Cu for immune-PET and with 177Lu for radio- reaction in aqueous media with the preformed complex, or
immunotherapy, respectively. even in vivo, following a two-step administration.
Zirconium-89 (with physical half-life 78.4 h and 22.7%
abundance β+ emission) is an additional promising radionu-
clide. Its quite long physical half-life allows PET imaging on Key Learning Points
multiple days after administration, thus ensuring a longitudi- • Developments of gallium chemistry and 68Ge/68Ga
nal follow-up of the biodistribution pattern of labeled thera- generator technology have boosted the interest over
peutic antibodies (mAbs) or other compounds. 68
Ga as a clinically significant radionuclide.
The critical issue in the use of 89Zr is to identify good • 68Ga needs good chelating groups, such as macro-
chelators able not only to allow the preparation of cycles containing nitrogen atoms, for acceptable
89
Zr-complexes in good yield but also, and most importantly, in vivo stability.
to ensure that they can withstand the long exposure to the • Macrocycles can be chemically linked to biologi-
in vivo environment so to prevent release of free 89Zr (bone cally active compounds (e.g., peptides) to yield
being the target tissue). This task is not simple due to the 68
Ga-labeled substrates.
characteristics of the coordination core of Zr. While many • 68Ga-DOTA-TOC (octreotide) targeting somatosta-
“classic” coordination cores used for other radiometals tin receptors and 68Ga-PSMA ligands targeting
proved inefficient, the best results have so far been obtained prostate-specific membrane antigen sites are the
using the core of siderophore, i.e., desferrioxamine; most of most common 68Ga-radiotracers in clinical use.
the bifunctional chelators currently under investigation fol-
3.1.9.2 Radiopharmaceuticals Labeled with Very

• Other radiometals (such as 64Cu and 89Zr) are being Short-Lived Radionuclides
assessed as additional radiotracers based on chela- The radionuclides with half-lives shorter than carbon-11
tion chemistry and peptide (protein) radiolabeling. normally require that their production source is next to the
• Cu and 89Zr have longer half-life than 68Ga and
64
PET examination room; therefore, logistical problems are
might be used to trace longer kinetics. frequently encountered, and there is a declining trend regard-
ing their clinical use, especially when reasonable imaging
alternatives become available (e.g., ultrasonography or MRI).
The shorter the half-life, the narrower is the variety of
tracers that may be used as precursors and transformed into
3.1.9 Miscellanea the final radiopharmaceutical; for half-lives below the
10 min half-life of 13N, only flow-through chemistry can be
3.1.9.1 124
I-Labeled Radiopharmaceuticals used. Although some attempts have been undertaken to pro-
Iodine-124 has a physical half-life of 4.2 days and a 23% duce 13N-labeled amino acids, the only clinically viable
abundance β+ emission. Although these features are subopti- 13
N-labeled product is [13N]ammonia, which can be pro-
mal for imaging, 124I can be used for theragnostic applica- duced either in-target or by Devarda’s alloy treatment of
tions. After having acquired all relevant pharmacokinetic irradiated pure water (reduction of the cyclotron-produced
data that may be necessary to build up an accurate treatment nitrite/nitrate aqueous solution). In both cases, processing
plan, the same substrate that has been used for imaging lasts only few minutes and ends in a calibrated syringe of
(labeled with 124I) can then be used at higher activities for the injectable solution. Accurate quantitation of myocardial
treatment (e.g., labeled with 131I or with other suitable radio- blood flow and efficient test/retest measurements with low
nuclide emitting β− particles). background due to the quick decay justify the use of [13N]
In principle, all the know-how developed and consoli- ammonia PET.
dated over the years on other radioisotopes of iodine (123I, The use of radiogases labeled with oxygen-15 (t1/2
I, I) can be quickly transferred on to 124I. Iodine-124 is
125 131
2.03 min, β+ abundance 100%) was adopted in the early
available as sodium iodide, and, as such, it can be used for phases of development PET imaging, particularly for studies
nucleophilic substitutions or, in combination with the use of on cerebral blood flow, perfusion, and oxygen metabolism,
an oxidizing agent, for electrophilic substitutions. Classic during which a combination of different tracers was used,
oxidizing agents commonly used for radioiodination, such as i.e., [15O]water (for measuring flow), [15O]CO (for measur-
chloramine-T or Iodogen®, have been successfully used to ing blood volume), and [15O]O2 (for measuring oxygen
prepare 124I-labeled radiopharmaceuticals. Well-established extraction). Alternative methods have currently largely
prosthetic groups, such as the Bolton-Hunter reagent, and replaced the use of these tracers (which also involved a rele-
newer bifunctional compounds have been used when direct vant radiation exposure burden), and only few research cen-
labeling was not possible due to the lack of radioiodination- ters are still using this approach, in particular to measure
prone moiety (e.g., tyrosine or histidine residues in peptides) regional blood flow (mostly in the brain and/or
or when it was necessary to carefully target the introduction myocardium).
of the radiolabel via the prosthetic group. The ultrashort-lived rubidium-82 (t1/2 76 s, β+ abundance
Small molecules, peptides, and proteins have been 100%) is produced by the 82Sr/82Rb generator, which is com-
labeled and tested—with variable results in terms of clinical mercially available and meets all the requirements in terms
use. Table 3.13 lists the most widely used of efficacy and safety. Given the extremely short half-life of
124
I-radiopharmaceuticals as well as the radiolabeling route the radionuclide (a potassium analog), the entire generator,
adopted. equipped with shielding, pumps, eluting solution, and any
Table 3.13 Biologically active substrates under development for theranostic applications as radioiodinated agents
Radioiodination method Short description Acronym Biochemical target
Oxidative radioiodination Antibody fragment 124
I-mini-Ab F16SIP Treatment of NSCC
5-iodo-2′-deoxyuridine 124
I-IUDR Tumor cell proliferation
Iododestannilation EGFR inhibitor 124
I-ML08 EGFR expression
EGFR kinase activity 124
I-IPQA
Cyclin-dependent kinase 4/6 inhibitors 124
I-labeled CKIA Tumor cell proliferation
Isotopic exchange with iodide 2-nitromisonidazole moiety 124
I-IAZA Tumor hypoxia
124
I-IAZG
Epinephrine analog 124
I-MIBG Adrenergic activity (NET)
other formulating accessories, is connected to the patient 3.2 ection II: Positron-Emitting
S
through an i.v. line, and the eluate is directly administered to Radiopharmaceuticals for Clinical Use
the patient. Although blood flow measurements obtained
with this method are not identical to those obtained with 3.2.1 Sodium 18F-Fluoride
[13N]ammonia, this system has the advantage to allow PET
myocardial blood flow measurements without the need of an The microelement fluorine is a normal constituent of the
on-site cyclotron; furthermore, measurements can be hydroxyapatite crystals forming the mineral matrix of the bone.
repeated under different pathophysiologic conditions (e.g., at Sodium 18F-fluoride is one of the first radiopharmaceuticals
rest and during a pharmacologic stress test) in the same approved by the US Food and Drug Administration (in 1972)
imaging session. for clinical use as a bone-seeking agent. Technical difficulties in
imaging the 511 keV annihilation photons with the instrumenta-
tion available at that time subsequently led to its withdrawal in
1984 from the approval list, and 18F-fluoride was replaced by
Key Learning Points
diphosphonates labeled with 99mTc, emitting γ-rays with opti-
• 124
I is being considered for theragnostic application
mal energy for imaging with standard gamma cameras. The
both as diagnostic/therapeutic use by dose escala-
development of clinical PET scanners has revived interest in the
tion and in “sister imaging” before therapy with
clinical use of sodium 18F-fluoride as a PET bone-scanning
131
I-labeled compounds.
agent with much greater diagnostic performance than gamma
• 124
I is not optimal from the imaging and dosimetric
camera imaging with 99mTc-diphosphonates.
point of view.
Upon its i.v. administration, 18F-fluoride undergoes a high
• 124
I relatively long half-life and the consolidated
and rapid bone uptake which is combined with a fast, biexpo-
knowledge on radioiodination reactions may find
nential blood clearance; these kinetic features result in high
application in protein/peptide radiolabeling.
bone-to-background ratios reached within a short time, so that
• The very short half-life radionuclide 13N has its
imaging can be acquired as early as 1 h (or less) after adminis-
optimal application in PET quantitation of regional
tration. The intensity and extent of 18F-fluoride uptake through-
myocardial blood flow but can be used only
out the skeleton reflects blood flow and bone remodeling.
on-site.
The mechanism of uptake in the bone is based on an ion
• Ultrashort-lived 82Rb (from long-lived 82Sr genera-
exchange process thereby each fluoride ion exchanges for an
tor) can be used in clinical settings as a commercial
hydroxyl ion on the surface of the newly formed hydroxy-
medical device.
apatite crystals, followed by chemisorption into the crystal-
line matrix with prolonged retention (until subsequent
remodeling of the bone). About 20% of injected activity is
excreted through the urinary tract within 2 h after
3.1.10 What Might Be Next administration.
Although the normal biodistribution of 18F-fluoride is
Over 4000 compounds have been reported in the literature rather uniform, the axial skeleton has a higher uptake than
radiolabeled with one of the many existing positron-emitting the appendicular skeleton, and the bone around joints dis-
radionuclides. Yet, the number of imaging agents actually plays higher uptake than the shafts of long bones. Increased
reaching the stage of clinical practice is very limited. 18
F-fluoride uptake can be observed in both sclerotic and
Although the regulatory burden is often a recurrent theme as lytic metastases.
“the” attrition factor in reaching out a more widespread
application of new PET tracers, addressing the multifaceted,
variable disease manifestations and at the same time the Key Learning Points
intrinsic, profound molecular characteristics of the tracers • Sodium 18F-fluoride is a PET bone-seeking
remains a hard challenge. On the other hand, molecular com- radiotracer.
plexity is now seen as an opportunity. In fact, an emerging • It follows the metabolic pathway of fluorine being
trend in radiolabeling is to develop nanostructured materials incorporated into the hydroxyapatite crystals that
and elaborate these complex constructs as multi-probe trac- form the mineral matrix of the bone.
ers, e.g., optical/MRI/PET constructs that could combine the • It can be used to image bone remodeling, such as in
in vitro biochemistry at cellular level with the in vivo both sclerotic and lytic tumor metastases.
findings.
3.2.2 Flow/Perfusion Agents on-site by the 82Sr/82Rb generator (82Sr decaying with a phys-
ical half-life of 25.55 days). The positron range of 82Rb
As mentioned in Section I of this chapter, the simplest PET- (8.6 mm) and its relatively low extraction fraction (about
based approach to image blood flow and tissue perfusion of 65% at each pass) are suboptimal as compared to other PET
various organs is based on the administration of 15O-labeled tracers for myocardial perfusion. In addition, 82Rb extraction
small molecules that undergo virtually unrestricted diffu- can be decreased by severe acidosis, hypoxia, and ischemia.
sion, such as water ([15O]H2O; see here below) or radioactive Thus, uptake of 82Rb in the myocardium varies as a function
gases as [15O]O2 by itself, 15O-labeled carbon monoxide of the combined effect of blood flow, metabolism, and myo-
([15O]CO), and 15O-labeled carbon dioxide ([15O]CO2). Since cardial cell integrity.
the physical half-life of 15O is about 2 min, an important
advantage of using these 15O-labeled agents is the possibility 3.2.2.3 [13N]Ammonia
to perform repeated measurements at short intervals during a Ammonia labeled with 13N (physical half-life 9.96 min) is
single imaging session (e.g., every 15 min). This feature used for quantitative assessment of myocardial perfusion.
enables to investigate the pathophysiology of blood flow/per- The favorable positron range (2.53 mm) and an 80% extrac-
fusion in different conditions, for instance, at rest and under tion fraction at each pass allow acquisition of intermediate-
stress when evaluating myocardial perfusion or during sen- to-high-quality images. The uptake of [13N]ammonia is
sorial deprivation then again during activation tests when nonlinearly correlated with regional myocardial blood flow,
evaluating brain perfusion. as doubling the flow leads to a 70–80% increase in tracer
The very short physical half-life of PET tracers labeled with uptake. Furthermore, inhibition of glutamine synthetase with
15
O or with other cyclotron-produced short-lived radionuclides l-methionine reduces metabolic trapping of [13N]ammonia,
(e.g., 11C) that are not available locally through the use of radio- suggesting that the glutamic acid-glutamine reaction is pri-
nuclide parent/daughter generators constitutes an important marily responsible for ammonia trapping into the myocar-
limitation to a wider utilization of the 15O-labeled PET tracers. dium. [13N]Ammonia circulates in the blood in the ionic
Among the tracers mentioned above, [15O]H2O remains the form [13N]NH4+ that is converted back to [13N]NH3 upon
preferred PET agent for quantitative perfusion imaging. entering the extravascular space. Since [13NH3] is lipophilic,
it can freely diffuse across the sarcolemma. Into the myocar-
3.2.2.1 [15O]H2O dial cell, [13N]NH3 is metabolically trapped by conversion to
In principle, [15O]H2O is an ideal radiotracer for quantitative 13
N-glutamine. In healthy subjects, myocardial retention of
flow measurements with PET, particularly myocardial blood [ N]NH3 is heterogeneous, with higher uptake in the inter-
13
flow. In fact, this tracer diffuses freely across the myocardio- ventricular septum than in the lateral wall. Furthermore, high
cyte membrane, with an extraction rate close to 100% at each uptake in the liver can interfere with evaluation of the infe-
pass, not affected by metabolic factors. Since the physical rior wall. Although the lung uptake of [13N]ammonia is usu-
half-life of 15O is 2.06 min, its utilization requires the avail- ally low, it may increase in patients with left ventricular
ability of an on-site cyclotron. As a freely diffusible tracer, dysfunction or with chronic pulmonary disease as well as in
[15O]H2O concentration in the blood pool is high, thus deter- smokers.
mining suboptimal imaging due to a low myocardial-to-back- Although the relatively short half-life of 13N requires
ground signal. This requires subtraction of the blood pool access to a cyclotron for radiotracer production, it has the
activity from the original image by the inhalation of [15O]CO, advantage to permit rest-stress assessment of myocardial
which binds irreversibly to hemoglobin so that the blood pool blood flow in the same session. Similar to SPECT, PET per-
activity can be digitally subtracted to visualize myocardium. fusion images can be graded visually and perfusion defects
[15O]H2O has been validated and extensively used for quantita- reported in terms of their extent, severity, and location.
tion of myocardial blood flow and coronary flow reserve. However, cardiac PET with radiotracers such as [13N]ammo-
Digital subtraction techniques and parametric imaging pro- nia offers the unique opportunity to estimate absolute myo-
duced by automated software packages have been imple- cardial blood flow and coronary reserve. This type of
mented to display perfusion at the voxel level in the form of quantitation requires dynamic acquisitions and compartmen-
parametric imaging depicting absolute quantitative flow val- tal analysis of radiotracer kinetics.
ues, thus facilitating the use of [15O]H2O in clinical practice—
provided of course that an in-house cyclotron is available. 3.2.2.4 18
F-Flurpiridaz
18
F-Flurpiridaz is an analog of the insecticide pyridaben, an
3.2.2.2 82
Rb-Chloride inhibitor of NADH-ubiquinone oxidoreductase, also known as
Rubidium-82 is a potassium analog taken up by the myocar- mitochondrial complex-1 (MC-1, a component of the electron
dium through active transport mediated by the Na+/K+ transport chain). 18F-Flurpiridaz inhibits MC-1 by competing
ATPase-pump. Although the physical half-life of 82Rb is for binding with ubiquinone without affecting the viability of
extremely short (1.25 min), this radiotracer is produced by myocardial cells. Uptake of this radiotracer in the myocar-
dium is rapid, followed by slow washout. In a phase II clinical 3.2.3 Energy Substrates
trial, cardiac PET with 18F-flurpiridaz showed superior image
quality and higher diagnostic accuracy than SPECT imaging. 3.2.3.1 2-Deoxy-2-[18F]fluoro-d-glucose
Absolute myocardial blood flow and coronary reserve may be Replacing the hydroxyl group on position 2 of glucose with
measured with 18F-flurpiridaz, yielding values in a similar a hydrogen atom yields a glucose analog, 2-deoxy-d-glu-
range as those measured with other radiotracers. cose, that is taken up by cells through the glucose transport-
ers (GLUTs) in a competitive manner with glucose. Once
3.2.2.5 Novel 18F-Labeled Myocardial Perfusion inside the cells, 2-deoxy-d-glucose is phosphorylated by
Agents hexokinase to form 2-deoxy-d-glucose-6-phosphate,
18
F-Fluoro-dihydrorotenone (18F-FDHR) is derived from rote- which—unlike glucose-6-phosphate—cannot progress
none, a neutral lipophilic compound that binds to MC-1 in the through the glycolytic pathway. Therefore, 2-deoxy-d-glu-
electron transport chain. The myocardial retention of this tracer cose-6-phosphate accumulates within the cell and is not
is higher and less affected by flow than that of 201Tl and better metabolized further. As an antagonist of glucose metabo-
correlated with flow both at 1 and 15 min after injection. lism, 2-deoxy-d-glucose has been investigated as a possible
18
F-p-Fluorobenzyl triphenyl phosphonium (18F-FBnTP, a therapeutic agent in conditions of high glucose metabolic
lipophilic cation) is taken up by myocardiocytes by passive demand, as typically observed in tumor tissues.
diffusion and accumulates mainly in the mitochondria. This The radiolabeled counterpart of 2-deoxy-d-glucose is
radiotracer (which is excreted mainly by renal clearance) 2-deoxy-2-[18F]fluoro-d-glucose (or [18F]FDG), which is
accumulates in the myocardium within 5 min after adminis- obtained by substituting the hydroxyl group at the C-2 posi-
tration and is efficiently retained up to 90 min. Activity in the tion of the native glucose molecule with the positron emitter
blood pool and in the lungs decreases over time, leading to 18
F (Fig. 3.13). [18F]FDG shares with 2-deoxy-d-glucose the
favorable uptake ratios and high-quality images at later time same transmembrane transport and intracellular metabolism
points after administration. Nonetheless, no clinical investi- characteristics that lead to virtually irreversible intracellular
gations with 18F-FBnTP have been reported yet. Due to accu- accumulation (Fig. 3.14).
mulation in the mitochondria as a function of the
mitochondrial membrane potential, this tracer has also been
defined as a “PET voltage sensor,” and its potential as an
imaging marker of apoptosis has also been suggested. a CH2OH
18
F-4-Fluorophenyl triphenyl phosphonium (18F-FTPP) is O
an additional lipophilic cationic tracer that accumulates in OH
OH
the mitochondria as function of their transmembrane poten-
tial. Early clinical investigations with this tracer yield myo- OH
cardial blood flow values comparable with those obtained 18F
with [13N]ammonia PET.

b
Key Learning Points

• PET perfusion imaging is the unique technique that
allows noninvasive absolute quantification of
regional blood flow and flow reserve.
• PET perfusion radiotracers (15OH2O; 82Rb; 13NH3)
have short half-lives that permit repeated measures
of blood flow in the same session and under stress-
ors stimuli.
• With the exception of 82Rb, the availability of these
radiotracers is limited by the need for on-site
cyclotron.
• 18F-Flurpiridaz is a promising agent with higher
extraction, higher image resolution, and longer
half-life than previously employed blood flow Fig. 3.13 Chemical structure (a) and tridimensional molecular repre-
sentation of (b) of [18F]FDG. Color codes for constituting elements:
radiotracers and is currently undergoing clinical green, 18F; red, O; gray, H; light blue, C (reproduced with permission
validation. from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di medicina
nucleare – Tecniche e applicazioni. Milan: Springer; 2010)
GLUT I & III

[18F]FDG-1-PO4 [18F]Fru-6-PO4 [18F]FDG-1-P-GL
[18F]FDG Hexokinase
[18F]FDG [18F]FDG-6-PO4
Glucose-6-
phosphatase
Nucleus
Mitochondria
Fig. 3.14 After entering cells through the same mechanism as native FDG-6-PO4 is not a substrate for any of the metabolic routes that
glucose (GLUT system), [18F]FDG is converted to [18F]FDG-6-PO4 glucose-6-PO4 normally undergoes, therefore remaining trapped
by the enzyme hexokinase, in the same manner as native glucose. The inside cells (modified from: Volterrani D, Erba PA, Mariani G, eds.
reverse enzyme glucose-6-phosphatase (which would revert [18F] Fondamenti di medicina nucleare – Tecniche e applicazioni. Milan:
FDG-6-PO4 back to [18F]FDG, freely diffusible outside the cell) oper- Springer; 2010)
ates at very low levels and almost exclusively in hepatocytes. [18F]
Based on the concept that glucose is the sole energy sub- subsequent studies with the use of [18F]FDG PET for imag-
strate for neurons, [18F]FDG was initially used in humans to ing various types of cancers.
map glucose metabolism in the brain. Exploring brain Upon systemic i.v. injection, [18F]FDG rapidly distributes
metabolism either under physiologic conditions or in patients in the body fluids and is taken up by various tissues through
with different abnormalities of the central nervous system glucose transporters and trapped intracellularly. Similarly as
has remained for some years the exclusive utilization of [18F] for glucose, [18F]FDG distributes to tumor tissue proportion-
FDG in humans, also because early PET technology limited ally to regional blood flow, and it is transported into the cell
size of the gantry to accommodate the head only of the sub- by facilitated diffusion mediated by specific membrane glu-
ject undergoing a PET scan. Current indications for PET cose transporters (GLUT family). Overexpression of
with [18F]FDG still include, among others, characterization GLUT-1 has been shown to be closely related to [18F]FDG
of patients with various neurological disorders (dementing uptake in human cancers and is also supposed to be an intrin-
disorders, epilepsy, movement disorders, and neuro-oncol- sic marker of hypoxia. In this regard, vascular growth in
ogy), in addition to indications in oncology and detection of tumor tissue is frequently insufficient to meet the local meta-
viable ischemic myocardium. bolic demand. The consequent chronic hypoxia causes the
The current predominant use of [18F]FDG in oncology release of specific biochemical mediators, like the hypoxia-
stems from studies performed in 1924, when Warburg et al. inducible factor-1a (HIF-1a), which is considered to support
reported that cancer cells have an increased glucose metabo- tumor growth by inducing angiogenesis via the expression of
lism, based on the observation that, even under aerobic con- vascular endothelial growth factor (VEGF). Recent studies
ditions, these cells produce large amounts of lactic acid from suggest that GLUT-1 overexpression is linked to the HIF-1a
the degradation and oxidation of glucose. This increased gly- production and that hypoxic conditions correspond to a
colytic activity is called Warburg effect and can be traced higher [18F]FDG uptake; moreover, tumor hypoxia activates
using PET imaging. Soon after the development of [18F]FDG the anaerobic glycolytic pathway. Once into the cell, [18F]
as a PET imaging agent, Som et al. pioneered studies demon- FDG is phosphorylated by mitochondrial hexokinase to form
strating that this tracer can accumulate in various spontane- [18F]FDG-6-phosphate. The elevated glycolysis of cancer
ous and transplanted tumors in animals. The first study with cells is associated with an increased hexokinase production
the use of [18F]FDG PET in patients with cancer was pub- and activity. Moreover, in cancer cells [18F]FDG-6-phosphate
lished by DiChiro and colleagues in 1982, showing that [18F] cannot be dephosphorylated or progress through other meta-
FDG uptake was higher in high-grade gliomas than in low- bolic pathways as glucose-6-phosphate would do, because of
grade gliomas. This investigation opened the avenue for all its nature of a modified molecule of glucose. As a result, [18F]
FDG-6-phosphate is trapped into the cell (see Fig. 3.14). PET imaging when evaluating patients with cancer.
Although [18F]FDG-6-phosphate may be converted back to Moreover, several malignancies with low glucose metabolic
[18F]FDG, the enzyme glucose-6-phosphatase is either at rate (such as early prostate cancer) or well-differentiated
very low levels or absent in most cancers. The overexpres- cancers or tumors with a predominant mucinous component
sion of GLUT transporters combined with the increased may not display high [18F]FDG uptake. Therefore, there is
hexokinase activity and trapping of the molecule results in an great interest in developing novel PET probes to target more
increased accumulation of [18F]FDG in cancer cells that is specific biomarkers of cancer proliferation.
significantly higher than in benign tissues. Biodistribution of On the other hand, evaluation of patients with suspected
[18F]FDG can be affected by several physiologic factors, the infection/inflammation constitutes per se a current indication
most important being blood glucose levels that can reduce for PET with [18F]FDG.
radiotracer uptake through competition. Therefore, before Finally, although the primary energy substrate for cardio-
performing a PET scan with [18F]FDG, patients should fast myocytes is fatty acid, these cells also utilize glucose as a
to reduce competition with plasma glucose. This require- metabolic substrate. In the assessment of myocardial viabil-
ment is especially important in diabetic patients, in whom ity, ischemic myocardium, which has downregulated contrac-
high glucose levels can be present that can interfere with tile function, utilizes more glucose than surrounding tissue,
imaging quality and results. with an adequate oxygen supply. When comparing the meta-
PET images are qualitatively interpreted in terms of the bolic pattern of myocardium to the distribution of perfusion,
presence, distribution, and intensity of radiotracer uptake viable ischemic regions demonstrate a relative increase in
into the suspected cancer mass. Intensity of radiopharmaceu- [18F]FDG compared to perfusion and compared to surround-
tical uptake relative to the normal tissue has been used to ing tissue. Specific preparation protocols have been described
differentiate benign from malignant lesions. Strategies have to optimize the diagnostic yield of the investigation—mostly
been developed for qualitative classification of regional [18F] based on glucose load inducing raised insulin levels.
FDG accumulation, such as the intensity of uptake compared
to liver uptake. Alternatively, [18F]FDG uptake is semi-quan- 3.2.3.2 Free Fatty Acids
titatively assessed using the standardized uptake value (SUV) The transport rate of free fatty acids (FFAs) in the brain,
according to the following formula: heart, and tumors is regulated by the FFA concentration in
SUV = Activity (Bq g−1)/[injected activity (Bq)/patient blood. Although FFAs can freely diffuse across cell mem-
weight (g)] branes, specific transporters, such as FATP1 and FATP6,
Although compartmental modeling approaches, origi- facilitate their transmembrane transport. Inside the cells,
nally proposed by Patlack and Sokoloff, allow absolute FFAs are either metabolized to triglycerides and sterols or
quantification of tumor glucose metabolism (in terms of transported into the mitochondria for energy production
moles of glucose/mass of tissue per time unit), technical through β-oxidation. Carnitine palmityl-transferase I and II
complexity limits their use in the clinical practice. (CPT I–II) transport fatty Acyl-CoA from the cell cytosol
[18F]FDG is the PET probe most commonly employed in into mitochondria, and oxidation generates acetyl-CoA for
nuclear medicine. More than 90% of oncologic PET scans the Krebs cycle to produce energy (ATP). In the cytosol of
use [18F]FDG, since glucose metabolism is increased in most normal cells, acetyl-CoA is converted to malonyl-CoA by
lung, colorectal, pancreatic, esophageal, stomach, head and acetyl-CoA carboxylase inhibitors reducing the CPT activity
neck, cervical, ovarian, and breast cancers, as well as mela- and transport of FFAs into the mitochondria. In tumor tissue,
noma and most types of lymphoma. A 2009 survey of the although FFA synthesis is increased to match the energy
National Oncologic PET Registry of the United States demand linked to the enhanced proliferation rate, the mito-
included results from nearly 41,000 [18F]FDG PET scans chondrial oxidation pathway is altered, and energy produc-
performed in more than 34,000 patients at 1368 centers. tion is shifted toward glucose utilization. This phenomenon
Thirty-five percent of the scans were for initial staging, 36% favors FFA utilization for production of phospholipids and
for restaging after treatment, and 29% for evaluation of dis- sterols.
ease recurrence. The [18F]FDG PET findings resulted in FFA consumption can be investigated with PET imaging
change in patient management in 38.0% of the cases. by tracing acetate metabolism through the use of [11C]ace-
However, despite being an outstanding radiotracer with tate (Fig. 3.15). Acetate is transported into the normal cell by
many advantages for use in oncology, [18F]FDG is not a spe- the monocarboxylate transporter; once inside the cell, [11C]
cific radiotracer for oncology, since it cannot distinguish a acetate can enter the Krebs cycle, to be rapidly converted to
high metabolic rate associated with cancer from increased tricarboxylic acids intermediates and eventually to [11C]CO2,
glucose metabolism related to other causes, such as infec- which can be found in the exhaled air. In this regard, [11C]
tion/inflammation. This can cause false positive results on acetate washout has been intensively studied as an index of
H a O
O
11C
H3C OH
H C 11C
b
OH
H
Fig. 3.15 Chemical structure of [11C]acetate
Fig. 3.16 Chemical structure (a) and tridimensional molecular repre-

the Krebs cycle activity. Since [11C]acetate has a high first- sentation (b) of [11C]palmitate. Color codes for constituting elements:
pass extraction, it is particularly suitable to assess myocar- light blue, C; red, O (reproduced with permission from: Volterrani D,
dial blood flow. Erba PA, Mariani G, eds. Fondamenti di medicina nucleare – Tecniche
There is however another possible metabolic pathway for e applicazioni. Milan: Springer; 2010)
[11C]acetate, i.e., conversion to acetyl-CoA by acetyl-CoA
synthase (present both in the cytosol and mitochondria) and i.v. injection [11C]palmitate undergoes rapid oxidation, stor-
FFA synthase. Fatty acid synthase is expressed predomi- age, or even redistribution as radiolabeled triglycerides. This
nantly in hepatocytes, but is also present in the white adi- results in the formation of several different 11C-labeled
pose, endometrial, and intestinal tissues and is overexpressed metabolites, which must be identified and subtracted from
in several malignancies including prostate and lung cancers. blood activity in order to obtain a correct input function.
Since overexpression of the enzyme is associated with tumor Therefore, dynamic studies with [11C]palmitate require fre-
aggressiveness and poor prognosis, [11C]acetate has been quent blood sampling and cumbersome identification and
used clinically for PET imaging in patients with prostate, measurements of 11C-metabolites, which may be estimated
renal, brain, or hepatocellular tumors or other cancers with directly or through indirect assessment based on the appear-
low GLUT expression and [18F]FDG uptake. However, ance of [11C]CO2 in the blood. As a consequence, the use of
important discrepancies have been reported on the clinical [11C]palmitate in humans is restricted to clinical
usefulness of PET with [11C]acetate, probably due to the use investigations.
of non-standardized imaging protocols. A recently published
meta-analysis of [11C]acetate PET in prostate cancer con-
cluded that further high-quality studies are needed to assess
its usefulness in clinical practice. Key Learning Points
The fatty acid analog 14-(R,S)-18F-fluoro-6-thia- • [18F]FDG is a glucose analog that follows the glyco-
heptadecanoic acid (18F-FTHA) can be used to assess fatty lytic pathway, accumulates into the cells, and is not
acid metabolism. 18F-FTHA is a long-chain FFA that is taken metabolized further.
up by tissues to enter mitochondria or to be incorporated into • [18F]FDG can be used to characterize neurological
complex lipids. In mitochondria, 18F-FTHA undergoes the disorders, to assess myocardial viability, and to
initial steps of β-oxidation, but it then remains trapped identify sites of inflammation/infection.
because further oxidation is blocked by its sulfur hetero- • The predominant use of [18F]FDG is in oncology to
atom. Because of this property, 18F-FTHA has provided good trace the increased glycolytic activity of tumor
target-to-background signal on PET images and appears masses (Warburg effect). PET imaging allows diag-
promising in the investigation of tumor fatty acid metabo- nosis, staging of the disease, and response to
lism. Unfortunately, tracer kinetics is not completely eluci- therapy.
dated, and the existing information is mostly limited to the • Free fatty acid consumption and utilization can be
heart, skeletal muscle, and liver, mainly derived from studies investigated in vivo with PET using [11C]acetate,
in small animals. 18
F-fluorothiaheptanoic acid, and [11C]palmitate.
[11C]Palmitate (Fig. 3.16) may in principle be preferable • [11C]Acetate is most frequently employed in clini-
to F-FTHA to trace FFA metabolism, because it is indistin-
18 cal studies to assess the increased lipid metabolism
guishable from endogenous palmitate and therefore reflects in tumor and to evaluate myocardial blood flow. Its
the rapid and somewhat complex metabolism of fatty acids. use is confined to PET centers with on-site cyclo-
Unfortunately, this property is also a major obstacle to derive tron facilities.
accurate estimates of [11C]palmitate kinetics. In fact, upon
3.2.4 Substrates for Phospholipid Synthesis a H3C 11CH

3
N+
Choline is necessary for the synthesis of phospholipids in HO
cell membranes, as well as for methyl metabolism, choliner- CH3
gic neurotransmission, transmembrane signaling, and lipid-
cholesterol transport and metabolism. The biochemical b
process explored with choline-based PET tracers is mem-
brane synthesis. After entering into cells through the activity
of specific transporters, choline is phosphorylated by the
enzyme choline kinase (CK). Phosphorylcholine is further
incorporated into phosphatidylcholine (lecithin) that is a
major phospholipid of all membranes. Through choline
dehydrogenase and betaine aldehyde dehydrogenase, phos-
phorylcholine is metabolized into betaine and finally, through
the enzyme choline acetyltransferase, into acetylcholine.
Since cell membranes duplicate at the same rate as the
rate of cell duplication, tumor cells incorporate choline rap-
idly to meet the need of rapid synthesis of cell membranes.
The levels of choline and phosphorylcholine are increased in
a variety of tumor cells, mirroring enhanced activation of sentation (b) of [11C]choline. Color codes of constituting elements:
choline uptake and phosphorylation. In slowly proliferating red, O; gray, H; light blue, C; deep blue, N (reproduced with permission
tumors, high phospholipid metabolite levels are related to from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di medicina
alterations in choline transport, incorporation, and utiliza- nucleare – Tecniche e applicazioni. Milan: Springer; 2010)
tion. Interestingly, malignant transformation is associated
with an increase in the cellular transport and phosphoryla-
tion of choline, as well as in increased expression of CK. a H3C
18
Choline can be directly labeled with 11C (yielding [11C] F
N+
choline, which is biochemically indistinguishable from
HO
native choline), or alternatively 18F can be used to label suit- CH3
able choline analogs. The successful use of [11C]choline
(Fig. 3.17) as a PET tracer for the detection of prostate can- b
cer was reported in 1998, leading to clinical investigations in
patients with other malignancies as well, including brain
tumors, lung cancer, esophageal cancer, colon cancer, blad-
der cancer, and other types of cancers. Physiologic uptake of
[11C]choline is observed in all glands (pituitary, salivary
glands, pancreas), as well as in the liver, kidney, bowel, and
stomach. As with all 11C-labeled tracers, the short physical
half-life of 11C (20 min) limits the use of [11C]choline to PET
centers that have an on-site cyclotron facility. To overcome
these limitations, analogs of choline labeled with 18F have
been developed, such as 18F-fluoro-ethylcholine (Fig. 3.18)
and 18F-fluoro-methylcholine (Fig. 3.19); these two tracers
have similar biodistribution patterns, with higher urinary
excretion than [11C]choline. In particular, 18F-fluoro-
methylcholine mirrors metabolic processing of native cho- Fig. 3.18 Chemical structure (a) and tridimensional molecular repre-
line more closely than 18F-fluoro-ethylcholine and is therefore sentation (b) of 18F-fluoro-ethylcholine. Color codes of constituting ele-
ments: red, O; gray, H; light blue, C; deep blue, N; green, 18F (reproduced
the PET imaging agent commercially available. with permission from: Volterrani D, Erba PA, Mariani G, eds.
[11C]choline and 18F-fluoro-methylcholine are part of the Fondamenti di medicina nucleare – Tecniche e applicazioni. Milan:
current armamentarium for the detection of prostate cancer Springer; 2010)
a H3C
18 • Choline is one of the main precursors for the syn-
N+ F
thesis of phospholipids in cell membranes, as well
HO as for methyl metabolism, cholinergic neurotrans-
CH3
mission, transmembrane signaling, and lipid-cho-
b lesterol transport and metabolism.
• In patients with malignancies, choline-based radio-
pharmaceuticals are utilized to explore with PET
imaging the biochemical process of membrane
synthesis.
• [11C]Choline is biochemically indistinguishable
from native choline and has been widely used for
PET imaging, mainly in patients with prostate can-
cer among other malignancies that do not overex-
press the GLUT system.
• The logistic limitations of 11C linked to its short
physical half-life (20 min, which restrict its use to
centers with an in-house cyclotron) have been over-
come with the use of 18F-labeled choline analogs,
such as 18
F-fluoro-ethylcholine and
Fig. 3.19 Chemical structure (a) and tridimensional molecular repre- 18
F-fluoro-methylcholine.
sentation (b) of 18F-fluoro-methylcholine. Color codes of constituting
elements: red, O; gray, H; light blue, C; deep blue, N; green, 18F (repro-
duced with permission from: Volterrani D, Erba PA, Mariani G, eds.
Fondamenti di medicina nucleare – Tecniche e applicazioni. Milan:
Springer; 2010)
3.2.5 Substrates for Protein Synthesis
and, to a lesser degree, for brain, bladder, and non-small cell As the basic building blocks of proteins, amino acids have
lung cancers. In particular, early prostate cancer cells fre- many physiologic roles, primarily for the synthesis of struc-
quently display low expression of the GLUT transport sys- tural proteins during cell replication, but also interacting
tem and low glucose consumption, thus resulting in frequent with energy substrates, and finally as precursors of bioactive
false negative results on [18F]FDG PET. Instead, PET with molecules (hormones, mediators of cell-to-cell interaction,
choline radiotracers is particularly effective in patients with etc.). In cancer cells the levels of protein synthesis are
prostate cancer, high Gleason scores (8–10), and high or increased for survival, replication, and invasion. As a conse-
increasing level of the prostate-specific antigen (PSA). PET quence, amino acid transport across plasma membranes by
with 18F-fluoro-methylcholine is currently used mostly to specific transporters is upregulated. Therefore, molecular
monitor the recurrence of prostate cancer after treatment imaging of amino acid metabolism is an attractive target,
with radical prostatectomy, radiotherapy, or hormonal ther- especially for malignancies where PET imaging with [18F]
apy alone, as it yields useful information for patient FDG has certain limitations.
management. In the past, the most widely used amino acid for tumor
imaging with PET was [11C]methionine (Fig. 3.20). Although
the mechanism of uptake in tumor cells is not completely
elucidated, it probably reflects substrate transport across the
tumor cell membrane and does not seem to be affected by
Key Learning Points
hypoxia. [11C]Methionine has been used mostly for investi-
• Because of their enhanced proliferation, tumor cells
gating brain tumors, where the use of [18F]FDG may be prob-
utilize greater amounts than normal cells of sub-
lematic because of the physiologically high uptake in the
strates for the synthesis of cell membranes.
normal brain gray matter. Instead, background uptake of this
• Phospholipids are essential components of cell
radiotracer in the normal brain is very low, and non-tumor
membranes.
conditions, such as edema and fibro-necrosis, do not exhibit
• Phospholipid precursors enter cells through the
significant uptake. However, despite the high target-to-back-
action of specific transporters, after which they are
ground ratio, [11C]methionine cannot be used for differenti-
usually phosphorylated by specific kinases.
ating benign from malignant cancers. Whereas, [11C]
methionine has advantages in assessing whether there is
tumor recurrence versus posttreatment fibrosis/necrosis even also in nuclear medicine centers without an in-house cyclo-
in patients with low-grade brain cancers, in whom [18F]FDG tron. The exact transport mechanisms of 18F-FET (which is
is not very helpful. The reported sensitivities and specifici- currently the most widely used amino acid tracer for brain
ties in the evaluation of the remnant or the relapse of the tumor imaging) are not fully clarified. 18F-FET belongs to
disease are 87% and 89%, respectively. the class of amino acids transported across the cell mem-
O-(2-18F-fluoroethyl)-l-tyrosine (18F-FET; see Fig. 3.21) brane via specific l-amino acid transporters, a system that is
was developed in the 1990s to provide an 18F-labeled amino especially active at the blood-brain barrier. Another possible
acid PET tracer with a longer physical half-life than the mechanism is a Na+-dependent activity associated to trans-
11
C-compounds, more suitable for routine clinical applications membrane transport. Cancers typically exhibit higher 18F-
FET uptake than nonneoplastic lesions, which may be
a O helpful in differential diagnosis. A meta-analysis of 13 18F-
FET PET studies with a total of 462 patients yielded a
S pooled 82% sensitivity and 76% specificity for the diagnosis
H11
3 C OH
of primary brain tumors (gliomas). However, diagnostic
applications in non-oncological diseases with increased
NH2
uptake of 18F-FET have been described, including brain
b abscesses, epilepsy, demyelinating processes, and in tissue
adjacent to cerebral ischemia or hematomas. Thus, 18F-FET
PET may be helpful in the preliminary assessment of equiv-
ocal brain lesions, but histological evaluation remains nec-
essary under most circumstances to provide a definite
diagnosis.
Furthermore, PET with 18F-FET is useful to assess
response of brain tumors to therapy at an early stage of treat-
ment with high accuracy, as well as to detect tumor recur-
rence during the follow-up. Such assessments can be
problematic even when high-resolution MR morphological
Fig. 3.20 Chemical structure (a) and tridimensional molecular repre- imaging is available.
sentation (b) of [11C]methionine. Color codes of constituting elements: The synthetic amino acid anti-1-amino-3-18F-fluoro-
red, O; gray, H; light blue, C; deep blue, N; yellow, S (reproduced with
permission from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di cyclobutanecarboxylic acid (18F-FACBC or 18F-fluciclovine;
medicina nucleare – Tecniche e applicazioni. Milan: Springer; 2010) see Fig. 3.22) is also transported through the l-amino acid
transporters (LAT) as well as by the system of alanine-ser-
O
ine-cysteine transporters (ASCT), which is primarily
a involved in the transmembrane transport of glutamine. In
many cancers, including prostate cancer, a metabolic shift
OH occurs when glutamine is used as an alternative energy
source instead of glucose. This metabolic shift is associated
18F NH2
with overexpression of the amino acid transporters (LAT1
O
b H H
N
OH

18F
sentation (b) of O-(2-18F-fluoroethyl)-l-tyrosine. Color codes of consti-
tuting elements: red, O; gray, H; light blue, C; deep blue, N; green, 18F
(reproduced with permission from: Volterrani D, Erba PA, Mariani G, Fig. 3.22 Chemical structure of Anti-1-amino-3-18F-fluoro-
eds. Fondamenti di medicina nucleare – Tecniche e applicazioni. Milan: cyclobutanecarboxylic acid (also known as 18F-FACBC, or
Springer; 2010) 18
F-fluciclovine)
and ASCT2) and with a more aggressive pattern of growth of to 6-18F-fluorodopamine, which in turn can be stored in
prostate cancer. 18F-fluciclovine can be used to trace in vivo secretory vesicles by the vesicular monoamine transporter
the increased activity of membrane transporters and for iden- (VMAT), thus remaining effectively trapped inside the cell.
tifying high-risk patients. Furthermore, the biodistribution of Alternatively, 6-18F-fluorodopamine can be degraded by
18
F-fluciclovine is advantageous versus [18F]FDG, because other enzymes, such as monoamine oxidase (MAO), yield-
low urinary excretion helps images interpretation of the PET ing breakdown products that are rapidly cleared from the
images in the pelvis and abdomen. The higher diagnostic cell.
performance of 18F-fluciclovine versus conventional imag- In order to prevent early decarboxylation of 18F-DOPA to
ing, especially for recurrent prostate cancer, has been dem- 6- F-fluorodopamine in extracerebral tissues and thus
18
onstrated by different clinical studies, whereas a relatively improve image quality, it is possible to administer carbidopa
low specificity limits the potential role of the radiotracer in [l-α-hydrazino-α-methyl-β-(3,4-dihydroxyphenyl) propi-
the primary diagnosis of prostate cancer. onic acid], an AADC inhibitor. This enhances striatal uptake
The amino acid dihydroxyphenylalanine (DOPA) con- by increasing the 18F-DOPA concentration in the plasma and
tains two hydroxyl groups on the third and fourth positions decreasing renal excretion. Carbidopa is used also to enhance
of the phenol ring; for PET imaging, it can be labeled with 18
F-DOPA uptake by tumor cells when imaging patients with
either 11C to yield the [11C]DOPA tracer or with 18F to yield neuroendocrine tumors (NETs). Physiological uptake/accu-
3,4-dihydroxy-6-18F-fluoro-phenylalanine (or 18F-DOPA; see mulation of 18F-DOPA occurs in the basal ganglia, liver, pan-
Fig. 3.23). l-DOPA is the precursor of the neurotransmitters creas, adrenal glands, gallbladder, biliary tract, kidneys,
dopamine, norepinephrine, and epinephrine (altogether ureters, and urinary bladder.
known as catecholamines). Therefore, PET with 18F-DOPA PET/CT with 18F-DOPA PET/CT is used clinically mainly
is used for functional imaging of pathologies in which to assess the extent of degeneration of presynaptic dopami-
enhanced intracellular transport and decarboxylation of nergic neurons (as in Parkinson’s disease), to image brain
DOPA are the diagnostic targets, allowing visualization of tumors, and to image well-differentiated NETs, such as med-
sympathetic cells. 18F-DOPA enters cells through the amino ullary thyroid cancer, pheochromocytoma, paraganglioma,
acid transport systems (LAT1) for large neutral amino acids, and congenital hyperinsulinemic hypoglycemia.
which are present in nearly all cells. The enzyme aromatic
amino acid decarboxylase (AADC) metabolizes 18F-DOPA
Key Learning Points
a OH • Phospholipid synthesis and turnover can be investi-
gated with PET using [11C]choline, 18F-fluoro-
HO ethylcholine, and 18F-fluoro-methylcholine.
O
• PET with positron-labeled choline tracers is partic-
NH2 ularly effective in tumors with low metabolic rate
18F
HO and mostly used to monitor the recurrence of pros-
tate cancer after treatment with radical prostatec-
b tomy, radiotherapy, or hormonal therapy.
• [11C]Methionine is the most widely used amino acid
for tumor imaging with PET, particularly for brain
cancers. It is supposed to reflect substrate transport
across the tumor cell membrane and does not seem
to be affected by hypoxia.
• 18F-fluoro-ethyl-tyrosine is an amino acid trans-
ported across the cell membrane via specific
l-amino acid transporters that has high diagnostic
accuracy for the diagnosis of primary brain tumors
(gliomas) and their recurrence after treatment.
• 18F-DOPA is a precursor of neurotransmitters (cat-
echolamines), and it is mainly used to assess the
Fig. 3.23 Chemical structure (a) and tridimensional molecular repre- extent of degeneration of presynaptic dopaminergic
sentation (b) of 3,4-dihydroxy-6-18F-fluoro-phenylalanine (or neurons (as in Parkinson’s disease), to image brain
18
F-DOPA). Color codes of constituting elements: red, O; gray, H; light
blue, C; deep blue, N; green, 18F (reproduced with permission from:
tumors, and to image well-differentiated neuroen-
Volterrani D, Erba PA, Mariani G, eds. Fondamenti di medicina nucle- docrine tumors.
are – Tecniche e applicazioni. Milan: Springer; 2010)
3.2.6 Substrates for DNA Synthesis porter 1 (hENT1), whose expression may be upregulated in
tumor cells. 18F-FLT follows the salvage pathway of DNA
Cell proliferation as a biological target is particularly appeal- synthesis and, like thymidine, undergoes phosphorylation by
ing in cancer imaging, as it is a characteristic feature of thymidine kinase 1 (TK1), to form 18F-FLT-monophosphate.
tumor development and growth. The nucleoside thymidine is In quiescent cells, TK1 activity is virtually absent, while it is
utilized for DNA synthesis by proliferating cells during the increased in proliferating cells, particularly in the S-phase of
S-phase of the cell cycle; unlike other nucleosides, thymi- the cell cycle. 18F-FLT-monophosphate may be phosphory-
dine is not incorporated into RNA. There are two main path- lated into 18F-FLT-diphosphate and 18F-FLT-triphosphate. At
ways leading to DNA synthesis. The “salvage pathway” this point of the pathway to DNA synthesis, 18F-FLT metabo-
recycles nucleoside precursors from outside the cell, while lism differs from that of native thymidine, as 18F-FLT-
the de novo endogenous pathway methylates deoxyuridine triphosphate is not incorporated into DNA and thus remains
monophosphate to form thymidine monophosphate through trapped in the cytosol. The rate-limiting step for intracellular
the action of thymidylate synthetase. The “salvage” pathway 18
F-FLT accumulation is initial phosphorylation by TK1.
is more specific for DNA synthesis than the de novo path- Phosphorylated 18F-FLT can be dephosphorylated by the
way, because precursors of the latter are also utilized for enzyme 5′-deoxynucleotidase, although this reaction occurs
RNA synthesis. at a slow rate. Thus, the accumulation of 18F-FLT-
[11C]Thymidine (Fig. 3.24) has initially been used for monophosphate, diphosphate, and triphosphate nucleotides
PET imaging of cell proliferation in vivo. However, this forms the basis for PET imaging. 18F-FLT is catabolized in
radiotracer is suboptimal for PET imaging because its rapid the liver with the production of 18F-FLT-glucuronide, which
catabolism results in the release and recirculation of labeled is released in the blood and excreted through the urinary sys-
metabolites in the blood; therefore, arterial blood sampling, tem. Therefore, 18F-FLT-glucuronide is the only metabolite
analysis of plasma metabolites, and complex compartmental in blood to be considered for kinetic modeling. In addition to
modeling are required for image interpretation and TK1 activity, uptake and retention of 18F-FLT are influenced
quantitation. also by the balance between the salvage pathway and de
The radiolabeled thymidine analog 18F-fluoro-3′-deoxy-3- novo DNA synthesis.
l-fluorothymidine (18F-FLT; see Fig. 3.25) is more suitable The intensity of 18F-FLT uptake correlates with the extent
for clinical use, both because of the longer physical half-life Ki-67 expression, a histopathologic marker of cell prolifera-
of the radionuclide and because it is more resistant to phos- tion. In various cancers (such as lung, breast, colon cancers
phorylase-mediated degradation in vivo. 18F-FLT is taken up as well as lymphomas), 18F-FLT uptake has been demon-
by cells through the same mechanisms as native thymidine. strated to be more tumor-specific than [18F]FDG uptake with
Transmembrane transport is facilitated by nucleoside trans- regard to inflammatory lesions, even though some 18F-FLT
porters, especially the human equilibrative nucleoside trans- uptake can be detected also in some inflammatory tissues
involving proliferating immune cells. Interest is growing
O also in the use of this tracer to assess response to antitumor
a b
CH3
HN
11C a O b
O N
O CH3
HN
HO
HO O N
O
HO
18F

Fig. 3.24 Chemical structure (a) and tridimensional molecular representation (b) of 18F-fluoro-3′-deoxy-3-l-fluorothymidine (18F-FLT).
sentation (b) of [11C]thymidine. Color codes of constituting elements: Color codes of constituting elements: red, O; gray, H; light blue, C;
red, O; gray, H; light blue, C; deep blue, N (reproduced with permission deep blue, N; green, 18F (reproduced with permission from: Volterrani
from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di medicina D, Erba PA, Mariani G, eds. Fondamenti di medicina nucleare –
nucleare – Tecniche e applicazioni. Milan: Springer; 2010) Tecniche e applicazioni. Milan: Springer; 2010)
therapy. Since 18F-FLT uptake in the cancer mass is consid- maceuticals provide excellent image quality with better spatial
ered to trace the degree of DNA synthesis and cell prolifera- resolution (3–5 mm) and diagnostic performance than the
tion, effective treatment should reduce tracer uptake. This γ-emitting analogs (e.g., 111In-DTPA-pentetreotide), particu-
rationale has been demonstrated to hold true, as several pub- larly if the tumor lesions are embedded in organs with high
lications have reported that the extent of 18F-FLT uptake physiologic uptake (e.g., the liver) and in case of small lesions
reflects the effects of anticancer therapies and is related to (<1.5 cm). These molecular probes differ in their affinity pro-
changes in proliferation rates as determined by ex vivo anal- file for the somatostatin receptor subtypes. In particular, the
yses. 18F-FLT uptake has potential as an imaging biomarker affinity of 68Ga-DOTA-TATE for the SSTR2 subtype is approx-
for response assessment. imately tenfold higher than that of 68Ga-DOTA-TOC. Whereas,
68
Ga-DOTA-NOC has high affinity for the SST2, SSTR3, and
SSTR5 subtypes.
Key Learning Points
68
Ga-labeled somatostatin analogs have mostly been used
• [11C]Thymidine and 18F-FLT are taken up by cells for imaging of thoracic and GEP-NETs. They can be used
through the same mechanisms as native thymidine also for imaging of focal congenital hyperinsulinism, breast
and can be used to study DNA synthesis and cell cancer, medulloblastoma, supratentorial primitive neuroec-
proliferation in vivo. todermal tumors, meningioma, neuroblastoma, pheochro-
• The rate-limiting step for intracellular 18F-FLT mocytoma/paraganglioma, and potentially in all solid tumors
accumulation is phosphorylation by thymidine that express somatostatin receptors.
kinase 1, whose activity is increased in proliferating
cells. 3.2.7.2 Estrogen and Androgen Receptor-
• The intensity of 18F-FLT uptake correlates with the Based Agents
extent Ki-67 expression, a histopathologic marker
18
F-Fluoroestradiol (18F-FES) binds to the estrogen recep-
of cell proliferation. tors and thus allows in vivo mapping of the estrogen receptor
• 18F-FLT uptake has a high potential as an imaging status of tumor lesions. This is particularly interesting in
biomarker for response assessment. breast cancer patients, where 18F-FES has been used to pre-
dict response to hormonal therapy.
The androgen analog 16-β-18F-fluoro-5α-dihydrotestosterone
(18F-FDHT) binds with high affinity to the androgen receptor,
3.2.7 P
ositron-Emitting Agents Based and promising results have been obtained using this agent for
on Ligand-Acceptor Interaction PET imaging in patients with prostate cancer. Assessment of the
androgen receptor status could provide important information
3.2.7.1 Somatostatin Receptor-Based Agents for patient selection and to predict response to therapy.
Somatostatin receptors, which are structurally related mem-
brane glycoproteins, are expressed in various normal tissues, 3.2.7.3 PSMA-Based Agents
including the central nervous system (CNS), anterior pitu- The prostate-specific membrane antigen (PSMA) is a type II
itary, thyroid, pancreas, gastrointestinal tract, spleen, and integral membrane glycoprotein (100–120 kDa), with an
adrenals. Five different subtypes of somatostatin receptors intracellular component, a transmembrane component, and a
(SSTR1–SSTR5) have been characterized, expressed with large extracellular domain, that was first detected in a human
variable density on tumor cell surface among various types prostatic carcinoma cell line. PSMA is a homologue of
of tumors. Expression of SSTR1 and SSTR2 are predomi- N-acetyl-l-aspartyl-l-glutamate peptidase I, an enzymatic
nant in gastro-entero-pancreatic neuroendocrine tumors protein which is physiologically active in the central nervous
(GEP-NETs). system, where it cleaves the neurotransmitter N-acetyl-l-
Several 68Ga-labeled somatostatin analogs have been evalu- aspartyl-l-glutamate into N-acetylaspartate and glutamate.
ated for PET imaging of neuroendocrine tumors, based on the In malignant tissue, it has been suggested that PSMA is
use of a suitable chelator for binding the radiometal, i.e., involved in angiogenesis, as increased PSMA is overex-
1,4,7,10-tetraazacyclododecane-N,N′,N″,N‴-tetraacetic acid pressed in the stroma adjacent to neovasculature of solid
(or DOTA). 68Ga-[DOTA0,Tyr3]octreotide (68Ga-DOTA-TOC) tumors. Due to its selective overexpression in 90–100% of
is the first 68Ga-labeled somatostatin analog developed for clin- prostate cancer lesions, PSMA is a reliable tissue marker for
ical use in cancer patients. Other 68Ga-labeled somatostatin prostate cancer and is considered an ideal target for theranos-
analogs have been developed, with either increased sensitivity tic applications. Increased PSMA expression is correlated
or more favorable affinity profile, such as 68Ga-[DOTA,1-Nal3] with increased tumor grade, advanced pathological stage,
octreotide (68Ga-DOTA-NOC) and 68Ga-[DOTA0,Tyr3]octreo- aneuploidy, and biochemical recurrence. Most importantly,
tate (68Ga-DOTA-TATE; see Fig. 3.26). These PET radiophar- PSMA is overexpressed when prostate cancer lesions become
Fig. 3.26 Chemical structure OH

(a) and tridimensional
a
molecular representation (b) OH
of 68Ga-[DOTA0,Tyr3] O O
octreotate (68Ga-DOTA- O
H
TATE). Color codes of NH N NH
constituting elements: red, O; N
N N NH H
gray, H; light blue, C; deep
blue, N; yellow, S; fuchsia, 68Ga O O
68
Ga (reproduced with S
O
permission from: Volterrani HO N N
D, Erba PA, Mariani G, eds. S HN
Fondamenti di medicina O
O O H
nucleare – Tecniche e NH2
applicazioni. Milan: NH N
HO HO N O
Springer; 2010)
H
O
H3C OH H 3C OH
androgen-independent. This feature makes PSMA particu- 68

Ga, 18F, 11C, 64Cu, and 86Y. One of these ligands, 68Ga-Glu-
larly valuable, since it has potential as an early indicator of urea-Lys(Ahx)-HBED-CC (also known as 68Ga-PSMA-11),
tumor progression after androgen-deprivation therapy. is currently the most extensively validated PSMA-based
Moreover, PSMA expression is a prognostic factor for dis- PET probe for clinical use in prostate cancer patients.
ease recurrence. HBED-CC stands for N,N′-Bis(2-hydroxy-5-(ethylene-
Although its denomination suggests a high degree of beta-carboxy)benzyl)ethylenediamine N,N′-diacetic acid, a
prostate specificity, PSMA is also expressed in several nor- highly effective chelator for labeling various compounds
mal tissues (such as the salivary glands, proximal renal with radiometals such as 68Ga (among others). 68Ga-PSMA-11
tubules, epididymis, ovary, the luminal side of the ileum- has a strong binding affinity to PSMA and is efficiently
jejunum, and astrocytes in the CNS), as well as in the neo- internalized into prostate cancer cells. Excellent contrast on
vasculature of tumors other than prostate cancer, including the images is seen in 1 h postinjection, even in patients with
bladder, pancreas, lung, and kidney cancers. low PSA levels.
Antibody-based PSMA PET agents targeting the extra- Further modifications of PSMA-11 have resulted in the
cellular domain of PSMA include 64Cu- and 89Zr-labeled production of a novel small PSMA-ligand, 68Ga-PSMA-617,
PSMA antibodies and antibody fragments, as well as that has a significantly higher binding affinity to PSMA asso-
64
Cu-labeled aptamers. More recently, small high-affinity ciated with a highly efficient internalization into prostate
ligands acting as PSMA antagonists have been labeled with cancer cells. In addition to 68Ga, PSMA-617 can be labeled
with 111In, 177Lu, and 90Y and therefore may potentially be 3.2.8 Amyloid Imaging Agents
used for imaging as well as for therapy.
PSMA ligands labeled with 18F (still in an experimental Alzheimer’s disease (AD) is characterized by the formation
phase of investigation) should result in improved spatial res- of insoluble aggregates of β-amyloid protein (β-amyloid
olution (due to the shorter positron range in tissues of 18F plaques) and neurofibrillary tangles (NFTs), which consist of
versus 68Ga) and more accurate quantitation. N-[N-[(S)-1,3- aggregates of hyper-phosphorylated tau protein. Although
dicarboxypropyl]carbamoyl]-4-18F-fluorobenzyl-l-cysteine AD can be suspected (with variable degrees of probability)
(18F-DCFBC) is a first-generation low-molecular-weight, in subjects presenting with progressive impairment of mem-
urea-based inhibitor of PSMA which binds to prostate can- ory and cognitive functions, a definitive diagnosis has tradi-
cer cells that express high PSMA levels. The basic structure tionally been based on the post-mortem histologic
is the same as the 68Ga-based agents, although labeling with demonstration of the presence of β-amyloid plaques and
18
F does not require the presence of a chelator for binding the NFTs in the brain using either the Congo red stain or the
radiometal. fluorescent dye thioflavin T.
A novel second-generation PSMA-ligand, 18F-DCFPyL, In 2004, an analog of thioflavin T labeled with 11C was
has higher affinity and faster clearance than 18F-DCFBC, developed and was denominated as 11C-Pittsburgh com-
thus resulting in higher tumor-to-background ratios and bet- pound B (11C-PiB). 11C-PiB binds to β-amyloid plaques with
ter detection of lower-grade or smaller-size prostate cancer high affinity, but not to amorphous amyloid plaques nor to
compared with 18F-DCFBC. Thus, 18F-DCFPyL represents a NFTs. Non-specific binding is seen mainly in white matter.
highly promising alternative to 68Ga-PSMA-HBED-CC for Specific cortical binding has been reported in more than 90%
PET imaging in prostate cancer. of patients with a clinical diagnosis of AD, whereas healthy
An additional PSMA-based agent, EuK-Subkff-68Ga- controls have cortical binding 1–5-fold lower than the cere-
DOTAGA (68Ga-PSMA I&T) can alternatively also be labeled bellum, which usually serves as a reference structure.
with 177Lu for therapeutic applications. However, some healthy elderly controls show higher 11C-PiB
binding, although the clinical implications of this amyloid
deposition are not clear. The apolipoprotein E ε4 allele
(APOE4) is frequently associated with β-amyloid plaques
Key Learning Points and increased uptake of 11C-PiB. Subjects with mild cogni-
• 68
Ga-labeled somatostatin analogs have been devel- tive impairment and 11C-PiB PET-positive scan have a higher
oped for PET imaging of tumor with somatostatin risk to develop manifest dementia than those with negative
receptors on cell surface. imaging. For several years 11C-PiB has been the most
• 68
Ga-DOTA-TOC, 68
Ga-DOTA-NOC, and employed PET radiotracer for clinical investigations on brain
68
Ga-DOTA-TATE are currently used in neuroendo- amyloid and AD, with sensitivities reported to range from
crine cancer PET imaging; these molecular probes 83% to 100% while specificities from 46% to 88%.
differ in their affinity profile for the somatostatin Suboptimal specificity suggests that 11C-PiB may be a bio-
receptor subtypes. marker of other diseases, such as mild cognitive impairment
• 18
F-Fluoroestradiol and 16-β-18F-fluoro-5α- caused by vascular causes or cerebral amyloid angiopathy.
dihydrotestosterone can be used to map receptor Furthermore, the widespread clinical use of 11C-PiB is lim-
status of tumor lesions and to predict response to ited by the very short physical half-life of ¹¹C (about 20 min),
hormonal therapy. which requires the availability of an on-site cyclotron. In
• PSMA is a tissue marker for prostate cancer and is order to overcome the above limitations, novel PET radio-
considered an ideal target for theranostic applica- tracers for β-amyloid plaques imaging have been developed
tions. Increased PSMA expression is correlated and labeled with 18F.
with increased tumor grade, advanced pathological 2-[3-18F-fluoro-4-(methylamino)phenyl]-6-benzothiazolol
stage, biochemical recurrence, and hormonal ther- (or 18F-Flutemetamol, approved for commercial use as
apy resistance. Vizamyl™) is the 3-fluoro-derivative of PiB. This radiotracer
• 68
Ga-labeled PSMA-ligands constitute currently the has a good safety profile with high affinity to β-amyloid
most extensively validated PET probes for clinical plaques and a good cortical-to-cerebellar uptake ratio.
use in prostate cancer patients. Amyloid imaging with 18F-Flutemetamol is best performed
about 90 min after i.v. injection. Cortical uptake can be
assessed qualitatively or semiquantitatively by using relative ity to tau than to β-amyloid plaques and very low white mat-
standardized uptake value ratios (SUVRs), the cerebellar cor- ter binding. Therefore, PET with 18F-flortaucipir displays
tex serving as a reference region. PET with 18F-Flutemetamol high imaging contrast between deposits of tau aggregates
can discriminate between clinically diagnosed AD or mild and background. Autoradiography studies of post-mortem
cognitive impairment from elderly healthy controls (sensitiv- tissues have consistently reported a strong binding affinity of
ity range, 91–97%; specificity range, 85–88%). 18
F-flortaucipir to tau in AD, in contrast to its much weaker
(E)-4-(2-(6-(2-(2-(2-([18F-fluoroethoxy)ethoxy)ethoxy)pyr- binding affinity to straight filament tau in non-AD tauopa-
idin-3-yl)vinyl)-N-methyl benzenamine (or 18F-Florbetapir, thies. However, the kinetic pattern of 18F-flortaucipir binding
approved for commercial use as Amyvid™) displays a binding in vivo is relatively slow, thus causing the SUVR values to
pattern similar to PiB, with a high affinity for β-amyloid increase even after 60 min postinjection; this kinetic feature
plaques. PET images are obtained 60 min after injection. limits the reproducibility of quantitative estimates.
Investigations in patients with variable cognitive status and 18
F-Flortaucipir also exhibits a high affinity for melanin-pro-
using histopathology at autopsy as the gold standard have ducing cells, including the substantia nigra, skin, retinal pig-
reported high diagnostic accuracy, with 100% specificity. ment epithelium, and melanomas. These distinct patterns of
Furthermore, patients who were found to be amyloid-positive tau-selective radiotracer binding suggest a potential role of
at PET with 18F-Florbetapir exhibited progressive worsening 18
F-flortaucipir as an imaging biomarker in different disease
in cognitive and functional status compared with those who conditions, both neurodegenerative diseases and non-CNS
were amyloid-negative. However, as for the other β-amyloid diseases.
imaging agents, differences in scan interpretation can deter-
mine significant variability in diagnostic accuracy. The use of
SUVRs and semiautomated software for quantification Key Learning Points
reduces inter-reader variability in image interpretation. • 11
C-Pittsburgh compound B, an analog of thioflavin
4-[(E)-2-(4-{2-[2-(2-18F-fluoroethoxy)ethoxy]ethoxy}phe- T, has been the first radiotracer developed for PET
nyl)vinyl]-N-methylaniline (or 18F-Florbetaben, approved for imaging of β-amyloid plaques.
commercial use as NeuraCeq) is a polyethylene glycol stil- • 11
C-PiB has been employed PET for clinical inves-
bene derivative with high specificity for β-amyloid plaques tigations on brain amyloid and Alzheimer’s disease,
without binding to tau protein, to tissue from frontotemporal with high diagnostic sensitivity.
lobe dementia, or brain tissue from dementing patients with • 18
F-Flutemetamol, 18
F-Florbetapir, and
Lewy bodies. Phase III trials have demonstrated that 18
F-Florbetaben have been recently introduced as
18
F-Florbetaben PET imaging can identify amyloid deposi- 18
F-labeled radiotracers for imaging amyloid
tion in the prodromal phase of AD and has high diagnostic deposition.
accuracy (comparable to that of PET with 11C-PiB or [18F] • These radiotracers have demonstrated high diag-
FDG) in the identification of AD. Interestingly, a study in nostic accuracy in the identification of β-amyloid
patients with multiple sclerosis demonstrated decreased brain accumulation in the prodromal phase of neu-
uptake in the white matter affected by the disease, thus open- rodegenerative disorders.
ing new horizons in amyloid imaging. • 18
F-Flortaucipir is the most widely employed PET
In addition to the three commercially available radiophar- radiotracer for in vivo imaging of tau aggregates in
maceuticals described above, PET radiotracers targeting the human brain and melanin-producing cells.
aggregated forms of tau protein have been developed. The
ideal tau PET tracer should display high selectivity for tau
aggregates associated with low/absent binding to β-amyloid
plaques. Selective radiotracers for tau aggregates have been 3.2.9 Tissue Hypoxia Imaging Agents
described, such as 7-[6-18F-fluoro-3-pyridinyl]-5H-
pyrido[4,3-b]indole (or 18F-flortaucipir), 18F-T807, 18F- Besides conditions linked to insufficient blood supply to normal
T808, and 18F-THK5351. Evidence is emerging that the organs (e.g., the brain and heart), hypoxia is a common feature
binding patterns observed with PET in various neurodegen- of solid cancers. In fact, uncontrolled growth of tumor tissue
erative disorders reflect the clinical and histopathological may not be mirrored by adequate vascularization, so that the
progression of the specific disease. The most widely core of the mass more distant from newly formed blood vessels
employed 18F-flortaucipir has a 25-fold greater binding affin- at the growing edge of the tumor becomes hypoxic and fre-
a 18
F-FMISO metabolites, thus resulting in tracer accumula-
OH
tion. Since 18F-FMISO only accumulates in hypoxic cells
18
N N F
with functioning nitroreductase enzymes, the uptake of this
radiotracer can be considered an index of both tissue hypoxia
and tissue viability.
N
Hypoxia imaging with PET can help optimize radiation
O O
therapy planning by identifying radioresistant regions of
b hypoxia in the tumor. Unfortunately, PET imaging with these
radiotracers suffers from some important limitations. In par-
ticular, PET scanners have a spatial resolution in the order of
mm, while tumor hypoxia is a cellular phenomenon with a
micrometer scale. Moreover, PET probes accumulate only in
viable tissue whose density may be very low in marked
hypoxia. Finally, uptake of these radiotracers is due to pas-
sive transmembrane diffusion, and the blood pool activity
remains high for several hours, thus reducing imaging qual-
ity and requiring delayed acquisition of PET images in order
to obtain quantitative estimations. Despite these limitations,
clinical studies performed in patients with head and neck or
lung cancers have demonstrated the ability to quantify
hypoxia and to predict patient prognosis. Thus, PET imaging
remains one of the few possible strategies to investigate
Fig. 3.27 Chemical structure (a) and tridimensional molecular repre- in vivo hypoxia and to provide comprehensive description of
sentation (b) of 18F-Fluoroimidazole (18F-FMISO). Color codes of con- tumor hypoxic states and hence better guide clinical assess-
stituting elements: red, O; gray, H; light blue, C; deep blue, N; green, ment and therapy design.
18
F (reproduced with permission from: Volterrani D, Erba PA, Mariani
G, eds. Fondamenti di medicina nucleare – Tecniche e applicazioni.
Milan: Springer; 2010)
Key Learning Points
• F-Fluoroimidazole is the most extensively investi-
18
quently necrotic. Hypoxia induces a metabolic shift toward glu- gated PET agent to image tissue hypoxia.
cose consumption, promotes angiogenesis and cancer • 18F-FMISO enters cells where it is reduced by nitro-
invasiveness, and also causes resistance to radiotherapy and reductase enzymes to remain trapped in cells with
chemotherapy. Although several different PET probes have reduced oxygen partial pressure.
been described to image tissue hypoxia, which constitutes an • Hypoxia imaging with PET is highly promising for
appealing target for molecular imaging, 18F-fluoroimidazole the choice of optimal treatments and therapy indi-
(18F-FMISO; see Fig. 3.27) remains the radiotracer most exten- vidualization, in particular during radiation therapy
sively investigated. The other PET agents evaluated to image planning.
tissue hypoxia include the nitroimidazole family (e.g.,
18
F-FAZA, 18F-HX4-25183380, 18F-EF5-112500504, 18F-EF3-
6918571) or compounds based on methylthiosemicarbazone
incorporating a copper atom, such as copper(II)-diacetyl-bis(N4- 3.2.10 Neoangiogenesis Imaging Agents
methylthiosemicarbazone) (ATSM) or copper-pyruvaldehyde-
bis(N4-methylthiosemicarbazone) (PTSM), that can be labeled In order for tumors to grow and metastasize, cancers must
with different positron-emitting copper radionuclides (i.e., 60Cu, stimulate the development of new vasculature (neoangiogen-
62
Cu, or 64Cu). However, most of these novel PET tracers have esis). Unlike normal blood vessels, tumor blood vessels are
not progressed through all the phases of clinical validation. chaotic, irregular, and leaky, which leads to irregular deliv-
Through passive transmembrane diffusion, 18F-FMISO ery of nutrients and therapeutic agents to the cancer mass.
enters cells where it is reduced by nitroreductase enzymes to Tumor vasculature is emerging as an important target for
remain trapped in cells with reduced oxygen partial pressure. anticancer therapy, and clinical trials in various malignancies
In fact, in normally oxygenated cells, the parent compound is are ongoing to assess the efficacy of antiangiogenic therapy
quickly re-oxydated, and the resulting metabolites diffuse with monoclonal antibodies against vascular endothelial
back to the extracellular compartment. Whereas, in hypoxic growth factor (VEGF), a central mediator of angiogenesis.
cells the low oxygen partial pressure prevents re-oxidation of For example, HuMV833 (a humanized anti-VEGF monoclo-
nal antibody) has been labeled with 124I and investigated in a

phase 1 clinical trial. • Zr-bevacizumab PET imaging is promising as an
89
After preliminary investigations in experimental animal early predictive biomarker for anti-neoangiogenic-
models, another anti-VEGF antibody, bevacizumab, has directed treatment.
been labeled with 89Zr for PET imaging in patients with • 64Cu-DOTA-PEGEcRGDyK2 and 18F-galacto-RGD
breast cancer, ovarian cancer, neuroendocrine tumors, renal are specific PET tracer agonists of αvβ3 receptor that
cell carcinoma, and brain gliomas. Preliminary results sug- is overexpressed on the activated endothelial cells
gest that 89Zr-bevacizumab PET imaging might be useful as of tumor neovasculature and on most tumor cells.
an early predictive biomarker for anti-VEGF-directed treat- • 18F-fluciclatide is a novel PET probe for αvβ3 expres-
ment. Moreover, in small animal studies, 64Cu-DOTA- sion that has provided promising results in various
VEGF121 has exhibited fast, specific uptake in highly malignancies.
vascularized tumors with a prominent level of VEGFR-2
expression.
Integrins constitute a superfamily of transmembrane
receptors that mediate cell-to-cell adhesion. Among other 3.2.11 Apoptosis Imaging Agents
functions, integrin-mediated adhesion modulates signaling
cascades that control cell motility, survival, proliferation, Cells undergoing apoptosis express phosphatidylserine on the
and differentiation. Imaging based on integrin-derived agents outer leaflet of their cell membrane. Annexin V, an endoge-
has a great potential, as they play important roles in tumor nous human protein with a high affinity for phosphatidylser-
metastasis and angiogenesis. ine, has been radiolabeled with both single-photon and PET
Currently, the most extensively studied integrin system is radionuclides (99mTc and 18F). The feasibility of imaging with
αvβ3, a receptor for vitronectin primarily expressed on the annexin V has been evaluated in small-sized studies in humans.
surface of platelets but also involved in the activation of Novel small-molecule probes included in the ApoSense
phagocytosis by macrophages and dendritic cells. This family (which can be defined as a novel technology for func-
receptor, which is also overexpressed on the activated endo- tional molecular imaging) have been designed to detect the
thelial cells of tumor neovasculature and on most tumor complex cellular alterations occurring during apoptosis.
cells, binds to the arginine-glycine-aspartic acid (RGD)- Among such nuclear probes, 2-(5-18F-fluoro-pentyl)-2-
containing components of the extracellular matrix, which are methyl-malonic acid (or 18F-ML-10, Fig. 3.28) is the first
significantly upregulated in tumor vasculature. Specific PET PET radiopharmaceutical for molecular imaging of apopto-
tracer agonists to this receptor have been developed, includ- sis that has demonstrated promising results in several small-
ing 64Cu-DOTA-PEGEcRGDyK2 and 18F-galacto-RGD. scale clinical trials. In particular, increased uptake of
18
F-Galacto-RGD successfully imaged αvβ3 expression in 18
F-ML-10 has been observed on PET/CT imaging in patients
human tumors, with good tumor-to-background ratio and with brain metastases treated with radiation therapy, although
highly promising results in patients with breast and lung can- with some heterogeneity in uptake intensity within the tumor
cers, sarcomas, and other solid tumors. However, the synthe- mass. Although the precise mechanism of radiotracer uptake
sis of 18F-Galacto-RGD is a rather complex radiochemistry
process, which has so far hampered its broad use in large
18F
clinical studies. The novel PET probe called 18F-fluciclatide
(chemically defined as Ac(1)-Lys(Unk)-Cys(2)-Arg-Gly-
Asp-Cys(2)-Phe-Cys(1)-Unk) contains two RGD sequences
and has great potential for imaging to assess response to anti-
angiogenic therapy. Promising results have been obtained
with this tracer in patients with various malignancies (includ- O
ing breast, brain, lung, and thyroid cancers, as well as mela-
nomas and sarcomas), and 18F-fluciclatide is expected to
C
become commercially available in the near future for PET OH
imaging of αvβ3 expression.
C OH
Key Learning Points

O
• PET offers the unique possibility to image in vivo
angiogenesis associated with tumor growth. Fig. 3.28 Chemical structure of 2-(5-18F-fluoro-pentyl)-2-methyl-
malonic acid (18F-ML-10)
remains unclear, these findings illustrate the potential of 18F- Croteau E, Renaud JM, Richard MA, Ruddy TD, Bénard F, et al. PET
metabolic biomarkers for cancer. Biomark Cancer. 2016;8(Suppl
ML-10 as a PET radiopharmaceutical for clinical imaging of 2):61–9.
apoptosis in tumors. DeGrado TR, Reiman RE, Price DT, Wang S, Coleman
Caspases are a family of intracellular cysteine proteases RE. Pharmacokinetics and radiation dosimetry of 18F-fluorocholine.
that are responsible for the initiation and execution of apop- J Nucl Med. 2002;43:92–6.
Doss M, Kolb HC, Walsh JC, Mocharla V, Fan H, Chaudhary A, et al.
tosis. In particular, caspase-3/7 is an attractive biological Biodistribution and radiation dosimetry of 18F-CP-18, a potential
marker for apoptosis. Clinical trials with promising results apoptosis imaging agent, as determined from PET/CT scans in
are ongoing with the use of 18F-CP18, the first PET radio- healthy volunteers. J Nucl Med. 2013;54:2087–92.
pharmaceutical developed for imaging of apoptosis based on Driessen RS, Raijmakers PG, Stuijfzand WJ, Knaapen P. Myocardial
perfusion imaging with PET. Int J Cardiovasc Imaging.
a substrate for caspase-3/7. 2017;33:1021–31.
Eckelman WC, Boyd M, Mairs RJ. Principles of molecular targeting
for radionuclide therapy. In: Strauss HW, Mariani G, Volterrani D,
Larson SM, editors. Nuclear oncology – from pathophysiology to
clinical applications. New York, NY: Springer; 2017. p. 35–66.
Key Learning Points Ell PJ, Gambhir SS, editors. Nuclear medicine in clinical diagnosis and
• 2-(5-18F-fluoro-pentyl)-2-methyl-malonic acid is a treatment. 3rd ed. New York, NY: Churchill Livingston; 2004.
PET radiopharmaceutical developed for molecular Elsinga PH. Trends on the role of PET in drug development. Singapore:
imaging of apoptosis that has demonstrated promis- World Scientific; 2012.
Feinendegen LE, Shreeve WW, Eckelman WC, Bahk YW, Wagner HN
ing results in human cancers. Jr. Molecular nuclear medicine: the challenge of genomics and pro-
• 18F-CP18 is the first radiotracer developed for PET teomics to clinical practice. Berlin: Springer Science & Business
imaging based on a substrate for caspase, a family of Media; 2012.
intracellular proteases associated with apoptosis. Filss CP, Cicone F, Shah NJ, Galldiks N, Langen KJ. Amino acid PET
and MR perfusion imaging in brain tumours. Clin Transl Imaging.
2017;5:209–23.
Gaykema SB, Brouwers AH, Lub-de Hooge MN, Pleijhuis RG,
Timmer-Bosscha H, Pot L, et al. 89Zr-bevacizumab PET imaging in
primary breast cancer. J Nucl Med. 2013;54:1014–8.
Grassi I, Morigi JJ, Nanni C, Fanti S. FDG and other radiopharma-
Further Reading ceuticals in the evaluation of liver lesions. Clin Transl Imaging.
2014;2:115–27.
Barthel H, Sabri O. Clinical use and utility of amyloid imaging. J Nucl Hara T. 18F-fluorocholine: a new oncologic PET tracer. J Nucl Med.
Med. 2017;58:1711–7. 2001;42:1815–7.
Barwick T, Bencherif B, Mountz JM, Avril N. Molecular PET and Herbert JC, Eckelman WC, Neumann RD, editors. Nuclear medicine –
PET/CT imaging of tumour cell proliferation using F-18 fluoro- diagnosis and therapy. New York, NY: Thieme Medical Publishers;
L-thymidine: a comprehensive evaluation. Nucl Med Commun. 1996.
2009;30:908–17. Herholz K, Ebmeier K. Clinical amyloid imaging in Alzheimer’s dis-
Baum RP, Rösch F. Theranostics, gallium-68, and other radionuclides: ease. Lancet Neurol. 2011;10:667–70.
a pathway to personalized diagnosis and treatment. Berlin: Springer IAEA. Good practice for introducing radiopharmaceuticals for clinical
Science & Business Media; 2012. use. Vienna: International Atomic Energy Agency (IAEA); 2015.
Beer AJ, Kessler H, Wester H-J, Schwaiger M. PET imaging of integrin IAEA. Operational guidance on hospital radiopharmacy. Vienna:
αvβ3 expression. Theranostics. 2011;1:48–57. International Atomic Energy Agency (IAEA); 2008.
Bouchelouche K, Turkbey B, Choyke PL. PSMA PET in prostate Ido T, Wan CN, Casella V, Fowler JS, Wolf AP, Reivich M, et al.
cancer – a step towards personalized medicine. Curr Opin Oncol. Labeled 2-deoxy-D-glucose analogs: 18F-labeled 2-deoxy-2-fluoro-
2016;28:216–21. D-glucose, 2-deoxy-2-fluoro-D-mannose and 14C-2-deoxy-2-fluoro-
Brown J. Effects of 2-deoxyglucose on carbohydrate metabo- D-glucose. J Labeled Compounds Radiopharm. 1978;24:174–83.
lism: review of the literature and studies in the rat. Metabolism. Itsenko O, Goméz-Vallejo V, Llop J, Koziorowski J. On 11C chemis-
1962;11:1098–112. try reviews – surveying and filling the gaps. Curr Organ Chem.
Chacko A-M, Divgi CR. Radiopharmaceutical chemistry with 2013;17:2067–96.
iodine-124: a non-standard radiohalogen for positron emission Jacobson O, Kiesewetter DO, Chen X. Fluorine-18 radiochemistry, label-
tomography. Med Chem. 2011;7:395–412. ing strategies and synthetic routes. Bioconjug Chem. 2015;26:1–18.
Choi Y, Ha S, Lee YS, Kim YK, Lee DS, Kim DJ. Development of Khalil MM. Basic science of PET imaging. New York, NY: Springer;
tau PET imaging ligands and their utility in preclinical and clinical 2016.
studies. Nucl Med Mol Imaging. 2018;52:24–30. Klingensmith WC III. The mathematics and biology of the biodistri-
Chondrogiannis S, Marzola MC, Al-Nahhas A, Venkatanarayana TD, bution of radiopharmaceuticals – a clinical perspective. New York,
Mazza A, Opocherand G, et al. Normal biodistribution pattern and NY: Springer; 2016.
physiologic variants of F-DOPA PET imaging. Nucl Med Commun. Knaapen P, de Haan S, Hoekstra OS, Halbmeijer R, Appelman YE,
2013;34:1141–9. Groothuis JG, et al. Cardiac PET-CT: advanced hybrid imaging for
Christensen NL, Jakobsen S, Schacht AC, Munk OL, Alstrup AKO, the detection of coronary artery disease. Neth Heart J. 2010;18:90–8.
Tolbod LP, et al. Whole-body biodistribution, dosimetry, and metab- Kowalsky RJ, Falen SW, editors. Radiopharmaceuticals in nuclear
olite correction of [11C]palmitate: a PET tracer for imaging of fatty pharmacy and nuclear medicine. 3rd ed. Washington, DC: American
acid metabolism. Mol Imaging. 2017;16:1–9. Pharmacists Association; 2011.
Kuik WJ, Kema IP, Brouwers AH, Zijlma R, Neumann KD, Dierckx Volterrani D, Larson SM, editors. Nuclear oncology – from patho-
RAJO, et al. In vivo biodistribution of no-carrier-added 6-18F-fluoro- physiology to clinical applications. New York, NY: Springer; 2017.
3,4-dihydroxy-L-henylalanine (18F-DOPA), produced by a new p. 85–98.
nucleophilic substitution approach, compared with carrier-added Owunwanne A, Patel M, Sadek S, editors. The handbook of radiophar-
18
F-DOPA, prepared by conventional electrophilic substitution. J maceuticals. New York, NY: Springer; 1995.
Nucl Med. 2015;56:106–12. Pauwels EK, Ribeiro MJ, Stoot JH, McCready VR, Bourguignon M,
Kwekkeboom DJ, Kam BL, van Essen M, Teunissen JJ, van Eijck CH, Mazière B. FDG accumulation and tumor biology. Nucl Med Biol.
Valkema R, et al. Somatostatin-receptor-based imaging and therapy 1998;25:317–22.
of gastroenteropancreatic neuroendocrine tumors. Endocr Relat Price EW, Orvig C. Matching chelators to radiometals for radiopharma-
Cancer. 2010;17:R53–73. ceuticals. Chem Soc Rev. 2014;43:260–90.
Lappin G, Temple S. Radiotracers in drug development. Boca Raton, Price TW, Greenman J, Stasiuk GJ. Current advances in ligand design
FL: CRC Press; 2006. for inorganic positron emission tomography tracers 68Ga, 64Cu, 89Zr
Lütje S, Heskamp S, Cornelissen AS, Poeppel TD, van den Broek and 44Sc. Dalton Trans. 2016;45:157092–724.
SAMW, Rosenbaum-Krumme S, et al. PSMA ligands for radio- Reivich M, Kuhl D, Wolf A, Greenberg J, Phelps M, Ido T, et al.
nuclide imaging and therapy of prostate cancer: clinical status. Measurement of local cerebral glucose metabolism in man with
Theranostics. 2015;5:1388–401. 18
F-2-fluoro-2-deoxy-D-glucose. Acta Neurol Scand Suppl.
Lyoo CH, Cho H, Choi JY, Ryu YH, Lee MS. Tau positron emission 1977;64:190–1.
tomography imaging in degenerative parkinsonisms. J Mov Disord. Rice SL, Roney CA, Daumar P, Lewis JS. The next generation of
2018;11:1–12. positron emission tomography radiopharmaceuticals in oncology.
Mach RH. Small molecule receptor ligands for PET studies of the cen- Semin Nucl Med. 2011;41:265–82.
tral nervous system — focus on G protein coupled receptor. Semin Sandström M, Velikyan I, Garske-Román U, Sörensen J, Eriksson B,
Nucl Med. 2017;47:524–35. Granberg D, et al. Comparative biodistribution and radiation dosim-
Maddahi J, Packard RR. Cardiac PET perfusion tracers: current status etry of 68Ga-DOTATOC and 68Ga-DOTATATE in patients with neu-
and future directions. Semin Nucl Med. 2014;44:333–43. roendocrine tumors. J Nucl Med. 2013;54:204–10.
Martínez G, Vernooij RW, Fuentes Padilla P, Zamora J, Bonfill Cosp Sarparanta M, Demoin DW, Cook BE, Lewis JS, Zeglis BM. Novel
X, Flicker L. 18F PET with florbetapir for the early diagnosis of positron-emitting radiopharmaceuticals. In: Strauss HW, Mariani G,
Alzheimer’s disease dementia and other dementias in people Volterrani D, Larson SM, editors. Nuclear oncology – from patho-
with mild cognitive impairment (MCI). Cochrane Database Syst physiology to clinical applications. New York, NY: Springer; 2017.
Rev. 2017;(11):CD012216. https://doi.org/10.1002/14651858. p. 129–72.
CD012216.pub2. Schelhaas S, Heinzmann K, Bollineni VR, Kramer GM, Liu Y, Waterton
Martínez G, Vernooij RW, Fuentes Padilla P, Zamora J, Flicker L, JC, et al. Preclinical applications of 3′-deoxy-3′-[18F]fluorothymi-
Bonfill CX. 18F PET with flutemetamol for the early diagnosis dine in oncology – a systematic review. Theranostics. 2017;7:40–50.
of Alzheimer’s disease dementia and other dementias in people Schuster DM, Nanni C, Fanti S. PET tracers beyond FDG in prostate
with mild cognitive impairment (MCI). Cochrane Database Syst cancer. Semin Nucl Med. 2016;46:507–21.
Rev. 2017;(11):CD012884. https://doi.org/10.1002/14651858. Sols A, Crane RK. Substrate specificity of brain hexokinase. J Biol
CD012884. Chem. 1954;210:581–95.
Martínez G, Vernooij RW, Fuentes Padilla P, Zamora J, Flicker L, Som P, Atkins HL, Bandoypadhyay D, Fowler JS, MacGregor RR,
Bonfill CX. 18F PET with florbetaben for the early diagnosis of Matsui K, et al. A fluorinated glucose analog, 2-fluoro-2-deoxy-D-
Alzheimer’s disease dementia and other dementias in people with glucose (F-18): nontoxic tracer for rapid tumor detection. J Nucl
mild cognitive impairment (MCI). Cochrane Database Syst Rev. Med. 1980;21:670–5.
2017;11:CD012883. https://doi.org/10.1002/14651858.CD012883. Strauss HW, Mariani G, Volterrani D, Larson SM, editors. Nuclear
Mathis CA, Mason NS, Lopresti BJ, Klunk WE. Development of posi- oncology – from pathophysiology to clinical applications.
tron emission tomography β-amyloid plaque imaging agents. Semin New York, NY: Springer; 2017.
Nucl Med. 2012;42:423–32. Subramanian G, Rhodes BA, Cooper JF, Sodd VJ. Radiopharmaceuticals.
Mertens K, Slaets D, Lambert B, Acou M, De Vos F, Goethals I. PET Society of Nuclear Medicine: New York, NY; 1975.
with 18F-labelled choline-based tracers for tumour imaging: a review Theobald T, editor. Sampson’s textbook of radiopharmacy. 4th ed.
of the literature. Eur J Nucl Med Mol Imaging. 2010;37:2188–93. London: Pharmaceutical Press; 2010.
Molecular Imaging and Contrast Agent Database (MICAD) – Trembath L, Newell M, Devous MD Sr. Technical considerations in
NCBI – NIH. Accessible at https://www.ncbi.nlm.nih.gov/books/ brain amyloid PET imaging with 18F-Florbetapir. J Nucl Med
NBK5330/. Technol. 2015;43:175–84.
Morris E, Chalkidou A, Hammers A, Peacock J, Summers J, Keevil Podoloff DA, et al. NCCN task force: clinical utility of PET in a variety
S. Diagnostic accuracy of 18F amyloid PET tracers for the diagnosis of tumor types. J Natl Compr Canc Netw. 2009;7(Suppl 2):S1–S26.
of Alzheimer’s disease: a systematic review and meta-analysis. Eur Vallabhajosula S. Molecular imaging – radiopharmaceuticals for PET
J Nucl Med Mol Imaging. 2016;43:374–85. and SPECT. Berlin-Heidelberg: Springer; 2009.
Mossine AV, Brooks AF, Jackson IM, Quesada CA, Sherman P, Cole van Asselt SJ, Oosting SF, Brouwers AH, Bongaerts AH, de Jong JR,
EL, et al. Synthesis of diverse 11C-labeled PET radiotracers via direct Lub-de Hooge MN, et al. Everolimus reduces 89Zr-bevacizumab
incorporation of [11C]CO2. Bioconjug Chem. 2016;27:1382–9. tumor uptake in patients with neuroendocrine tumors. J Nucl Med.
Nagengast WB, de Vries EG, Hospers GA, Mulder NH, de Jong 2014;55:1087–92.
JR, Hollema H, et al. In vivo VEGF imaging with radiolabeled Varrone A, Asenbaum S, Vander Borght T, Booij J, Nobili F, Någren K,
bevacizumab in a human ovarian tumor xenograft. J Nucl Med. et al. EANM procedure guidelines for PET brain imaging using [18F]
2007;48:1313–9. FDG, version 2. Eur J Nucl Med Mol Imaging. 2009;36:2103–10.
Núñez Miller R, Pozo MA. Non-FDG PET in oncology. Clin Transl Villemagne VL, Doré V, Burnham SC, Masters CL, Rowe CC. Imaging
Oncol. 2011;13:780–6. tau and amyloid-β proteinopathies in Alzheimer disease and other
Orsini F, Lorenzoni A, Puta E, Mariani G. Positron-emitting radiophar- conditions. Nat Rev Neurol. 2018;14:225. https://doi.org/10.1038/
maceuticals for diagnostic applications. In: Strauss HW, Mariani G, nrneurol.2018.9.
Wang X, Feng H, Zhao S, Xu J, Wu X, Cui J, et al. SPECT and PET Wild D, Bomanji JB, Benkert P, Maecke H, Ell PJ, Reubi JC, Caplin
radiopharmaceuticals for molecular imaging of apoptosis: from ME. Comparison of 68Ga-DOTANOC and 68Ga-DOTATATE PET/
bench to clinic. Oncotarget. 2017;8:20476–95. CT within patients with gastroenteropancreatic neuroendocrine
Warburg O, Wind F, Negelein E. The metabolism of tumors in the body. tumors. J Nucl Med. 2013;54:364–72.
J Gen Physiol. 1927;8:519–30. Xia C, Dickerson BC. Multimodal PET imaging of amyloid and tau
Waxman AD, Herholz K, Lewis DH, Herscovitch P, Minoshima S, pathology in Alzheimer disease and non-Alzheimer disease demen-
Ichise M, et al. Society of nuclear medicine procedure guideline for tias. PET Clin. 2017;12:351–9.
FDG PET brain imaging. Version 1.0, approved February 8, 2009. Xu Z, Li XF, Zou H, Sun X, Shen B. 18F-Fluoromisonidazole in tumor
Accessed at www.snmmi.org. hypoxia imaging. Oncotarget. 2017;8:94969–79.
Welch MJ, Redvanly CS, editors. Handbook of radiopharmaceuticals:
radiochemistry and applications. Hoboken, NJ: Wiley; 2003.
Wick AN, Drury DR, Nakada HI, Wolfe JB. Localization of the pri-
mary metabolic block produced by 2-deoxyglucose. J Biol Chem.
1957;224:963–9.
Radiopharmaceuticals for Therapy
4
Federica Orsini, Federica Guidoccio, and Giuliano Mariani
Contents
4.1 Introductory Background 100
4.2 Beta-Emitting Radionuclides 101
4.2.1 Iodine-131 (131I) 101
4.2.2 Yttrium-90 (90Y) 102
4.2.3 Rhenium-188 (188Re) and Rhenium-186 (186Re) 102
4.2.4 Lutetium-177 (177Lu) 103
4.2.5 Phosphorus-32 (32P) 103
4.2.6 Strontium-89 (89Sr) 103
4.2.7 Samarium-153 (153Sm) 104
4.2.8 Holmium-166 (166Ho) 104
4.3 Radionuclides Emitting Alpha Particles 104
4.3.1 Radium-223 (223Ra) 105
4.4 Radionuclides Emitting Auger Electrons 105
4.5 Radiopharmaceutical-Based Emission of β Particles
−
106
4.5.1 Sodium 131I-Iodide 106
4.5.2 Meta-[131I]Iodo-Benzyl-Guanidine (131I-MIBG) 107
4.5.3 Radiopharmaceuticals for Radioimmunotherapy 108
4.5.4 131I-Lipiodol and 188Re-HDD-Lipiodol 110
4.5.5 90Y- and 166Ho-Microspheres 110
4.5.6 90Y- or 177Lu-Labeled Somatostatin Analogs 111
4.5.7 32P-Orthophosphate 113
4.5.8 89 Sr-Chloride (89SrCl2) 113
4.5.9 153Sm-Lexidronam (153Sm-EDTMP) 114
4.5.10 186/188Re-Etidronate (186/188Re-HEDP) 114
4.6 Alpha-Emitting Radiopharmaceuticals 115
4.6.1 223Ra-Dichloride 115
References 116
Learning Objectives
• Understand the basic concepts on the physical and bio-
logical events underlying the radiobiological effects
caused by radionuclides utilized for therapy.
F. Orsini (*) • Learn the general concepts of radiobiological damage
Nuclear Medicine Unit, “Maggiore della Carità” University caused by radionuclides emitting β− particles.
e-mail: federica.orsini@maggioreosp.novara.it
• Learn the main physical and chemical characteristics of
radionuclides emitting β− particles commonly employed
F. Guidoccio · G. Mariani
for therapy.
Research and Advanced Technologies in Medicine and Surgery, • Learn the general concepts of radiobiological damage
University of Pisa, Pisa, Italy caused by radionuclides emitting α++ particles.

100 F. Orsini et al.
• Learn the main physical and chemical characteristics of tion density along the track of radiation, whether electro-
the radionuclide emitting α++ particles commonly magnetic radiation or particle radiation. Low-LET radiations
employed for therapy—radium-223. produce ionizations only sparsely along their track as is typi-
• Understand the potential for therapy of radionuclides cal for γ-rays or X-rays. High-LET radiations (more typical
emitting very low-energy β− particles—the Auger for electrically charged particles) are more destructive to bio-
electrons. logical material than low-LET radiations. In fact, with the
• Become familiar with the radiopharmaceuticals com- same activity/dose, low-LET radiations induce few reactive
monly used for therapy, with special attention to their (biochemically toxic) radicals within a cell/tissue, whereas
in vivo pharmacokinetics and radiation dosimetry esti- high-LET radiation transfers most of their energy in a small
mates to normal tissues/organs. region of the cell. If these ionizations are in the nucleus, the
result is damage to the DNA that is more difficult to repair.
Table 4.1 lists the main physical characteristics (energy, tis-
4.1 Introductory Background sue path, and LET) for the particles that are (or can be) uti-
lized for radionuclide therapy.
Diagnostic radiopharmaceuticals emit γ or β+ radiation, the In addition to particle emission causing a cytocidal radio-
former traveling at distances long enough to be detected out- biological effect, some radionuclides used for therapy also
side the body for imaging, and the latter originating in turn emit γ-rays or β+ particles, which do not contribute to the
high-energy photons (through the annihilation process that effectiveness of therapy (and may even lead to increased irra-
takes place upon interaction with β− particles) that can simi- diation to nontarget tissues), but can be used to image the
larly be detected outside the body for imaging. Instead, ther- distribution of the therapeutic agent. These images can be
apeutic radiopharmaceuticals exert their radiobiological used to calculate the dose delivered to the lesion (and other
action through the emission of electrically charged particles tissues) by the therapeutic agent.
(α++ particles, β− particles, Auger electrons) that deposit their Therapeutic radiopharmaceuticals can be administered
energy within a much more limited range in tissue—from systemically by intravenous administration or orally, intra-
fraction of a micrometer to a few millimeters at most. This cavitarily, or intra-arterially. The majority of radionuclide
energy deposition within a restricted radius from the decay/ therapies are administered for treating various forms of
emission point can induce radiation damage in the cancer with either curative or palliation purposes. Less fre-
surrounding cells that causes the desired radiobiological
quently, radionuclide therapy is utilized for benign disease,
effect—cell death. Depending on their energy, the β− parti- such as treatment of hyperthyroidism.
cles typically travel a few millimeters from their emission Selection of the most appropriate radionuclide for therapy
point, while α++ particles typically travel few tens of microm- must consider the type of radiation, energy, effective half-life
eters in tissues. The radionuclides emitting electrically on the target tissue, and chemistry in relation to the carrier
charged particles with energy suitable for therapy are linked molecule.
to carrier molecules capable of (selectively) transporting the
radiotracers to the target tissues [1].
The goal of radionuclide therapy is to cause irreversible
damage to tumor cell DNA, resulting in death of the tumor Key Learning Points
cell [2] (see also Chap. 11 of this book “Principles of • Radionuclides utilized for therapy exert their radio-
Radiation Biology and Dosimetry for Nuclear Medicine biological action through the emission of electri-
Procedures”). Biologic response to ionizing radiation cally charged particles (α++ particles, β− particles,
depends on many factors: cell radiosensitivity, location of Auger electrons) that deposit their energy within a
radionuclide within the cell or outside, and physical proper- limited range in tissue—from fraction of a microm-
ties such as absorbed dose and linear energy transfer (LET) eter to a few millimeters.
of the emitted radiation. Linear energy transfer reflects the • Radionuclides emitting electrically charged parti-
amount of energy deposited by radiation per unit length of cles suitable for therapy are linked to carrier mole-
travel in soft tissue, expressed in kiloelectronvolt per cules capable of (selectively) transporting the
micrometer (keV/μm). It can also be defined as the ioniza- radiotracers to the target tissues.
• Radionuclide therapy causes irreversible damage to
Table 4.1 Physical characteristics (energy, path in soft tissue, and tumor cell DNA causing death of the tumor cell.
LET) of different particle emissions utilized for radionuclide therapy • Biologic response to ionizing radiation depends on
Emission Energy (min-max) Soft tissue path LET (keV/μm) cell radiosensitivity, location of the radionuclide
α++ particles 5–9 MeV 40–100 μm 80 within the cell or outside, and physical properties
Auger 450 eV–2 keV 0.002–0.5 μm 4–26 such as absorbed dose and linear energy transfer of
electrons the emitted radiation.
β− particles 0.05–2.3 MeV 50–1200 μm 0.2
4 Radiopharmaceuticals for Therapy 101
space is similar to a tangled ball of yarn, and the maximum

• Concomitant emission of γ-rays or β+ particles by path range in tissues reported for each radionuclide is
the radionuclide used for therapy can be used to reached by only a few of the electrons emitted by that
image whole-body distribution of the therapeutic radionuclide. In reality, approximately 90% of the energy
radiopharmaceutical. associated with β− emission (as well as with β+ emission) is
• Therapeutic radiopharmaceuticals can be adminis- deposited within about 20% of the maximum tissue range
tered systemically by intravenous administration or reported for each radionuclide.
orally, intracavitarily, or intra-arterially. The path length of β− particles in tissues results in the so-
called crossfire effect, whereby cells that have not concen-
trated the radionuclide are exposed to potentially lethal
radiation from adjacent cells that concentrated the therapeu-
4.2 Beta-Emitting Radionuclides tic agent (Fig. 4.3). This effect can enhance the effectiveness
of the treatment, especially in tissues with a heterogeneous
Radionuclide therapy usually employs radionuclides emit- distribution in the target tissue.
ting β− particles. These negatively charged particles have a The radiopharmaceuticals currently employed in clinical
relatively low LET (0.2 keV/μm) and a continuous spec- practice for therapy are listed in Table 4.2, while the main
trum of energy. After their emission, the daughter nucleus radionuclides emitting β− with suitable energy for therapeu-
has one more proton and one less neutron than the original tic applications are described here below.
radionuclide. The radionuclides currently employed for
therapy have maximum path lengths in soft tissues (or
water-equivalent) that range from 2 to 12 mm, depending 4.2.1 Iodine-131 (131I)
on maximum energy of the particle (Fig. 4.1). Although
common terminology refers to a linear measure of path Iodine-131 is produced by fission of uranium-235 or by neu-
length, radioactive emission occurs always isotropically (in tron bombardment of stable tellurium in a nuclear reactor. It
all directions) and must therefore be considered in terms of has a half-life of 8.0 days and decays to stable xenon-131
a “sphere” around the emission point. Furthermore, the β− emitting, in addition to γ-rays, β− particles with maximum
particles do not travel in a straight path (as γ-rays and energy of 606 keV (247 keV with 1.8% abundance, 334 keV
X-rays do) but are continuously deviated in a random man- with 7.2% abundance, 606 keV with 89.7% abundance,
ner in all directions because of interactions with the sur- 806 keV with 0.7% abundance). The maximum path range is
rounding matter (Fig. 4.2). Therefore, from a 2.4 mm in tissues (mean 0.7 mm).
three-dimensional point of view, the path of an electron in
Fig. 4.1 Correlation between 12

maximum energy (MeV) and
maximum path in water/soft 90Y
tissue (mm) of the main 188
β− particle-emitting Re
10
radionuclides employed for
therapy in nuclear medicine
Maximum pathway in soft tissue (mm)
166Ho
32P
8
89Sr
186Re
4 198 149
Au Pm
153S
131I
117Lu
2 212Pb
67Cu 165Rh
0
0.0 0.5 1.0 1.5 2.0 2.5
Maximum Energy (MeV)
1000
500
Z coordinate (mm)
-500
-1000
1000
1000
0 500
0
Y coordinate (mm) -500
-1000 -1000 X coordinate (mm)
Fig. 4.2 Simulation of the tracks in water of β− particles emitted by a repeatedly from their initial direction. Most of the energy (about 90%)
point source of 131I (full energy spectrum); simulation has been per- associated with the emissions is deposited within about 20% of the
formed for only 100 particles, a number too limited to show the whole maximum path range reported for each radionuclide (image provided by
spectrum of possibilities (including the maximum range, which is courtesy of Prof. Alberto Del Guerra, Department of Physics
reached only by an extremely small fraction of the particles emitted). “E. Fermi”, University of Pisa and National Institute of Nuclear
Due to interactions with surrounding matter, the β− particles deviate Physics (INFN), Pisa Section, Pisa (Italy))
4.2.2 Yttrium-90 (90Y)
The radio-metal 90Y is produced by the generator system

90
Sr/90Y and decays to stable 90Zr. It has a half-life of 64 h and
emits high-energy β− particles (2.2 MeV) with a maximum
path length of 11–12 mm (mean 5.3 mm). The β− decay is
associated with pair production of β+ particles, although in a
minimal proportion (32 β+ particles per million decays); no
γ-emission occurs during decay of 90Y. 90Yttrium is coupled
to peptides and monoclonal antibodies through bifunctional
chelates, such as DOTA to produce a therapeutic
radiopharmaceutical.
There are currently three major clinical uses of 90Y: (1) to
label a monoclonal antibody for radioimmunotherapy of
lymphomas, (2) to bound to resin or glass microspheres for
Fig. 4.3 Schematic representation of the crossfire effect for particle- intra-arterial radio-embolization of liver tumors, and (3) to
emitting radionuclides employed for therapy (assumed to be β− parti- label somatostatin analogs for peptide receptor radionuclide
cles for this representation). Because of delivery of energy within a therapy of neuroendocrine neoplasms.
certain range from the point of emission, the radiotoxic effect is not
limited to the cells that specifically accumulate the radiopharmaceutical
but also involves nearby cells (either they be neoplastic or tumor cells)
4.2.3 Rhenium-188 (188Re) and Rhenium-186
(186Re)
In the chemical form of sodium 131I-iodide, this radio-
nuclide is extensively employed for treating benign and Rhenium-188 is a group VII metal with a half-life of 16.9 h that
malignant thyroid diseases. Other therapeutic applications decays to stable osmium-188 with emission of β− particles with
involve linking of 131I of more complex molecules, such as maximum energy of 2.12 MeV having maximum range of
meta-iodobenzylguanidine (MIBG) or monoclonal 11 mm in tissue (mean 3.5 mm). It also emits γ-rays with energy
antibodies. of 155 keV (15%), favorable for gamma camera imaging.
Table 4.2 Localization mechanisms and indications for the radiophar- (186Os) with the emission of β− particles with maximum
maceuticals most commonly used in clinical practice for radionuclide energy of 1.09 MeV and maximum path range of 4.5 mm in
therapy
tissue (mean 1.1 mm); it also emits a low abundance (9%) of
Localization
γ-ray suitable for gamma camera imaging (137 keV).
Radiopharmaceutical mechanism Indications
131
I-Iodide Transmembrane Treatment of
Both 186Re and 188Re have labeling chemistry properties
active transport: Na+/ malignant or benign similar to those of technetium; for this reason, these radionu-
I− symporter (NIS) thyroid disease clides offer the ability to utilize vast radiochemistry experi-
[131I]MIBG Accumulation in Treatment of neural ence accumulated with 99mTc.
intracellular crest-derived After conjugation with bone-seeking bisphosphonates,
neurosecretory neuroendocrine
vesicles tumors these two radionuclides are currently employed clinically for
ATPase-linked treatment of bone pain from skeletal metastases; 188Re-lipiodol
energy-dependent (a radiolabeled suspension of fatty acids) is used for treat-
process and ment of hepatocellular carcinoma.
pinocytosis
90
Y-Ibritumomab Antigen-antibody RIT of NHLs
tiuxetan binding (anti-CD20)
131
I-Tositumomab Antigen-antibody RIT of NHLs 4.2.4 Lutetium-177 (177Lu)
binding (anti-CD20)
90
Y- or Mechanical trapping Intra-arterial therapy Lutetium-177 is a radionuclide emitting β− and γ with a
166
Ho-Microspheres (intra-arterial of liver tumors physical half-life of 6.7 days. It decays to stable hafnium
micro-embolization)
(177Hf). The β− particles emitted by 177Lu have a maximum
131
I- or Mechanical trapping Intra-arterial therapy
188
Re-Lipiodol and pinocytosis of liver tumors energy of 497 keV (78.6%), 384 keV (9.1%), and 176 keV
177
Lu- or 90Y-SST Receptor/ligand Treatment of (12.2%). It also emits γ-rays with 113 keV (6.6%) and
analogs binding neuroendocrine 208 keV (11%) energies, suitable for gamma camera imag-
tumors ing. The maximum soft tissue penetration depth is about
32
P-Chloride Ion exchange Treatment of
2 mm (mean 0.67 mm). It is produced in a nuclear reactor
metastatic bone
disease with “direct” or “indirect” reactor production routes based
89
Sr-Chloride Ion exchange Treatment of on neutron irradiation of 176Lu or 176Yb, respectively [4].
metastatic bone Currently, this radionuclide is used for labeling soma-
disease tostatin analogs for peptide receptor radionuclide treatment
186
Re- or Chemisorption Treatment of
(PRRT) of neuroendocrine tumors (NETs). However, addi-
188
Re-HEDP metastatic bone
disease tional possible uses of 177Lu are growing, including conjuga-
153
Sm-EDTMP Chemisorption Treatment of tion with bisphosphonates (e.g., 177Lu-EDTMP) as
metastatic bone bone-seeking radiopharmaceuticals for treatment of bone
disease pain from skeletal metastases and conjugation with ligands
223
Ra-Chloride Ion exchange Treatment of
for the prostate-specific membrane antigen (PSMA) for ther-
metastatic bone
disease apy of inoperable prostate cancers.
RIT radioimmunotherapy, NHLs non-Hodgkin’s lymphomas, SST
somatostatin
4.2.5 Phosphorus-32 (32P)
Rhenium-188 can be produced with relatively high spe- Phosphorus-32, a pure emitter of β− particles, is produced in
cific activity by direct production in a nuclear reactor by irra- a nuclear reactor. It has a physical half-life of 14.3 days and
diation of enriched rhenium-187. For clinical purposes it is a maximum energy of 1.71 MeV (100%), resulting in a max-
produced in the no-carrier-added form using a generator sys- imum tissue range of 8 mm (mean 3 mm).
tem based on its parent radionuclide wolfram-188 (also Treatment with 32P in the chemical form of orthophosphate
known as tungsten-188), with a half-life of 69 days. The is indicated for palliation of bone pain from skeletal metasta-
188
W/188Re generator allows local availability of 188Re on ses, polycythemia vera, and essential thrombocythemia.
demand to label kits to produce therapeutic agents [3].
Rhenium-186 can be produced in low neutron flux nuclear
reactors (that are widely available throughout the world) by 4.2.6 Strontium-89 (89Sr)
direct neutron activation of enriched rhenium-185; its 90-h
half-life permits distribution also to sites distant from the Reactor-produced 89Sr is an alkaline earth metal radionuclide
production facility. Rhenium-186 decays to osmium-186 with a half-life of 50.5 days. It is an almost pure emitter of
β− particles with a maximum energy of 1.49 MeV (100%

abundance) and a very low γ-ray emission (0.01% abun- have not concentrated the radionuclide are exposed
dance) with a 0.91 MeV photopeak that precludes external to potentially lethal radiation from adjacent cells
imaging. The maximum path length of its β− particles in soft that concentrated the therapeutic agent.
tissues is 7.2 mm (mean 2.4 mm). • 131
I has a physical half-life of 8 days and decays
In the chemical form of 89SrCl2 (dichloride salt), its cur- emitting both β− particles with maximum path
rent clinical use is for treatment of bone pain from skeletal range of 2.4 mm in tissues and γ-rays suitable for
metastases. gamma camera imaging.
• 90
Y has a physical half-life of 64 h and decays emit-
ting β− particles with maximum path length of
4.2.7 Samarium-153 (153Sm) 11–12 mm in tissues; there is also a very low frac-
tion of decay associated with pair production of β+
This radionuclide is fission-produced by neutron capture particles (32 particles per million decays).
using a target enriched in 152Sm. It has a physical half-life of • 188
Re has a physical half-life of 16.9 h and decays
1.9 days and emits β− particles with a maximum energy of emitting both β− particles with maximum path
0.81 MeV (20%), 0.71 MeV (50%), and 0.64 MeV (30%). range of 11 mm in tissues and γ-rays suitable for
153
Sm also emits γ-rays with and energy of 103 keV (29% gamma camera imaging.
abundance), which are useful for gamma camera imaging. • 186
Re has a physical half-life of 90 h and decays
The maximum tissue range of the β− particles emitted by emitting both β− particles with maximum path
153
Sm is 3 mm (mean 0.5 mm). range of 4.5 mm in tissues and γ-rays suitable for
The prevalent use of 153Sm is in the chemical form of a gamma camera imaging.
radiolabeled bisphosphonate (153Sm-EDTMP), employed in • 177
Lu has a physical half-life of 6.7 days and decays
the clinical routine for treatment of bone pain from skeletal emitting both β− particles with maximum path
metastases. length of 2 mm in tissues and γ-rays suitable for
gamma camera imaging.
• 32
P has a physical half-life of 14.3 days and emits β−
4.2.8 Holmium-166 (166Ho) particles with maximum path range of 8 mm in tissues.
• 89
Sr has a physical half-life of 50.5 days and emits
This radionuclide is produced by neutron capture using a tar- β− particles with maximum path range of 7.2 mm in
get enriched in 165Ho oxide. It decays with a half-life of tissues.
26.8 h to erbium-166 emitting β− particles with maximum • 153
Sm has a physical half-life of 1.9 days and decays
energies of 1.84 MeV (48% abundance) and 1.77 MeV (48% emitting both β− particles with maximum path
abundance), associated with emission of γ-rays with energies range of 3 mm in tissues and γ-rays suitable for
of 81 keV (5.4% abundance) and 1379 keV (1.13% abun- gamma camera imaging.
dance). The maximum tissue range of the β− particles emit- • 166
Ho has a physical half-life of 26.8 h and decays
ted by 166Ho is 8.4 mm (mean 4 mm). emitting both β− particles with maximum path
The prevalent use of 166Ho is currently in the bound form range of 8.4 mm in tissues and γ-rays suitable for
to glass microspheres for transarterial radio-embolization of gamma camera imaging.
liver tumors.
4.3 adionuclides Emitting Alpha

R
Key Learning Points Particles
• The radionuclides emitting β− particles of suitable
energy most commonly utilized for therapy include Alpha particles are positive helium nuclei; their emission leads
I, Y, 177Lu, 32P, 89Sr, 153Sm, and 166Ho.
131 90
to a daughter nucleus with two fewer protons and two fewer
• The maximum path length in tissues of these radio- neutrons than the original radionuclide. These monoenergetic
nuclides is directly correlated with their maximum particles are usually emitted with high energy (5–9 MeV) and
energy and varies between 2 and 12 mm. produce a high density of ionization along their short linear
• For each radionuclide, approximately 90% of the tracks of 40–100 μm, corresponding to about five to ten cell
energy associated with β− emission is deposited diameters. The high therapeutic potential of these particles is
within about 20% of the maximum tissue range. due to their relatively large mass, based on which they have
• The path length of β− particles in tissues results in numerous interactions with nuclei of the cells they encounter
the so-called crossfire effect, whereby cells that along their path; this phenomenon results in high-LET values
of the α particles, typically in the 80–100 keV/μm range. Their
biologic effect is independent of the oxygenation state or cell 4.4 adionuclides Emitting Auger
R
cycle phase, as their interaction with the cell nucleus results in Electrons
single- and/or double-strand DNA breaks that are lethal for the
cell. The tumoricidal effect is maximal if the α particles are Auger electrons are low-energy orbital electrons emitted
emitted within the cell nucleus. after electron capture. Electron capture causes a vacancy,
Despite growing investigations on several α emitter radio- which is filled by electrons moving from an outer shell and
nuclides, there is currently only one approved radiopharma- thus initiating a cascade of electron transitions that shift the
ceutical available for clinical practice, radium-223 vacancy toward the outermost shell.
dichloride. Auger electrons have an extremely short path length in
tissues (2–500 nm), due to their very low energy (0.45–2 keV),
and LET ranging from 4 to 26 keV/μm. Localization of the
4.3.1 Radium-223 (223Ra) radionuclide within the cell nucleus is a strict requirement to
reach DNA (preferably within the DNA component of the
Radium-223 decays to stable lead-207 via a six-stage decay nucleus) and thus achieve cell killing; mitochondrial or cell
chain of short-lived daughter radionuclides, producing a surface localization does not produce significant cytotoxic
number of α-, β-, and γ-emissions with different energies and effects.
emission probabilities. The fraction of energy emitted from Thus, in principle the radiodosimetric burden would be
223
Ra and its daughters as α particles is 95.3% (energy range maximum to the target (tumor) tissue, provided that all cells
5.0–7.5 MeV). The fraction emitted as β− particles is 3.6% incorporate the radionuclide, and minimal in adjacent healthy
(average energies are 0.445 and 0.492 MeV), and the frac- tissues. Although theoretically very promising, treatment
tion emitted as γ-radiation is 1.1% (energy range of 0.01– with Auger electron-emitting radionuclides remains at the
1.27 MeV). It has a physical half-life of 11.4 days. moment only in a potential state, as it has not so far been
223
Ra can be produced efficiently in large amounts by a possible to transfer to humans the good results obtained in
227
Ac/223Ra neutron generator (half-life 21.8 years) that guar- experimental models (in vitro cell cultures or animal
antees availability of the radionuclide for long periods. 227Ac models).
is produced by neutron irradiation of natural 226Ra. Some of the radionuclides employed in diagnostic con-
The relatively long half-life of 223Ra provides sufficient ventional nuclear medicine (67Ga, 75Se, 111In, 125I, 123I, and
time for its preparation, distribution (including long distance 201
Tl) emit Auger electron, but no radiopharmaceutical is
shipment), and administration to patients. Its low irradiation currently employed in clinical practice for therapeutic pur-
is favorable from the point of view of handling, radiation poses. The only relevant clinical investigations based on
protection, and treatment on an outpatient basis. emission of Auger electrons have been carried out with the
Administered as a chloride salt, the calcium analog 223Ra use of 111In-DTPA-octreotide in patients with neuroendo-
has bone-seeking properties that make it useful for treating crine tumors. However, the pilot clinical investigations have
intractable bone pain from skeletal metastases. not proceeded any further after publication of the first reports
describing the superior therapeutic results obtained with
90
Y-labeled somatostatin analogs.
Key Learning Points
• High-energy α++ particles (5–9 MeV) produce a
high density of ionization along their short tracks of Key Learning Points
40–100 μm, corresponding to about five to ten cell • Auger electrons are low-energy orbital electrons
diameters. emitted after electron capture.
• The linear energy transfer associated with emission • The very low energy (0.45–2 keV) of Auger elec-
of α++ particles is much higher than that associated trons results in an extremely short path length in
with the emission of β− particles of similar energy. tissues (2–500 nm).
• The single- and/or double-strand DNA breaks • Localization of the radionuclide emitting Auger
caused by α++ particles are lethal for the cell, with a electrons within the cell nucleus is a strict require-
maximal tumoricidal effect when the particles are ment to reach DNA and thus achieve cell killing.
emitted within the cell nucleus. • Radionuclides emitting Auger electrons include
• Although the potential for antitumor efficacy of 67
Ga, 75Se, 111In, 125I, 123I, and 201Tl.
several radionuclides emitting α++ particles is being • Although potentially attractive for tumor therapy, it
explored, the only radionuclide currently used clini- has not so far been possible to transfer to humans
cally is 223Ra, which has a physical half-life of the promising results obtained with in vitro cell cul-
11.4 days. tures or animal models.
4.5 Radiopharmaceutical-Based Emission istration, sodium 131I-iodide is absorbed rapidly from the
of β− Particles upper gastrointestinal tract (90% in 60 min). The main excre-
tion route of iodine is through the kidneys, with renal clear-
4.5.1 Sodium 131I-Iodide ance equal to about 2–3% of renal blood flow in subjects
with normal renal function. Renal clearance of iodine is
Iodine-131 is commonly administered in the chemical form of reduced in hypothyroid subjects or in patients with impaired
sodium iodide for treating thyroid disorders. In a manner iden- renal function, while it is increased in hyperthyroidism. In
tical with that of native 127I− ions, the radioactive I− ions pres- euthyroid subjects with normal renal function, 50–75% of
ent in the circulation and in interstitial fluid are transported administered radioiodine is excreted in the urine within 48 h,
into the thyroid cells against a concentration gradient by an while about 10% of the administered activity is excreted
active transport mechanism (the Na+/I− symporter, NIS), through the fecal route, and there is a minor fraction of
located on the basolateral membrane of thyroid follicular cells. excretion with sweat and through the lungs with exhaled air
This active transport process concentrates iodide 20- to 40-fold (the latter being almost negligible). The biological effective
the plasma concentration under normal circumstances, a value half-life of radioiodine in plasma is about 12 h, whereas
that may in turn increase by an additional tenfold in the hyper- radioiodine concentrated in the thyroid gland is released with
functioning thyroid gland. Similarly as for other parameters of a half-life of about 6 days. Thus, after administration of
thyroid cell function and growth, the intracellular accumula- sodium 131I-iodide to an individual with normal thyroid func-
tion of iodine is regulated by serum TSH. Once inside the thy- tion, approximately 40% of the activity has an effective half-
roid cell, iodine is oxidized by the enzyme peroxidase; this life of 0.4 days, whereas the remaining 60% has an affective
starts the organification process that enables incorporation of half-life of 8 days.
iodine (either native or radioactive) into tyrosine to produce Although radiation safety regulations vary from country
molecules of monoiodotyrosine (MIT) and diiodotyrosine to country, radioiodine treatment for hyperthyroidism is gen-
(DIT) that are finally combined to constitute the two chemical erally performed on an outpatient basis, after adequate prep-
species of active thyroid hormones, thyroxine (T4) and triio- aration and dosimetric evaluation. Two methods are
dothyronine (T3). Thyroid hormones stored in the colloid con- commonly used to select the activity of 131I-iodide to be
tained inside the thyroid follicles are finally secreted into the administered: (1) administration of a fixed activity based on
circulation by pinocytosis, again following stimulation by clinical experience and (2) administration of a fixed radiation
TSH under physiologic conditions. dose on the basis of dosimetric calculations. The aim of
In addition to the concentration in the thyroid gland, radioiodine therapy in hyperthyroid patients is to induce per-
iodine accumulates in salivary glands, lacrimal glands, sweat manent euthyroidism or hypothyroidism.
glands, gastric glands, mammary gland, and the placenta. The most common indication for therapy with 131I-iodide
These physiological processes can induce unwanted side in patients with differentiated thyroid cancer derived from
effects when using 131I-iodide for high-dose therapy includ- the follicular epithelium (papillary and follicular carcinoma)
ing pain in the anterior cervical region (caused by inflamma- is for ablation of postsurgical remnants after thyroidectomy
tion induced by the cytotoxic action of radioiodine), in patients with intermediate- to high-risk cancer. In addi-
dysgeusia, anosmia, acute or chronic sialadenitis (sometimes tion, therapy with 131I-iodide is used for the treatment of
with persistent xerostomia), gastritis (with nausea and vom- locoregional recurrences and/or distant metastases, with
iting), and actinic cystitis. 131I uptake in the placenta and activities varying from 1.1 to 7.4 GBq [6]. For such high-
mammary gland would expose the fetus or the baby, respec- dose treatments, radiation safety regulations standing in sev-
tively, to an unacceptable radiation dose. eral countries require hospitalization in specially equipped
131
I-iodide is commonly used for treating both benign thy- rooms for adequate handling of radioactive wastes. Patients
roid diseases (hyperthyroidism caused by Graves’ disease or are discharged from the hospital when measurements of their
hyperfunctioning thyroid adenomas) and malignant thyroid radioactivity emissions fall below certain thresholds set
disease (differentiated thyroid carcinomas) [5]. according to local radiation safety regulations [7].
131
I-iodide is generally administered orally (capsules or As for all therapies with radionuclides, absolute contraindi-
liquid solutions), while the intravenous administration is cations to radioiodine treatment are pregnancy and lactation.
needed only in uncooperative patients or in case of uncon- Since the specific radioactivity of 131I-iodide is very high (typi-
trolled vomiting. To facilitate and accelerate intestinal cally >200 MBq/μg), the amount of iodine administered for
absorption, patients should fast for at least 6 h prior to and therapy with 131I-Iodide is considerably lower than the recom-
for 3 h after administration, and a low-iodine diet should be mended daily amount of iodine in a normal diet. Thus, allergy
undertaken for 2–3 weeks prior to therapy. to iodine is not considered a contraindication.
After i.v. injection, about 20% of circulating iodide is Approximately 4–7 days after administration of high-
extracted by the thyroid gland. The thyroid peak uptake/ dose 131I-iodide for therapy, whole-body scintigraphy is per-
accumulation occurs within 24–48 h of administration, with formed, based on the consistent γ-ray emission of 131I. The
about 50% of the maximum reached at 5 h. After oral admin- whole-body images may be supplemented by spot images
and SPECT/CT to visualize both physiologic and abnormal glioma, and neuroblastoma. Nonetheless, other neuroendo-
sites of radioiodine localization. These images provide infor- crine tumors such as medullary thyroid carcinoma,
mation for staging or restaging the disease and sometimes carcinoids, and gastro-entero-pancreatic (GEP) tumors can
depict previously unknown metastatic lesions. also be successfully treated, provided that diagnostic scans
The main radiation dosimetry estimates to patients fol- with 123I-MIBG show satisfactory uptake in the lesions [8].
lowing oral administration of 131I-iodide in individuals with a After slow i.v. infusion, the distribution pattern of
normal functioning thyroid gland (without thyroid-blocking 131
I-MIBG shows rapid initial uptake in the liver (33% of the
pretreatment) are reported here below, normalized to unit of administered activity), lungs (3%), myocardium (0.8%),
administered activity: spleen (0.6%), and salivary glands (0.4%). While uptake in
normally functioning adrenal medulla is so low that the adre-
• Effective dose 0.15 mSv/MBq nals cannot be visualized, hyperplastic adrenals show a high
• Tissues/organs with the highest values of absorbed dose: uptake of 131I-MIBG. About half of the administered activity
–– Thyroid 2.5 mGy/MBq is excreted in the urine within the first day, 70–90% of the
–– Gastric wall 0.12 mGy/MBq activity being cumulatively excreted within 2 days; the uri-
–– Urinary bladder wall 0.066 mGy/MBq nary metabolic breakdown products account for approxi-
mately 5–15% of the administered activity. About 3% of
MIBG excretion occurs via the gastrointestinal tract.
4.5.2 M
eta-[131I]Iodo-Benzyl-Guanidine Absolute contraindications to therapy with 131I-MIBG are
(131I-MIBG) pregnancy, lactation, and renal insufficiency requiring dialy-
sis. Relative contraindications include unmanageable urinary
This radiopharmaceutical (also known as 131I-iobenguane) incontinence, rapidly deteriorating renal function
was developed in the early 1980s to visualize tumors of the (GFR < 30 mL/min), progressive hematological and/or renal
adrenal medulla. It is a catecholamine analog (labeled 123I for toxicity because of prior treatments, and myelosuppression
diagnostic use) in which the iodinated benzyl group of brety- (total white cell count <3.0 × 109/L, platelets less than
lium is combined with the guanidine group of guanethidine 100 × 109/L).
(Fig. 4.4). Radioiodination is performed by the isotope The most important precaution in the preparation of
exchange method. patients for therapy with 131I-MIBG includes discontinuing
Due to its structural analogy with catecholamines, MIBG drugs interfering with MIBG uptake (such as labetalol, tricy-
is taken up by chromaffin cells via an active physiologic clic antidepressants, reserpine, and some sympathomimetic
uptake mechanism, the epinephrine transporter utilized for drugs) for an appropriate period before therapeutic
accumulating noradrenaline in neurosecretory granules. administration of 131I-MIBG.
Following depolarization induced by high transmembrane Free radioiodine produced during in vivo degradation of
flux of calcium ions, these molecules are then secreted 131
I-MIBG can cause accumulation of radioactivity in the
through an exocytosis mechanism. This uptake/accumula- thyroid and the gastrointestinal tract. Thyroid blocking using
tion process is abundantly expressed in the sympathetic gan- potassium or sodium iodide (KI or NaI) or Lugol’s solution
glia, the adrenal medulla, and in all tissues with high is necessary, generally starting 1 or 2 days prior to treatment
adrenergic innervation (myocardium, salivary glands). When and continuing for up to 14 days after treatment.
the drug is administered in high concentrations for therapy, Administration of 131I-MIBG must be performed very
passive diffusion also becomes important. slowly (over 1–4 h after dilution in 50 mL of saline), because
The therapeutic applications of 131I-MIBG include tumors the therapy dose levels involve injecting a sizable mass
of the sympathoadrenal system originating embryologically amount of a catecholamine analog that may induce the sud-
from the neural crest, such as pheochromocytoma, paragan- den release into circulation of catecholamines stored in neu-
rosecretory granules, with possible pharmacologic effects
such as arterial hypertension, tachycardia, nausea, and vom-
iting; nevertheless, the likelihood of such an occurrence is
NH
believed to be extremely low. For the same reason, the spe-
NH
cific radioactivity of 131I-MIBG for therapy should be higher
NH2 than that for diagnostic purposes (up to 1.8 GBq/mg), with
the aim of reducing the incidence of pharmacologic effects;
131I continuous monitoring of the patient’s vital signs during
administration is anyway recommended.
Possible side effects include nausea and vomiting within
the first 24 h after administration and temporary suppression
Fig. 4.4 Chemical formula (left) and 3D structure (right) of radioio-
dinated MIBG. Color codes: yellow, 131I; white, H; light blue, C; dark of the bone marrow (manifesting mostly with reduced platelet
blue, N counts) and deterioration of renal function. Possible long-
term effects are hypothyroidism (after inadequate thyroid on the surface of malignant cells, but not in normal cells; fur-
blockade) and persistent hematological effects (thrombocyto- thermore, it should form a stable complex with the antibody.
penia, myelosuppression). There is also sparse evidence for Another important issue concerns the origin and structure
induction of leukemia or secondary solid tumors, but this of the antibodies employed for RIT; many monoclonal anti-
occurrence is very rare and has been reported especially in bodies were in fact initially developed from murine cells,
conjunction with (long-standing) chemotherapy [9]. thus being intrinsically heterologous proteins with
There is currently no consensus on the optimal dosing immunogenic activity when administered to humans. This
strategy. Even if “fixed” therapeutic activities are possible, immunogenicity is much lower for chimeric antibodies
administration of individually tailored activities based on pre- (made up of components for 60% human and 40% murine)
therapy dosimetric estimates is to be preferred. The activity or for the humanized proteins (with 95% of human compo-
administered for each treatment and the interval(s) between nents, while generally only the binding site for the antigen is
possible multiple administrations (repeated treatments can be murine) developed through genetic engineering techniques.
considered at 6–8-month intervals) are mainly determined by In fact, the formation in the patient receiving these prepara-
hematological radiotoxicity and by tumor biology. tions of human anti-mouse antibodies (HAMA), human anti-
The main radiation dosimetry estimates to patients fol- chimeric antibodies (HACA), or human antihuman antibodies
lowing i.v. administration of 131I-MIBG are reported here (HAHA) can alter the pharmacokinetic profiles of the radio-
below, normalized to unit of administered activity: labeled monoclonal antibody or predispose to allergic reac-
tions upon repeat administration. The monoclonal products
• Effective dose 0.2 mSv/MBq currently used for the preparation of radiopharmaceuticals
• Tissues/organs with the highest values of absorbed dose: are characterized by greatly reduced immunogenicity com-
–– Liver 0.83 mGy/MBq pared to their predecessors.
–– Urinary bladder wall 0.59 mGy/MBq A final important issue that is crucial for the success of
–– Spleen 0.49 mGy/MBq RIT (especially in the case of solid tumors) has to do with
molecular size/weight of the radiolabeled preparation. In this
regard, intact immunoglobulin molecules are generally too
Key Learning Points large (about 150–160 kDa for IgG antibodies produced by
• Radiopharmaceuticals labeled with 131I commonly hybridomas) to diffuse efficiently throughout a tumor mass.
utilized for therapy include 131I-iodide and Such poor diffusion causes insufficient targeting of the radio-
131
I-MIBG. labeled antibodies to tumor cells located even a few cell
• Due to selective accumulation in normal thyroid diameters away from blood capillaries and constitutes a limi-
cells and in thyroid-derived tumor cells, 131I-iodide tation to the therapeutic potential of RIT.
is employed for treating patients with hyperthyroid- To overcome these limitations and reduce immunogenic-
ism or patients with differentiated thyroid carci- ity, smaller immunoglobulin fragments with much lower
noma derived from the follicular epithelium. molecular weight have been produced. A common step in
• As a structural analog of catecholamines, 131I-MIBG this process is enzymatic cleavage to remove the Fc frag-
is used for therapy primarily in patients with some ment, which is responsible for nonspecific binding of the
neuroendocrine tumors originating embryologi- intact molecule with immune cells (including macrophages,
cally from the neural crest, such as pheochromocy- dendritic cells, NK cells, B cells) causing high background
toma, paraganglioma, and neuroblastoma. uptake in the reticuloendothelial system throughout the body.
Pepsin digestion results in the production of the bivalent
F(ab′)2 fragment (molecular weight about 110 kDa), where
the two antigen-binding sites are still connected by the hinge
4.5.3 Radiopharmaceuticals region. Two smaller monovalent Fab fragments (each about
for Radioimmunotherapy 50 kDa in molecular weight and containing only one antigen-
binding site) are produced by papain digestion. Even smaller
Radioimmunotherapy (RIT) is based on the use of radiola- immunoreactive fragments, the single-domain binding site
beled monoclonal antibodies (or antibody fragments) target- with molecular weight between 12 and 15 kDa, can be pro-
ing epitopes of antigens expressed specifically on the surface duced using molecular engineering techniques.
of cancer cells. The binding with tumor-specific antigens Despite several ongoing clinical investigations, no radio-
deposits ionizing radiation within the neoplastic tissue caus- pharmaceuticals have been approved yet for RIT of solid
ing cytotoxicity. tumors, whereas two radiopharmaceuticals are currently
The ideal tumor-associated antigen to be targeted with this used for RIT in some hematologic diseases, as described
form of therapy should be densely and uniformly expressed hereinafter.
4.5.3.1 90 Y-Ibritumomab Tiuxetan diagnostic-level activity of the surrogate radiopharmaceuti-

90
Y-Ibritumomab tiuxetan is a murine monoclonal antibody cal 111In-ibritumomab tiuxetan. The early images (up to
that binds to lymphocytes expressing the CD20 antigen. 90Y about 24 h) typically show activity in the blood pool, with
is chelated to tiuxetan, which is covalently linked to the anti- significant accumulation in the liver and spleen, while accu-
body via the arginine and lysine residues contained in the mulation in the lung and bone is generally low.
antibody. This radiopharmaceutical has been approved for Radiodosimetric estimates have demonstrated an up to 850-
the treatment of low-grade or follicular B-cell NHL that has fold greater radiation dose to the tumor lesions than to nor-
relapsed during or after treatment with other anticancer mal organs.
drugs and newly diagnosed follicular NHL following a The main radiation dosimetry estimates to patients fol-
response to initial anticancer therapy. lowing i.v. administration of 90Y-ibritumomab tiuxetan are
The basis for therapy with 90Y-ibritumomab tiuxetan is reported here below, normalized to unit of administered
that the CD20 antigen is expressed on the surface of normal activity:
B lymphocytes (except pre-B cells and secretory B cells) but
especially on the surface of >90% of B-cell lymphomas. The • Tissues/organs with the highest values of absorbed dose:
CD20 antigen does not shed from the cell surface and does –– Spleen 7.35 mGy/MBq
not internalize upon antibody binding. The β− emission from –– Liver 4.32 mGy/MBq
90
Y induces cellular damage through the formation of free –– Lungs 2.05 mGy/MBq
radicals, not only in cells that have been directly targeted by
the radiolabeled antibody, but also in neighboring cells—due
to the crossfire effect. Key Learning Points
90
Y-Ibritumomab tiuxetan administration must be pre- • Radioimmunotherapy is based on the use of radio-
ceded by unlabeled monoclonal antibody (rituximab), as part labeled monoclonal antibodies or antibody frag-
of 90Y-ibritumomab tiuxetan regimen, in order to block CD20 ments targeting epitopes of antigens expressed on
antigens expressed on normal B cells. The unlabeled mono- the surface of cancer cells.
clonal antibody per se constitutes an approved pharmaceuti- • Chimeric antibodies and humanized proteins devel-
cal for immunotherapy regimens of NHLs, because of its oped through genetic engineering techniques mini-
intrinsic tumor cell-killing capability. mize the complications possibly arising in patients
90
Y-Ibritumomab tiuxetan is administered as a slow i.v. following repeat administration of these agents, such
infusion over 10 min; the agent has a half-life in plasma of as the formation of human anti-mouse antibodies
28–30 h. Over 7 days, a median of 7.2% of the injected activ- (HAMA), human anti-chimeric antibodies (HACA),
ity is excreted in the urine. or human antihuman antibodies (HAHA).
The administered activity is generally 15 MBq/kg body • 90Y-Ibritumomab tiuxetan, a murine monoclonal
weight with a maximum limit of 1.2 GBq (32 mCi). In antibody that binds to the CD20 antigen expressed
patients with mild thrombocytopenia (platelet count between on lymphocytes, is used for radioimmunotherapy of
100,000 and 149,000 per mm3), the administered activity relapsed or refractory lymphomas expressing the
should be reduced to 11 MBq/kg. Dose-limiting toxicity for CD20 antigen.
90
Y-ibritumomab tiuxetan is bone marrow suppression, which • 131I-Tositumomab, a murine monoclonal antibody
is reversible; bone marrow toxicity results in possible throm- that binds to a different epitope of the CD20 antigen
bocytopenia and neutropenia, especially in patient with base- expressed on lymphocytes, has similar indications
line platelet count between 100,000 and 149,000/mm3. The as 90Y-ibritumomab tiuxetan for radioimmunother-
median time to nadir is 7–9 weeks, and the median duration apy of lymphomas expressing the CD20 antigen;
of cytopenia is 3–5 weeks. 90Y-Ibritumomab tiuxetan should however, this radiopharmaceutical is no longer
not be administered to patients with >25% of bone marrow commercially available.
involved by lymphoma and/or impaired bone marrow reserve
because of, e.g., prior myeloablative therapies. Additional
contraindications include platelet count <100,000/mm3, neu-
trophil count <1500/mm3, hypocellular bone marrow (<15% 4.5.3.2 131
I-Tositumomab
cellularity or marked reduction in bone marrow precursors), 131
I-Tositumomab is a murine monoclonal antibody labeled
or history of failed stem cell collection [10]. with 131I that had been approved in the United States in 2003,
Since 90Y does not emit γ-rays and scintigraphic visual- with indications similar to those of 90Y-ibritumomab tiuxetan
ization of the secondary Bremsstrahlung X-rays is rather in patients with non-Hodgkin’s lymphomas. 131I-Tositumomab
poor, biodistribution can be evaluated by administering a is specifically directed against a different epitope of the same
CD20 surface antigen, and the reported response rate is More recently lipiodol labeled with 188Re using hexadecyl-
about 70%. However, it has recently withdrawn from the tetramethyl-diaza-decanethiol (HDD) as the chelating agent
market because of declining sales. [11] has shown promising results for radio-embolization in
As in the case of 90Y-ibritumomab, infusion of unlabeled patients with inoperable large and/or multifocal
“cold” antibody (tositumomab) was recommended before hepatocarcinomas.
administration of the radiolabeled antibody. 188
Re has potentially favorable physical characteristics,
Dosimetric estimations to determine the whole-body such as a shorter half-life than 131I (16.9 h versus 8 days), a
radiation-absorbed dose are possible on the basis of three β− emission of high energy (2.1 MeV) with ensuing good
sequential whole-body acquisitions after the administration cell-killing effect, and an associated 155 keV γ-emission
of a diagnostic activity (185 MBq) of 131I-tositumomab (with favorable for gamma camera imaging (for the purpose of
prior infusion of 450 mg of unlabeled tositumomab). dosimetric estimates). The 188Re-labeled agent represents
therefore an excellent alternative to the older 131I-labeled
agent.
4.5.4 131
I-Lipiodol and 188Re-HDD-Lipiodol
Transarterial chemoembolization with drug-loaded lipiodol, 4.5.5 90Y- and 166Ho-Microspheres

a particulate suspension of a mixture of iodized esters of
poppy seed oil fatty acid, is still employed for treatment of Resin or glass microspheres containing 90Y or 166Ho are used
inoperable hepatocellular carcinoma. The rationale for using for transarterial radio-embolization (TARE) of liver tumors.
transarterial lipiodol (which is performed under angiographic This therapeutic procedure, also defined as selective internal
monitoring) is its entrapment and embolization in the radiation therapy (SIRT), is performed through selective
microvessels of the tumor, based on the preferential blood catheterization of the main or of segmental branches of the
flow as well as retentions by pinocytosis in both the endothe- hepatic artery under angiographic monitoring. Radio-
lial cells and tumor cells. embolization is one of the several therapeutic options to
Most of the clinical experience with lipiodol radio- treat inoperable hyper-vascular lesions in the liver due to
embolization has been acquired with 131I-lipiodol, although primary hepatocellular carcinoma, cholangiocarcinoma, or
the most recent trials have been based on the use of metastatic tumors originating at other sites. Other therapeu-
188
Re-lipiodol. tic options include liver transplantation; systemic chemo-
After transarterial administration, more than 75% of the therapy; locoregional treatments, such as percutaneous
injected 131I-lipiodol remains in the liver, while the remainder ablative techniques (radio-frequency ablation, laser coagu-
distributes mainly to the lungs. Thyroidal uptake of free lation, cryotherapy, percutaneous ethanol injection); and
radioiodine released from 131I-lipiodol must be prevented by transarterial chemoembolization.
thyroid-blocking medication (saturated potassium iodide or The rationale for SIRT relies on radio-microembolization
Lugol’s solution) starting 2–3 days before treatment and con- of the hyper-vascular liver tumors, which receive their blood
tinuing for 10–15 days. Tumor/non-tumor uptake ratios in the supply primarily from the hepatic artery, whereas the normal
liver are generally >5, and more than 10% of the injected liver parenchyma receives blood primarily via the portal
radioactivity remains within the tumor with an effective half- vein. Thus, microspheres injected into the hepatic artery
life of about 6 days, longer than normal liver tissue. (most commonly into the right or left hepatic artery or even
Administered activity can be either a fixed amount of more selectively into the proper segmental artery) are trapped
2.4 GBq or defined on the basis of patient-specific dosimet- within the tumor microvasculature and deliver radiation
ric estimates. Due to the long half-life of 131I-lipiodol in the selectively to the tumor liver lesion. The two types of
tumor, current legislation in some countries requires hospi- 90
Y-microspheres currently available commercially (resin
talization for about 1 week, for radiation protection and glass microspheres) are biocompatible but not biode-
purposes. gradable nor metabolized.
Treatment is in general well tolerated and serious adverse 90
Y-resin microspheres are made of an acrylic polymer in
effects are very rare, while generic asthenia is commonly which 90Y is bound to the carboxylic group of the polymer
reported; likewise, hematologic toxicity is exceptionally after production of the microspheres. The spheres have a
rare, although blood cell counts may be altered due to the diameter between 20 and 60 μm and a specific radioactivity
cirrhosis-related hypersplenism. Interstitial pneumopathy of 50 MBq per sphere. There are about 40–80 million micro-
due to trapping and retention of the radiolabeled particle sus- spheres per vial, with a total activity of 3 GBq that can be
pension is reported as the main risk of this treatment. The aliquoted in the radiopharmacy to select the desired amount
highest values of absorbed dose are reported in the liver of 90Y activity. Because of their elevated number, resin
tumor, liver parenchyma, and lung. microspheres used for this treatment may have a moderate
embolic effect and must be delivered at a slow rate (no more about 37 μm) consisting of a poly(l-lactic acid) matrix con-
than 5 mL/min) to avoid reflux into the hepatic artery and taining 166Ho. In addition to its emission of β− particles
into other organs. The shelf life of the device is 24 h, which (with maximum energies of 1.74 and 1.85 MeV), this radio-
restricts clinical flexibility and patient scheduling. nuclide (that decays with a physical half-life of 26.8 h) also
In the 90Y-glass microspheres, 89Y is embedded in the emits low-energy γ-rays (81 keV) that are suitable for
glass matrix, to be subsequently activated to 90Y in a nuclear imaging with a conventional gamma camera. Therefore,
reactor just before shipment to the nuclear medicine center. pretreatment evaluation of tumor perfusion is possible by
Their medium diameter is 20–30 μm, and the specific activ- injecting intra-
arterially a scout activity of
ity is 2500 Bq per sphere. The number of 90Y-glass spheres 166
Ho-microspheres rather than 99mTc-MAA for preliminary
required for treatment ranges between 1.2 and 8 million, and SPECT/CT imaging. Furthermore, holmium has strong
the preparation is supplied in six sizes of activities: 3, 5, 7, paramagnetic properties, which make it possible to acquire
10, 15, and 20 GBq. The embolic effect linked to treatment MR imaging of the intrahepatic distribution of the
with the glass 90Y-microspheres is limited because a lower 166
Ho-microspheres injected intra-arterially. Although the
number of microspheres is injected intra-arterially. The shelf clinical experience acquired with this new preparation is
life of the glass spheres is 15 days from the date of calibra- limited, good disease control has been demonstrated in
tion; therefore, no physical manipulation is required to pre- about 70–75% of the patients submitted to therapy with
pare the desired, patient-based activity for administration, 166
Ho-microspheres. Different levels of activities adminis-
which is obtained simply by letting a certain activity to decay tered to patients have been investigated, ranging from 1260
to the desired level. to 5040 MBq per kg of body weight. In addition to their use
For both microsphere preparations, the biodistribution of for treatment of liver tumors, 166Ho-microspheres are also
90
Y-microspheres is usually predicted through prior intra- being investigated for other locoregional intra- arterial
arterial 99mTc-MAA administration and subsequent planar treatment, e.g., in patients with inoperable squamous cell
and SPECT imaging (possibly SPECT/CT). The images are head and neck cancers.
analyzed to estimate extrahepatic shunting to the lungs or the
gastrointestinal tract, as well as to evaluate the absorbed
doses to the target and the nontarget tissues [12].
Adequacy of 99mTc-MAA scintigraphy for simulation of Key Learning Points
therapy with the 90Y-microspheres is widely accepted, • Selective/superselective intra-arterial administra-
although not definitely proven; in fact, some discrepancies tion of particles that remain trapped at tumor sites
have been observed between the pretreatment 99mTc-MAA because of local microembolization is the basis for
scan and posttreatment imaging obtained either with a con- therapy of either primary or metastatic liver cancer
ventional gamma camera (based on the X-ray emission lesions.
linked to the Bremsstrahlung effect induced by the high- • Lipiodol, a particulate suspension of a mixture of
energy β− particles emitted by 90Y) or using a PET scanner iodized esters of poppy seed oil fatty acid, which
(based on the extremely low—but still imageable—emission can be loaded with either 131I or 188Re is used for
of β+ particles, about 32 positrons per million decays). Such intra-arterial therapy of hepatocellular carcinoma.
discrepancies are possibly due to the different nature of • The most commonly employed form of intra-arterial
99m
Tc-MAA versus the 90Y-microspheres or to occasional radio-embolization for treatment of either primary or
alterations of the original vascular anatomy produced by metastatic liver cancers is based on particles
angiography. Nevertheless, the acquisition of both pre- and (microspheres) containing either 90Y or 166Ho.
post-therapeutic images represents a suitable source of infor- • Microspheres containing 90Y can be made either of
mation for risk/benefit evaluation. glass or of a resin.
Three main methods are suggested by the manufacturer to
calculate the activity of resin 90Y-microspheres to be admin-
istered: (1) empirical method, (2) a body surface area (BSA)-
based method based on the body surface area (BSA), and (3) 4.5.6 90Y- or 177Lu-Labeled Somatostatin
the partition method. A more sophisticated evaluation can be Analogs
made using the voxel dosimetry approach or by Monte Carlo
modeling that provides information on dose distribution and Somatostatin (SST) is an oligopeptide comprising either 14
expected radiobiologic effects at the voxel level. or 28 amino acids produced endogenously mainly in the cen-
More recently, a new preparation of radiolabeled micro- tral nervous system; its recognized physiologic functions
spheres for TARE of liver tumors has become commer- include regulation of the secretion of various hormones, e.g.,
cially available, the 166Ho-microspheres (average diameter insulin and glucagon. Receptors for SST are overexpressed
in most neuroendocrine tumors; five different types of SST Therapy with the radiolabeled SST analogs is adminis-
receptors (transmembrane molecule weighing approximately tered by slow i.v. infusion over approximately 30 min.
80 kDa) have been identified, named SSTR1-5 (type 2 being Standard clinical protocols of PRRT consist of the systemic
divided into subtypes 2a and 2b). Native SST has a very short administration of a radiolabeled somatostatin analog, frac-
biological half-life in blood (3 min), limiting its use as a tionated in sequential cycles (usually 4–5), 6–9 weeks apart,
pharmacologic agent. Several stabilized SST analogs with until the intended total amount of radioactivity has been
longer half-lives have been developed, to treat patients with administered. The total activity of 90Y-DOTA-SST analog
neuroendocrine tumors expressing SST receptors (mostly to generally does not exceed 3.7 GBq/m2 body surface area,
palliate symptoms). Some of these analogs have been modi- while a flat activity of 7.4 GBq is considered for therapy with
fied by introducing a chelating agent to radiolabel the SST the 177Lu-DOTA-SST analogs.
analog (e.g., 111In-DTPA-octreotide, 68Ga-DOTA-TOC, The radiolabeled peptides are cleared through the kid-
68
Ga-DOTA-NOC, 68Ga-DOTA-TATE) (see Chaps. 2 and 3 neys, where they are reabsorbed and partially retained in the
of this book). The DOTA chelator, which allows more stable proximal tubules. Nephrotoxicity is the dose-limiting factor.
binding of metallic radionuclides emitting β− particles with Co-administration of negatively charged amino acids
energy suitable for therapy, has been selected for the produc- (l-lysine and/or l-arginine) competitively inhibits the proxi-
tion of SST analogs to be used as radiopharmaceuticals for mal tubular reabsorption of the radiopeptides leading to a
therapy. significant reduction of the radiation burden to the kidney.
The radiolabeled SST analogs that are currently used in Therefore, the standard protocol for therapy with either
clinical practice with the most promising results for therapy 177
Lu- or 90Y-labeled DOTA-SST analogs includes prior i.v.
include 90Y-DOTA-TOC and 177Lu-DOTA-TATE. Receptor- infusion of l-lysine and l-arginine (50–75 g each). Besides
mediated internalization and intracellular retention of the long-term renal toxicity, the other critical target is the bone
radio-peptide allow the so-called peptide receptor radionu- marrow; reduced bone marrow reserve may occur, and, very
clide therapy (PRRT) [13]. infrequently (1–2%), myelodysplastic syndrome and occa-
DOTA-TOC is a derivatized SST analog, abbreviated sionally leukemia have been sporadically reported.
form of [DOTA0,Tyr3]-octreotide (a modified octreotide), After injection, 177Lu-DOTA-TATE is rapidly cleared
where DOTA stands for the bifunctional chelating agent from the blood. In fact, at 4 h after administration,
1,4,7,10-tetraazacyclo-dodecane-1,4,7,10-tetraacetic acid, a 177
Lu-DOTA-TATE is concentrated in the kidneys, tumor
macrocyclic chelator that allows radiolabeling and formation lesions, liver and spleen, and in some patients in the pituitary
of stable complexes with both metallic isotopes, 90Y and gland and in the thyroid.
177
Lu. This peptide shows a high affinity for SSTR2 but lower General contraindications are pregnancy and kidney
affinities for SSTR5 and SSTR3. failure with creatinine clearance <30 mL/min. Candidates
DOTA-TATE is also a derivatized SST analog, abbrevi- for PRRT are basically all the SST receptor-expressing
ated form of [DOTA0,Tyr3,Thr8]-octreotide or [DOTA0,Tyr3]- tumors or metastatic/inoperable NETs. SSTR2-expressing
octreotate. This peptide has a six- to ninefold greater affinity tumors include GEP and bronchial neuroendocrine tumors
for SSTR2 than DOTA-TOC, but has no affinity for either but may also include patients with pheochromocytoma,
SSTR5 or SSTR3. paraganglioma, neuroblastoma, or medullary thyroid
Advantages of 177Lu-DOTA-TATE over 90Y-DOTA-TATE carcinoma.
include the emission of an associated γ-radiation, which Pre-therapy scintigraphy with 111In-pentetreotide is cur-
makes it possible to acquire conventional gamma camera rently the most validated method to assess for the presence of
scintigraphic imaging for radiation dosimetry estimates; fur- SSTR overexpression. Nevertheless, PET imaging with
thermore, the shorter tissue range of the β− particles emitted 68
Ga-DOTA-TOC or 68Ga-DOTA-TATE is now becoming the
by 177Lu allows a higher radiation dose be delivered to new pre-therapy standard evaluation, since these radiophar-
smaller tumors. The use of 177Lu also results in lower neph- maceuticals mimic in vivo patterns of biodistribution of their
rotoxicity of the PRRT regimens versus 90Y-labeled SST ana- 90
Y- or 177Lu-labeled counterparts used for PRRT.
logs (see further below). On the other hand, the use of
90
Y-labeled SST analogs (which emits β− particles with
higher energies and longer tissue ranges) would be more Key Learning Points
suitable for the treatment of larger tumors, in case of bulky • Somatostatin analogs used for therapy of neuroen-
and/or heterogeneous masses. Recently, a combination docrine tumors expressing somatostatin receptors
radionuclide therapy using 177Lu- and 90Y-labeled peptides can be labeled with either 90Y or 177Lu.
has been proposed in order to improve efficacy in patients • The metallic nature of these radionuclides requires
with tumors of various sizes and inhomogeneous receptor the use of a suitable bifunctional chelating agent
distribution.
Treatment with this radionuclide is therefore indicated in

(1,4,7,10-tetraazacyclo-dodecane-1,4,7,10- patients with multiple painful osteoblastic skeletal metasta-
tetraacetic acid, or DOTA) for radiolabeling of the ses exhibiting high uptake of 99mTc-labeled bisphosphonates
somatostatin analogs used for therapy. confirmed on a conventional bone scan prior to therapy.
• Different somatostatins with varying binding affin- Following i.v. administration of 32P-orthophosphate, 85%
ity for the somatostatin receptors have been devel- of the injected activity accumulates in the hydroxyapatite
oped, including octreotide and octreotate. crystals as inorganic phosphate, while the remainder local-
• Nephrotoxicity, the main side effect of therapy with izes in soft tissues, including active red bone marrow. In soft
radiolabeled somatostatin analogs, is minimized by tissues, 32P is predominantly distributed intracellularly, espe-
pre-administration of the negatively charged amino cially in the cytoplasm and cell nucleus, thus inducing non-
acids l-lysine and/or l-arginine that competitively negligible DNA damage. For this reason and also taking into
inhibit the proximal tubular reabsorption of the account the rather long beta β− particle path range in soft
radiopeptides and thus significantly reduce the radi- tissue, administration of this radiopharmaceutical causes a
ation burden to the kidney. significant radiation burden to the bone marrow, with possi-
• Neuroendocrine tumors candidates for the so-called ble hematopoietic side effects (leukopenia and thrombocyto-
peptide receptor radionuclide therapy (PRRT) are penia) reaching a nadir about 4–5 weeks after injection and
basically all the somatostatin receptor-expressing recovering at 6–7 weeks. Hematologic toxicity limits the use
tumors, including gastro-entero-pancreatic and of 32P-orthophosphate to those circumstances where other
bronchial neuroendocrine tumors; occasionally, less toxic bone-seeking radiopharmaceuticals are not
also some pheochromocytomas, paragangliomas, available.
neuroblastomas, or medullary thyroid carcinomas The therapeutic administered activity is usually 185–
express somatostatin receptors with high density. 370 MBq intravenously (often in divided doses) or 370–
444 MBq orally. Clearance occurs predominantly by renal
excretion.
Bone pain palliation is usually noticed within 14 days
4.5.7 32P-Orthophosphate after administration, with a range of 2 days to 4 weeks.
The response rate to therapy with 32P-orthophosphate is
This radiopharmaceutical was the first radiolabeled agent about 80%, and the mean duration of response is
used for the treatment of metastatic bone pain [14] and was 5 ± 2.5 months.
the most widely used agent until the 1980s. It should be mentioned that therapy with 32P-orthophosphate
Bone metastases are a frequent complication of advanced has also been successfully employed for treatment of polycy-
stages of neoplastic disease. Bone metastases occur in themia vera, based on the relatively high radiation burden
about 70% of patients with prostate cancer or breast cancer delivered to the active red bone marrow.
and in 30–50% of those with cancers of the lung, bladder, The main radiation dosimetry estimates to patients fol-
and thyroid. The major complications associated with skel- lowing administration of 32P-orthophosphate are reported
etal metastases are pain (initially mild to medium intensity here below, normalized to unit of administered activity:
and then increasing and becoming severe—requiring the
use of opioid medication for palliation), hypercalcemia, • Tissues/organs with the highest values of absorbed dose:
radicular compression, and/or spinal cord and pathological –– Bone surface 10 mGy/MBq
fractures. –– Red bone marrow 7.6 mGy/MBq
In patients with metastatic involvement of multiple skeletal –– Lower bowel wall 0.01 mGy/MBq
segments (therefore not amenable to external radiotherapy
and/or surgery), therapy with therapeutic bone-seeking radio-
pharmaceuticals constitutes an important strategy for pallia- 4.5.8 89Sr-Chloride (89SrCl2)
tion of bone pain. Therapeutic agents include
32
P-orthophosphate (rarely used today), 89Sr-chloride (infre- The use of the bone-seeking radiopharmaceutical
quently used today), 153Sm-EDTMP, and 186Re-HEDP, recently 89
Sr-chloride is indicated for palliation of bone pain in
the α-particle emitter 223Ra-dichloride (see further below). patients with painful skeletal metastases.
Accumulation of 32P in the bone mineral matrix mirrors Upon administration, 89SrCl2 quickly clears from the
the degree of active osteogenesis, thus being enhanced pri- blood; the 89Sr2+ ions follow the same in vivo metabolic fate
marily in areas of metastatic involvement (predominantly as calcium, with rapid incorporation into the inorganic min-
osteoblastic or mixed metastasis) with significantly greater eral matrix of the bone. Thus, the radiopharmaceutical selec-
concentrations than in the surrounding normal bone. tively localizes in bone mineral, with at least 50% of the
injected activity localized in the skeleton within 2–3 h post- uptake. The main excretory route is through the kidneys;
administration. About 30–35% of the administered activity almost 35% of the injected tracer is excreted in the urine by
remains in normal bone for 10–14 days postinjection, 6 h postinjection. The recommended administered activity is
whereas, retention in osteoblastic areas is as high as 85–90% 37 MBq/kg. Concomitant γ-emission of 153Sm allows gamma
at 3 months after administration. camera imaging to confirm satisfactory targeting of the
About two-thirds of the fraction of 89Sr not bound to the lesions.
bone is excreted in the urine, while about one-third is The onset of pain relief is generally noticed within
excreted through the fecal route. Urinary excretion is great- 2 weeks after radiopharmaceutical administration, with a
est in the first 2 days following administration. 89Sr-Chloride mean duration of pain relief of 3–4 months, but it may last up
is administered by slow i.v. injection, over about 1–2 min. to 11 months.
The recommended activity of 89SrCl2 is 150 MBq or Further relief of pain is achieved with repeat treat-
alternatively 1.5–2.2 MBq/kg body weight. The major side ments. Since the critical organ regarding radiation dosim-
effect is myelosuppression that may occur particularly with etry is the bone marrow, a minimum interval of 8 weeks
higher administered activity, reducing the platelet and leu- should be allowed between subsequent administrations in
kocyte counts by almost 25–30%. For this reason, patients order to allow for adequate recovery of bone marrow
should have a minimum platelet count of 60,000/mm3 and a function. However, hematological side effects are gener-
leukocyte count of 2400/mm3 at the time of ally mild.
administration. Transient exacerbation of the preexisting pain (flare phe-
Initial relief of pain is usually noticed within 3 days after nomenon) after 153Sm-EDTMP is reported by 12–20% of
administration, but it may be as late as 25 days. Mean dura- patients.
tion of pain relief is about 3–6 months; therefore, re- The main radiation dosimetry estimates to patients fol-
treatments with 89Sr may be considered every 3–6 months. lowing i.v. administration of 153Sm-EDTMP are reported
Complete remission of pain is experienced by 5–20% of the here below, normalized to unit of administered activity [15]:
patients after 89Sr treatment, while almost 80% of the patients
experience some relief of pain from osteoblastic metastasis. • Tissues/organs with the highest values of absorbed dose:
In 10% of the patients, there is an initial transient worsening –– Bone surface 6.8 mGy/MBq
of bone pain within 3 days of therapy that subsides in about –– Red bone marrow 1.5 mGy/MBq
a week (flare phenomenon). –– Urinary bladder wall 1 mGy/MBq
lowing administration of 89SrCl2 are reported here below,
normalized to unit of administered activity [15]: 4.5.10 186/188Re-Etidronate (186/188Re-HEDP)
• Tissues/organs with the highest values of absorbed dose: The bone-seeking agent hydro-ethylidene-di-phosphonate
–– Bone surface 17 mGy/MBq (etidronate, or HEDP) binds to hydroxyapatite crystals by
–– Red bone marrow 11 mGy/MBq forming hydroxide bridges, through the mechanism of che-
–– Lower bowel wall 4.7 mGy/MBq misorption (Fig. 4.5). HEDP can be radiolabeled with either
188
Re or 186Re, both radionuclides being suitable for thera-
peutic applications.
4.5.9 153Sm-Lexidronam (153Sm-EDTMP) Due to binding with plasma proteins, 186Re-HEDP is
cleared from the circulation with a longer plasma half-life
This bone-seeking radiopharmaceutical is indicated for relief than most of the other bone pain agents (41 ± 6 h). The mean
of bone pain in patients with confirmed osteoblastic metastatic effective half-life of labeled HEDP in the bone is about 16 h.
bone lesions that enhance on the radionuclide bone scan. Approximately 70% of the administered 186Re-HEDP activ-
153
Sm-EDTMP (Fig. 4.5) has biodistribution properties ity is excreted in the urine within 24 h in patients without
similar to those of other radiolabeled bisphosphonates, as it bone metastases. The standard activity administered for
accumulates avidly on the surface of newly formed hydroxy- therapy is 1295 MBq, given by i.v. injection. Pain relief
apatite crystals through a chemisorption mechanism. occurs early after administration.
Following i.v. administration, 153Sm-EDTMP clears rap- Although the typical tumor/bone marrow absorbed dose
idly from the blood with a half-life of 5.5 min; less than 1% ratio is 14:1, there may be some transient myelosuppression,
of the administered activity remains in the circulation 5 h which is important in patients with reduced bone marrow
postinjection. Up to 55–60% of injected activity remains sta- reserve. Typical marrow recovery times range from 4 to 6 weeks.
bly bound to the skeleton and accumulates in osteoblastic 188
Re-HEDP has similar biodistribution as 186Re-HEDP,
metastatic lesions in a 5:1 ratio compared with normal bone with rapid excretion in urine (60% ± 12% of administered
Fig. 4.5 3D structure of

radiolabeled bone-seeking
bisphosphonates employed
for palliation therapy of bone
pain from skeletal metastasis:
186
Re-HEDP (left) and
153
Sm-EDTMP (right). Color
codes: green, 186Re; yellow,
153
Sm; red, O; white, H;
fuchsia, P; light blue, C; dark
blue, N
186Re-HEDP 153Sm-EDTMP
activity in 48 h. The fraction of the dose excreted in the urine 223
RaCl2 is a calcium analog, with a biodistribution simi-
varies with renal function and extent of osteoblastic lar to 89SrCl2. It selectively accumulates in the bone, with
metastases). increased uptake in areas of bone metastases, by forming
The main radiation dosimetry estimates to patients fol- complexes with the bone mineral hydroxyapatite. The high
lowing i.v. administration of 188Re-HEDP are reported here LET of its α-emission (80 keV/μm) leads to a high frequency
below, normalized to unit of administered activity [15]: of double-strand DNA breaks in adjacent cells, resulting in a
localized antitumor effect on bone metastases. The α++
• Tissues/organs with the highest values of absorbed dose particle range from 223Ra is less than ten cell diameters (about
–– Bone surface 1.4 mGy/MBq 100 μm), which minimizes damage to the surrounding nor-
–– Red bone marrow 1.3 mGy/MBq mal tissue.
–– Lower bowel wall 0.57 mGy/MBq 223
Ra-Dichloride is administered over 1 min by i.v. injec-
tion; it then clears rapidly from the blood and distributes pri-
marily into the bone and bone metastases. Fifteen minutes
Key Learning Points after injection, about 20% of the administered activity
• Radiopharmaceuticals labeled with β− emitters used remains in the blood, while only <1% is still in the circulation
for therapy of skeletal metastases (mainly with the at 24 h. No specific accumulation is seen in other normal
purpose of bone pain palliation) include organs such as the heart, liver, kidney, gallbladder, stomach,
32
P-ortophosphate, 89Sr-chloride, 153Sm-EDTMP, and spleen.
188
Re-HEDP, and 186Re-HEDP. A major difference between 223Ra-dichloride and other
• All such bone-seeking radiopharmaceuticals local- bone-seeking radiopharmaceuticals is excretion occurs
ize at metastatic sites characterized by significant through the gastrointestinal tract, which causes some gastroin-
osteoblastic reaction, similarly as the diagnostic testinal toxicity. At 48 h postinjection, the cumulative fecal
bone-seeking tracers used for imaging. excretion is about 13%, with a range of 0–34%. Only 5% is
• The mechanism of accumulation at the metastatic excreted in the urine. However the rate of elimination of
site is based on chemisorption at the surface of the 223
Ra-dichloride from the gastrointestinal tract is influenced by
newly formed mineral component of the bone. the high variability in intestinal transit rates among patients.
There is little myelotoxicity and no long-term toxicity. In gen-
eral, most adverse events are mild to moderate in severity and
include nausea, bone pain, diarrhea, and fatigue. No dose-
4.6 Alpha-Emitting limiting hematologic toxicity has been observed in any
Radiopharmaceuticals patients.
The dose of 223Ra-dichloride is 55 kBq/kg of body weight.
4.6.1 223Ra-Dichloride A treatment cycle includes six administrations at 4-week
intervals, and it has been recognized that therapy with
This α++ emitting agent (trade name Xofigo) is indicated for 223
RaCl2 induces not only palliation of bone pain but also a
the treatment of patients with castration-resistant prostate significant benefit in survival.
cancer with symptomatic bone metastases and no known vis-
ceral metastatic disease.
The main radiation dosimetry estimates to patients fol- therapy for relapsed neuroblastoma. J Pediatr Hematol Oncol.
2003;25:543–7.
lowing i.v. administration of 223Ra-dichloride are reported 10. Tennvall J, Fischer M, Bischof Delaloye A, Bombardieri E, Bodei
here below, normalized to unit of administered activity: L, Giammarile F, et al. Dosimetry Committee, EANM procedure
guideline for radio-immunotherapy for B-cell lymphoma with
90
Y-radiolabelled ibritumomab tiuxetan (Zevalin). Eur J Nucl Med
Mol Imaging. 2007;34:616–22.
–– Osteogenic cells 1.152 mGy/MBq 11. Lee YS, Jeong JM, Kim YJ, Chung JW, Park JH, Suh YG, et al.
–– Red bone marrow 0.139 mGy/MBq Synthesis of 188Re-labelled long chain alkyl diaminedithiol for ther-
–– Lower large bowel wall 0.047 mGy/MBq apy of liver cancer. Nucl Med Commun. 2002;23:237–42.
12. Giammarile F, Bodei L, Chiesa C, Flux G, Forrer F, Kraeber-Bodere
F, et al. EANM procedure guideline for the treatment of liver cancer
and liver metastases with intra-arterial radioactive compounds. Eur
J Nucl Med Mol Imaging. 2011;38:1393–406.
Key Learning Points 13. Bodei L, Mueller-Brand J, Baum RP, Pavel ME, Hörsch D,

• Ra-Dichloride is the only approved α-emitting
223
O’Dorisio MS, et al. The joint IAEA, EANM, and SNMMI prac-
radiopharmaceutical. tical guidance on peptide receptor radionuclide therapy (PRRNT)
• 223Ra accumulates at metastatic sites because of its in neuroendocrine tumours. Eur J Nucl Med Mol Imaging.
2013;40:800–16.
chemical analogy with calcium. 14. Friedell HL, Storaasli JP. The use of radioactive phosphorus in the
• At variance with the most commonly used agents treatment of carcinoma of the breast with widespread metastases to
for treatment of bone metastases emitting β− parti- bone. Am J Roentgenol Radium Ther. 1950;64:559–75.
cles (that are excreted mainly through the kidneys), 15. Bodei L, Lam M, Chiesa C, Flux G, Brans B, Chiti A, et al. EANM
procedure guideline for treatment of refractory metastatic bone
223
Ra has a significant fecal excretion, which results pain. Eur J Nucl Med Mol Imaging. 2008;35:1934–40.
in some gastrointestinal toxicity.
Suggested Readings
References Eckelman WC, Boyd M, Mairs RJ. Principles of molecular targeting

for radionuclide therapy. In: Strauss HW, Mariani G, Volterrani D,
Larson SM, editors. Nuclear oncology – from pathophysiology to
1. Baum RP, editor. Therapeutic nuclear medicine. New York, NY:
clinical applications. New York, NY: Springer; 2017. p. 35–66.
Springer; 2014.
Ell PJ, Gambhir SS, editors. Nuclear medicine in clinical diagnosis and
2. Salvatori M, Cremonesi M, Indovina L, Chianelli M, Pacilio
treatment. 3rd ed. New York, NY: Churchill Livingston; 2004.
M, Chiesa C, et al. Radiobiology and radiation dosimetry in
Guidoccio F, Mazzarri S, Orsini F, Erba PA, Mariani G. Novel radio-
nuclear medicine. In: Strauss HW, Mariani G, Volterrani D,
pharmaceuticals for therapy. In: Strauss HW, Mariani G, Volterrani
Larson SM, editors. Nuclear oncology – from pathophysiol-
D, Larson SM, editors. Nuclear oncology – from pathophysiology
ogy to clinical applications. New York, NY: Springer; 2017.
to clinical applications. New York, NY: Springer; 2017. p. 173–98.
p. 35–66.
Herbert JC, Eckelman WC, Neumann RD, editors. Nuclear medicine –
3. Pillai MR, Dash A, Knapp FF Jr. Rhenium-188: availability from
diagnosis and therapy. New York, NY: Thieme Medical Publishers;
the 188W/188Re generator and status of current applications. Curr
1996.
Radiopharm. 2012;5:228–43.
IAEA. Good practice for introducing radiopharmaceuticals for clinical
4. Dash A, Pillai MR, Knapp FF Jr. Production of 177Lu for targeted
use. Vienna: International Atomic Energy Agency (IAEA); 2015.
radionuclide therapy: available options. Nucl Med Mol Imag.
IAEA. Operational guidance on hospital radiopharmacy. Vienna:
2015;49:85–107.
International Atomic Energy Agency (IAEA); 2008.
5. Seidlin SM, Marinelli LD, Oshry E. Radioactive iodine therapy;
ICRP. Radiation dose to patients from radiopharmaceuticals. ICRP
effect on functioning metastases of adenocarcinoma of the thyroid.
publication 53. Ann ICRP. 1988;18(1-4):149.
JAMA. 1946;132:838–47.
Kowalsky RJ, Falen SW, editors. Radiopharmaceuticals in nuclear
6. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ,
pharmacy and nuclear medicine. 3rd ed. Washington, DC: American
Nikiforov YE, et al. American Thyroid Association management
Pharmacists Association; 2011.
guidelines for adult patients with thyroid nodules and differentiated
Orsini F, Puta E, Lorenzoni A, Erba PA, Mariani G. Radiopharmaceuticals
thyroid cancer: the American Thyroid Association guidelines task
for therapy. In: Strauss HW, Mariani G, Volterrani D, Larson SM,
force on thyroid nodules and differentiated thyroid cancer. Thyroid.
editors. Nuclear oncology – from pathophysiology to clinical appli-
2016;26:1–133.
cations. New York, NY: Springer; 2017. p. 99–114.
7. Silberstein EB, Alavi A, Balon HR, Clarke SE, Divgi C, Gelfand
Owunwanne A, Patel M, Sadek S, editors. The handbook of radiophar-
MJ, et al. The SNMMI practice guideline for therapy of thyroid
maceuticals. New York, NY: Springer; 1995.
disease with 131I 3.0. J Nucl Med. 2012;53:1633–51.
Theobald T, editor. Sampson’s textbook of radiopharmacy. 4th ed.
8. Giammarile F, Chiti A, Lassmann M, Brans B, Flux G. EANM pro-
London: Pharmaceutical Press; 2010.
cedure guidelines for 131I-meta-iodobenzylguanidine (131I-mIBG)
Welch MJ, Redvanly CS, editors. Handbook of radiopharmaceuticals:
therapy. Eur J Nucl Med Mol Imaging. 2008;35:1039–47.
radiochemistry and applications. Hoboken, NJ: Wiley; 2003.
9. Weiss B, Vora A, Huberty J, Hawkins RA, Matthay KK. Secondary
myelodysplastic syndrome and leukemia following 131I-MIBG
Methods and Instrumentation
for Measuring Radioactivity 5
Maria Evelina Fantacci
Contents
5.1 Physical Quantities 118
5.1.1 Radiometric Quantities 118
5.1.2 Quantities Related to the Radiation-Matter Interactions 119
5.1.3 Dosimetric Quantities 120
5.1.4 Relationships Between Dosimetric Quantities 121
5.1.5 Measurement of Dosimetric Quantities 122
5.2 Instrumentation 123
5.2.1 Physical Principles of Ionization Detectors 123
5.2.2 Physical Principles of Scintillation Detectors 124
5.2.3 Physical Principles of Semiconductor Detectors 126
5.3 Detectors for Dosimetry 128
5.3.1 Ionization Detectors for Dosimetry 128
5.3.2 Film Dosimeters 130
5.3.3 Thermoluminescent Dosimeters (TLDs) 130
5.3.4 Probe for Uptake Measurements 131
5.3.5 Probes for Intraoperative Use 132
5.3.6 Well Counters 133
5.3.7 Liquid Scintillation Beta Counters 133
5.3.8 Dose Calibrators 133
5.3.9 Hand-Foot Contamination Monitors 134
Further Reading 135
Learning Objectives by charged secondary particles produced by indirectly

• Understand the fundamental principles underlying the ionizing radiation outside of that volume.
definition of quantities for radioactivity measurements. • Describe the relationships between the dosimetric quanti-
• Define subdivision of radioactivity measurements into ties in conditions of charged particle equilibrium.
radiometric quantities (radiation fields—source or field • Understand the main principles underlying operation of
quantities), quantities which describe the interactions radiation detectors: ionization detectors, semiconductor
between radiation and matter, and dosimetric quantities detectors, and scintillation (both organic and inorganic)
(which describe the energy transfer processes from radia- detectors.
tion to matter and the ability of radiation to ionize the • Understand the basic principles of detectors used for per-
atoms or molecules of the crossed medium). sonal dosimetry (film dosimeters and thermoluminescent
• Understand how, in conditions of charged particle equi- dosimeters).
librium, the energy dissipated outside a small spherical • Understand the basic principles of instruments for
volume is compensated by the energy dissipated therein measuring radioactivity based on ionization detectors
(free-air ionization chamber, cavity ionization cham-
ber, Geiger-Müller counter, dosimetric pen, and dose
M. E. Fantacci (*) calibrators).
Department of Physics “Enrico Fermi”, University of Pisa, • Understand the basic principles of instruments for mea-
Pisa, Italy suring radioactivity based on scintillation detectors and
e-mail: maria.evelina.fantacci@unipi.it

118 M. E. Fantacci
semiconductor detectors (hand-foot contamination radiodiagnostic). In the case of particle accelerators, the
monitors, well counters, liquid scintillation beta counters, parameters most directly concerned with radioprotection
probes for uptake measurements, and probes for intraop- are the energy of the accelerated particles and the average
erative use). power of the beam. In the case of X-ray tubes, the parame-
ters most directly c oncerned with radioprotection are the
potential difference applied to the tube, the filtrations, the
5.1 Physical Quantities current, the time, the half-value layer and the power. We
remind that:
We firstly introduce the physical quantities that are used to
describe the characteristics of the radiation fields and the • The potential difference (measured in kV) applied to the
characteristics of irradiated materials relative to their interac- tube determines the maximum energy of X-rays
tions with ionizing radiations. produced.
• Filtering changes the spectral distribution of the beam
(hardening, characteristic lines).
5.1.1 Radiometric Quantities • The anode current (measured in mA) determines, with the
time (s), the exposure (measured in mAs, i.e., mA × s),
Radiometric quantities describe the radiation fields. They related to the number of photons emitted.
can be subdivided in quantities related to the radioactive • The half-value layer (SEV) is the thickness (expressed in
sources and quantities related to the radiation fields. The mm) of a filter at which the intensity of radiation entering
source-related quantities are the quantities that are used to it is reduced by one half. SEV is conventionally expressed
describe and measure the characteristics of a radioactive in mm of Al for potential differences applied to the tube
source useful for radiation protection purposes. As radiation up to 120 kV and in mm of Cu for higher potential
sources we consider both radioactive sources (substances, differences.
preparations, or devices containing one or more radionu- • The power is the product of the potential difference
clides) and machines as ionizing radiation generators (parti- applied and the anode current. It is expressed in mA kV or
cle accelerators, X-ray tubes). in kW.
5.1.1.1 Source Quantities 5.1.1.2 Field Quantities

Since radioactive decays can be very complex, it is prefera- Intuitively, a radiation field can be described by counting the
ble to specify the intensity of the radioactive sources through number of particles present point by point.
the number of disintegrations per unit of time rather than by Particle fluence is defined as Φ1:
the number of particles or photons emitted. The activity of a
dN
source so is defined as: Φ=
dA
( m−2 ) (5.3)

dN
A=− 1Bq ( Becquerel ) = 1s −1 (5.1) where dN is the number of particles passing through the
dt
maximum section dA of an infinitesimal sphere having the
where dN is the number of nuclear decays occurring in the center at the point considered.
amount of radionuclide considered in the time interval dt. Particle fluence rate is defined as ϕ:
The activity does not provide information on the amount
of radioactive material present. We then define the specific d Φ d2 N
activity As, measured in Bq g−1, which represents the number
φ=
dt
=
dA dt
( m−2 × s−1 ) (5.4)

of disintegrations per unit of time per mass unit.
where d2N is the number of particles crossing the maximum
N 0 = Avogadro number section dA of an infinitesimal sphere, having the center at the
λN
As = M A = atomic mass ( g ) (5.2) point considered, within the time interval dt.
N0 M A Sometimes we can need to know the total energy deliv-
λ = decay constant
ered by the particles instead of their number. Then the parti-
The radiation energy emitted in a disintegration can be cle energy fluence is defined as Ψ:
derived from the decay scheme or measured by spectrom-
etry. The radiation spectrum is modified by the material 1
In the next definitions, we will use differential notation because every
interposed between source and the detector. Therefore, for definition must be applied, even in case of nonuniform radiation fields,
the sealed sources it is also necessary to know the filtration where fluence varies from point to point. Because of the statistical
nature of the radiation fields, the variables they are dealing with are
characteristics of the sealant. In the case of radiogenic
always random. The number of particles N is also random, whose dif-
machines, we distinguish between particle accelerators (for ferential dN is meant as the differential of the expected average number
radiotherapy or nuclear medicine) and X-ray tubes (for of particles.
5 Methods and Instrumentation for Measuring Radioactivity 119
dR We emphasize that the transfer of energy to matter by

Ψ=
dA
( J × m−2 ) (5.5) photons (which are indirectly ionizing particles) takes place

in two successive phases:
where dR is the sum of the radiant energies of all the parti-
cles passing through the maximum section dA of an infini- • The first step is the energy transfer to the secondary
tesimal sphere having the center at the point considered. charged particles of the medium.
Correspondingly, the energy fluence rate is defined as ψ: • The second phase consists in the dissipation of energy by
the secondary charged particles through the collisions
d Ψ d2 R they undergo in the matter (energy transfer to the medium).
ψ =
dt
=
dAdt
( J × m−2 × s−1 ) (5.6)
The properties of the material in relation to these two dis-
where d2R is the sum of the energies of all the particles pass- tinct phases can be represented by the introduction of appro-
ing through the maximum section dA of an infinitesimal priate coefficients of interaction. The first of these coefficients
sphere having the center at the point considered in the time is the μtr/ρ mass energy transfer coefficient, defined as:
interval dt.
µ tr 1 dEtr
ρ
=
ρ EN dl
( m2 × kg −1 ) (5.9)

5.1.2 Q
uantities Related to the Radiation- where dEtr is the energy of the N incident particles, each of
Matter Interactions the energy E, indirectly ionizing (of total initial energy
EN), transferred as kinetic energy of the secondary parti-
In order to assess the energy deposited in a certain region of cles due to the interactions in the length dl in a medium of
an irradiated medium, it is necessary to know, in addition to density ρ.
the quantities that characterize the radiation field, also the In calculating the mass energy transfer coefficient, all
quantities that specify the properties of the medium in rela- energy losses that do not reappear in the form of kinetic
tion to the interactions between radiation and matter. In par- energy of the secondary charged particles must be excluded.
ticular, it is necessary to subdivide these quantities into When we are interested in knowing the energy actually
quantities that express the likelihood of interact by photons stored in a certain volume element, it is necessary to use the
and quantities describing the characteristics of energy trans- μen/ρ mass energy absorption coefficient:
fer to the medium by charged particles.
µen µ tr
= (1 − g ) ( m 2 × kg −1 ) (5.10)
5.1.2.1 X- and Gamma Rays ρ ρ
Consider a length dl of a medium of density ρ, and suppose where g is the fraction of energy that the secondary charged
that a dN/N fraction of photons undergoes interaction. particles dissipate as the bremsstrahlung radiation in the
A mass attenuation coefficient μ/ρ of material of density material of density ρ.
ρ is defined for each fixed energy of the photons:
5.1.2.2 Charged Particles
µ 1 dN
=
ρ ρ N dl
( m 2 × kg −1 )

(5.7) Charged particles lose energy crossing biological matter,
especially by ionizing and excitating the electrons of the
The mass attenuation coefficient is the coefficient in the medium (energy losses by collision). For radiation protec-
exponential, from which it is possible to calculate the number tion purposes, we will consider only energy losses in the case
of particles that did not interact in the thickness l of the mate- of light charged particles (electrons and positrons) with rela-
rial considered, since N0 is the number of incident photons: tivistic energies. To describe the interactions of charged par-
µ
ticles with matter, some physical quantities are used, the
−  ρl
N = N0e ρ
(5.8) most important of which is the mass stopping power, defined
as:
With low-energy photons, we can neglect nuclear interac-
S 1 dE
tions, so the main processes to consider are the photoelectric =
ρ ρ dl
( J × m2 × kg −1 ) (5.11)
effect, the Compton effect, coherent scattering, and pairs
production. where S is the linear stopping power (defined as the expecta-
The mass attenuation coefficient is related to the total num- tion value of the rate of energy loss per unit path length (dE/
ber of interactions that the photons undergo in the medium. dx) of the charged particle) and dE is the energy lost by a
We remind that not all of these interactions cause a complete charged particle in the length dl in a medium of density ρ.
transfer of energy from the photon to the matter contained in Both collision losses and irradiation losses contribute to
the volume element considered. Being interested in local mass stopping power, and therefore it can be expressed as the
energy losses, we need to consider other coefficients. sum of these two contributions. Mass stopping power is not
120 M. E. Fantacci
enough to explain the effects of ionizing radiation on living moc2, of all the directly and indirectly ionizing particles leav-
matter. In fact, the secondary charged particles can have suf- ing the volume considered), and ΣQ is the energy expended
ficient kinetic energy to make traces distinct from that of the to increase the mass of the system.
primary charged particle (δ) and to produce ionization at a The energy imparted is a stochastic magnitude and is sub-
distance from this. It is then necessary to consider the spatial ject to random fluctuations that can be very large if it is
distribution of the energy transferred along the traces of related to particularly small volume elements and small
charged particles. To take this into account, the LET (linear number of ionizing particles. In dosimetry, however, its aver-
energy transfer) has been introduced: age value 〈E〉 is considered.
The absorbed dose, in a volume element of mass dm, is
 dE  defined:
 ( J × m or keV × µ m )
−1 −1
L∆ =  (5.12)
 dl  ∆ d E
D=
dm
( Gy = Gray, 1 Gy = 1J × kg −1 ) (5.14)
where dE is the energy dissipated from the initial charge par-
ticle in the length dl in collisions involving energy transfers where d〈E〉 is the average cumulative energy in a volume
lower than a predetermined Δ (usually expressed in eV). element of mass dm.
Only those transfers will be considered as energy locally The intensity or rate of absorbed dose is defined as:
transferred to the medium.
dD
D′ =
dt
( Gy × s−1 ) (5.15)

5.1.3 Dosimetric Quantities The absorbed dose plays a major role in evaluating
radiation-induced effects, although it is not sufficient to give
The dosimetric quantities describe the various phases of a complete quantitative interpretation. In particular, it is not
energy transfer processes from radiation to matter and the able to take into account the diversity of the effects induced,
ability of radiation to ionize the atoms or molecules of the with the same dose absorbed, by the different incident radia-
crossed medium. They can be expressed as the product of a tions in the different irradiated tissues and organs. For this
radiometric quantity for a constant characteristic of the reason, new quantities more suitable to express the likelihood
medium or for a quantity characteristic of the interactions of of the adverse effects of exposure to ionizing radiation have
the medium with the radiation. been introduced. These quantities are the equivalent dose H,
The nature of the energy transfer processes is discrete, and the which takes into account the different quality of the radiation,
values of the quantities used to describe them are subject to sta- and the effective dose E, which also takes into account the
tistically significant fluctuations, especially in case of small vol- different radiosensitivity of the different tissues and organs.
umes of matter or small number of particles. In this case, physical The equivalent dose is related to the dose absorbed by the
quantities defined in statistical terms, i.e., stochastic quantities, relationship:
must be used. The value of a non-stochastic quantity, once cer-
tain conditions are fixed, can be calculated, while in the case of H = DWr ( Sv = Sievert, 1Sv = 1J × kg −1 ) (5.16)

stochastic size, only the probability of each particular value can
be evaluated on the basis of a probability distribution. where Wr is the radiation weighting factor.
Microdosimetry uses stochastic quantities, defined in The radiation weighting factor represents the relative bio-
finite domains, that assume values that are discontinuous in logical effectiveness of the radiation and modifies the
space and time. absorbed dose to take into account the different biological
In radioprotection dosimetry non-stochastic quantities are effects of various types and energies of radiation at a micro-
used, which are continuous and differentiable functions of scopic level. It is identified according to the LET in water of
space and time. the radiation considered.
It is defined as the energy imparted by radiation in a cer- The effective dose is related to the dose equivalent by the
tain volume of the quantity: relationship:
E = Rin − Rout + ΣQ ( J ) (5.13) E = HWt ( Sv = Sievert, 1Sv = 1J × kg −1 ) (5.17)

where Rin is the radiant energy incident in the considered vol- where Wt is the tissue weighting factor.
ume (sum of energies, excluding the mass energy moc2, of all Overall, we can get the total effective dose to the body as:
directly and indirectly ionizing particles entering the volume
considered), Rout is the radiant energy leaving the volume E = ∑H i wi (5.18)

considered (sum of energies, excluding the mass energy i
The weighting factors for the various tissues are com- where dQ is the absolute value of the total charge of the ions
puted and published by the International Commission on of the same sign air-generated when all the electrons (posi-
Radiological Protection (ICRP). tive and negative) released from the photons in a volume ele-
As has already been observed, the transfer of the energy ment of dm mass have been completely stopped in the air.
to the medium takes place essentially in two successive The old measuring unit for the exposure was the roentgen:
phases. The first concerns the set in motion of the secondary 1 R = 2.58 × 10−4 C kg−1.
charged particles by the primary radiation. The second It should be noted that, in the definition of exposure, the
concerns the dissipation of energy by these secondary
ionization produced by the absorption of the bremsstrahlung
charged particles through the collisions they undergo in the radiation emitted by the secondary electrons released into
medium. Both processes should be considered for the calcu- the volume of interest should not be considered for the calcu-
lation of the absorbed dose. However, important dosimetric lation of dQ. Apart from this difference, which becomes sig-
information can be deduced also neglecting the study of the nificant only at high energies, exposure coincides with the
dissipation processes into the medium by the secondary ionization equivalent to kerma in air.
charged particles and only describing the phase of energy We can also define the intensity or rate of exposure:
transfer to secondary charged particles. This can be done by
dX
introducing the kerma k, whose definition is: X′=
dt
( C × kg −1 × s−1 ) (5.23)

dEtr
K= ( Gy ) (5.19)
dm
where dEtr is the sum of the initial kinetic energies of all 5.1.4 R
elationships Between Dosimetric
charged particles produced by indirectly ionizing particles in Quantities
a volume element of a specified material of mass dm, includ-
ing the energy that the charged secondary particles irradiate It is often useful to establish relationships between the dosi-
in the form of bremsstrahlung radiation. metric quantities, especially for practical purposes of mea-
The intensity or rate of kerma is also defined: surement, as some of them are difficult to measure, while in
special conditions they may be related to other more readily
dK
K′ =
dt
( Gy × s−1 ) (5.20) easily measurable quantities and can therefore be obtained.
Relatively simple relationships between some fundamen-
We can refer to the kerma (or kerma intensity) value in a tal dosimetric quantities can be established when, at the ref-
given material both in the air and in a point of another mate- erence point of a medium material irradiated with indirect
rial. This value will be obtained by placing a small amount of ionizing radiation, the conditions of “charged particle equi-
the specified material at the point of interest. If the two mate- librium” (CPE) occur. These conditions are intended to be
rials are different, the sample introduced must have a mass realized when the energy dissipated outside a small spherical
small enough to not cause appreciable disturbances to the volume, centered at the reference point, by secondary parti-
indirect radiation ionizing incident field. cles charged therein produced by indirectly ionizing radia-
Knowing the particle energy fluence at a point of a tion, is compensated by the energy dissipated therein by
medium whose mass attenuation coefficient is known, the charged secondary particles produced by indirectly ionizing
kerma can be determined. radiation outside of that volume. To illustrate the concept of
We remember that: equilibrium charged particles, consider the situations illus-
trated in Fig. 5.1. Suppose we have a mass m in a given vol-
dEtr dR µ tr 1 dEtr
K= ( 5.19 ) Ψ = ( 5.5) = ( 5.9 ) ume, and suppose we want to measure the absorbed dose at
dm dA ρ ρ EN dl an internal point in that volume. Suppose now that three indi-
From which rectly ionizing radiations have created three secondary elec-
trons in the points marked with an asterisk. It’s only the
µ tr dl µ tr
K= ρ EN ⋅ = Ψ (5.21) energy lost by these electrons on the way in the volume that
ρ dm ρ
contributes to determining the absorbed dose. In the case of
Another quantity widely used in dosimetry is the expo- Fig. 5.1b (electrons produced within the mass m and that
sure. It is the oldest of the dosimetric quantities introduced to complete their path within the mass m), the energy dis-
describe the ability of electromagnetic radiation to produce charged by radiation coincides with that deposited in the
ionization in the air. It is defined as: medium. In the case of Fig. 5.1c (electrons produced within
the mass of m which dissipate their energy partly within the
dQ
X =
dm
( C × kg −1 ) (5.22) mass and partly on the outside), the energy released by
the radiation is greater than that deposited in the medium. On
122 M. E. Fantacci
a 1 following relationship between the absorbed dose in the air

e– (Da) and exposure occur, i.e.:
* e–
* 2 D = K ( Gy ) (5.24)
m
e–
* Wa
3
Da = X ( Gy ) (5.25)
b e
where Wa is the mean energy needed to create an ion pair in
* e– the air (J), e is the charge of the electron (1.6 × 10−19 C), and
* e– m X is the exposure (C kg−1).
* e–
c 5.1.5 Measurement of Dosimetric Quantities
* Consider a uniform medium immersed in an ionizing radia-

* tion field, and suppose we want to measure the absorbed dose
m at a point of reference of that medium. To do this, we should
* intuitively place a cavity around the point considered and
introduce a dose-sensitive material, that is, a “dosimeter.” As
d we will see in paragraph 5.3, to have a dosimeter, it is suffi-
cient to have a detector whose response is a known function
* of the energy deposited by the ionizing particles. It is also
* necessary to consider that the introduction of the dosimeter in
* m the cavity perturbs the pre-existing field. Therefore, it is pos-
sible to establish a relationship between the value of the mea-
Fig. 5.1 Examples of situations of equilibrium and non-equilibrium sured dose and that of the dose at the same point of the
charged particles [http://www.fisicamente.net/FISICA/index-22.htm, unperturbed medium exclusively by taking special care.
accessed 9th July, 2017]
We recall that the transfer of energy to the medium by
primary X and gamma radiation can be described by the fol-
the contrary, in the case of Fig. 5.1d, the energy released by lowing two successive stages:
the radiation is less than that deposited in the middle. More
generally, intermediate situations, similar to those of 1. Primary radiation transfers energy to secondary charged
Fig. 5.1a, will be considered. In this case, the energy dis- corpuscular radiation (electrons and positrons).
charged from the radiation coincides with that deposited in 2. Secondary charged particles transfer energy in the

the middle if the energy dissipated by the electrons of type 1 medium by coulomb collisions.
within the volume considered is offset by the energy trans-
ported by the electrons of type 3 out of it. This is precisely We also recall that only in conditions of the charged par-
the situation that is used to indicate as a “condition of charged ticle equilibrium the energy deposited at a point of the
particle equilibrium.” medium coincides with the energy transferred at that point.
In practice, this equilibrium occurs when the two follow- So, we want to establish a relationship between the dose DG
ing conditions are valid at the same time: measured by a certain dosimeter placed in a cavity around
the reference point of the considered medium and the dose
1. The reference volume element must be immersed in a DM at the same point as the unperturbed medium. If the cav-
volume of material of such dimensions that the maximum ity is of any size, it is not possible to conceive a priori any
path of the secondary loads set in motion in the reference relationship. If the dimensions of the cavity are small com-
volume by indirect ionizing radiation does not allow them pared to the path of the secondary charged particles set in
to exit the total volume of the material in question. motion in the medium, so that they lose only a small fraction
2. The energy fluence of primary indirect ionizing radiation of their energy in it, and if the primary interactions in the
does not have to fluctuate appreciably on distances in the cavity can be neglected, then the dose absorbed into the cav-
order of that maximum path. ity depends only on the energy deposited in the cavity by the
secondary charged particles:
In conditions of charged particle equilibrium, the equiva-
DM = s DG (5.26)
lence between the absorbed dose and the kerma and the
S 5.2 Instrumentation

ρ
  col , M
s = (5.27) All ionizing radiation detection methods are based on a mea-
S sure of quantities influenced by the interaction of the radia-
ρ
  tion with the detector. Generally this quantity is a charge or
col , G
electric current, but it can also be heat, light, a chemical
Instead, if the cavity dimensions are much greater than effect, opacification of a transparent substance, or transfor-
the paths of the secondary charged particles set in motion in mation into a different type of radiation. The physical prin-
the medium, the ionization produced in the cavity will be due ciples most commonly used in radiation detectors are the
solely to the secondary charged particles set in motion ionization of a gas, the scintillation in an organic or inorganic
directly in the material occupying the cavity itself. In this material, and the production of electron-hole pairs in a
case (for primary monoenergetics), we will have: semiconductor.
 µen 
 ρ 
 M
DM = DG (5.28)
 µen  5.2.1 P
hysical Principles of Ionization
 ρ  Detectors
 G

If the chemical composition of the material filling the cav- Ionization detectors are based on the ionization produced in
ity is equal to that of the surrounding medium, 〈s〉 = 1, and if a gas by the energy dissipated by radiation. The number of
the absorption coefficients are the same, DM = DG, regardless positive electron-ion pairs that is produced is proportional to
of the cavity size. More generally, for DM = DG it is sufficient the energy deposited by the radiation. The proportionality
for the two media to have the same massive stopping power constant, called ionization potential W, depends on the type
and the mass absorption coefficient. In this case the two of gas and is greater than the corresponding binding energy,
means are said to be equivalent and the cavities since the stored energy is partially used also in excitation
homogeneous. processes. The number of pairs produced in relation to the
energy released by radiation is given by:
E
n= (5.29)
Key Learning Points W
Quantities for Radioactivity Measurements. Table 5.1 shows the values of the ionization potential for
The main quantities for radioactivity measurements some gases in case of irradiation with alpha particles and
are subdivided into radiometric quantities (which electrons.
describe the radiation fields—source or field quanti- The application of an electric field causes the acceleration
ties), quantities which describe the interactions of electrons and positive ions along the electric field toward
between radiation and matter, and dosimetric quanti- the anode and the cathode. Acceleration stops due to the
ties (which describe the various phases of energy impacts with the gas molecules when the maximum speed
transfer processes from radiation to matter and the (drift speed) is reached. For low electrical fields, recombina-
ability of radiation to ionize the atoms or molecules of tion phenomena may occur.
the crossed medium).
Charged Particle Equilibrium (CPE).
In CPE conditions, the energy dissipated outside a Table 5.1 Values of ionization potential (expressed in eV/ions pair)
for several gases irradiated with alpha particles and electrons (Glenn
small spherical volume, centered at the reference point, F. Knoll. Radiation Detection and Measurement, John Wiley & Sons,
by secondary particles charged therein produced by Inc., 2000)
indirectly ionizing radiation, is compensated by the Gas Fast electrons Alpha particles
energy dissipated therein by charged secondary parti- A 27.0 25.9
cles produced by indirectly ionizing radiation outside He 32.5 31.7
of that volume. When these conditions are verified, it is H2 38.0 37.0
possible to establish relationships between the dosi- N2 35.8 36.0
metric quantities, especially for practical purposes of Air 35.0 35.2
measurement. O2 32.2 32.2
CH4 30.2 29.0
124 M. E. Fantacci
Depending on the extent of the electric field applied, it is different mobility of electrons and positive ions. In flat
possible to subdivide the detectors that are based on the geometry, the picked signal also depends on the interaction
physical principle of the ionization of a gas in various opera- point of the radiation, and the proportionality with the
tion areas, as shown in Fig. 5.2. released energy is lost. In cylindrical geometry (Fig. 5.3), the
In region A, not all the produced charges are collected multiplication zone is limited and the signal is always pro-
because, due to the small value of the electric field, the portional to the released energy.
recombination process of the various ion-electron pairs is At even higher voltages than the proportional counter (E
remarkable. By increasing the potential applied, the time region, Geiger-Müller regime), even a few electrons moving
available for recombination decreases because the velocities toward the anode cause a heavy discharge which, however, is
along the direction of the field increase, thus allowing an independent of energy. In fact, for very high electrical fields,
increase of collected charges. In region B, called saturation the space charge influences the electric field near the anode,
or ionization chamber, the effects of recombination become and the proportionality is lost. An avalanche reaction is then
negligible and the collection charge is complete. In the ion- established, with emission of photons in the ultraviolet (UV)
ization chamber regimen, the electric field applied is suffi- frequency band by molecular excitement. The resulting dis-
cient to collect all the generated positive electron-ion pairs. charge can be switched off if there is a quenching gas that
All (and only) the charge generated by ionization is collected absorbs such photons. The response of such a detector is
from the electric field. slow and all the signals have the same amplitude.
In the C and D regions, the electric field is sufficiently Finally, increasing the voltage (F region) is no longer pos-
intense that the kinetic energy of the electrons produced is sible in any type of detector: the output pulse no longer
sufficient to ionize the atoms of the gas, thus producing an depends on the incident radiation, as a discharge occurs in
avalanche multiplication. Secondary ionization is still strictly the presence of radiation or not.
dependent on the primary one, and in this region the propor-
tional counters work. Indeed, in the proportional counter
system, the electric field is sufficiently high to accelerate the 5.2.2 P
hysical Principles of Scintillation
generated electrons enough to cause other ionizations. Detectors
The proportional counter works at higher voltages than the
ionization chamber, so that the pulses are amplified. The ava- Scintillators are materials that have a particular property
lanche has a typical drop shape, which is determined by the known as luminescence. In general, the luminescence process
consists in absorbing energy and re-emitting it in the form of
visible radiation following an excitation process that may be
A B C D E F
due to light, mechanical stress, chemical reaction, or heating.
Scintillation is the light emission process that accompanies
the interaction of a radiation with the scintillating substance
and draws its origin from the excitations and the ionizations
produced. The luminescent materials absorb energy and re-
emit them in the form of visible light. If the emission occurs
during or shortly after absorption (within 10−8 s), the process
is called fluorescence. If the emission is delayed (the excited
α
state is metastable), the process is called phosphorescence. In
β this case, the time between absorption and re-emission can last
from millisecond to hours (depending on the material).
Generally, the fast component is the one that dominates (and
that is used in radiation detectors). The emitted light is then
Fig. 5.2 Current I as a function of the potential V (arbitrary units) transmitted (directly or through light guides) to particular
E
V
E=
r ln(b / a)
b
r
Fig. 5.3 Proportional counter with cylindrical geometry (a = anode wire radius, b = cathode cylinder radius, r = distance from the wire, V = poten-
tial difference, and E = electric field)
devices (e.g., photomultiplier tubes) able to transform the light 5.2.2.1 Organic Scintillators
signals into measurable electrical signals. Organic scintillators are hydrocarbon compounds that con-
Although there are many scintillating materials, not all tain benzene ring structures. In these compounds, the scintil-
are suitable for constituting a radiation detector. To use a lation light derives from transitions of the free valence
scintillating material as a detector, the following conditions electrons of the molecules. These electrons are not associ-
must be met: ated with a particular atom in the molecule and occupy
molecular orbits. The energy levels of these materials can be
• The light produced must be proportional to the energy schematized as follows (Fig. 5.4):
deposited. The energy released by radiation excites both fundamen-
• The material must be transparent to the wavelength of its tal and vibrational levels. Single-state excitations decay in
output. ≤10 ps without emitting radiation (internal conversion).
• Induced luminescence decay time must be short to have From the S1 state, it is easy to decay in the fundamental state
rapid impulse signals. S0 with fluorescence light emission in 1–10 ns. Similarly the
• Its refractive index should be as close as possible to that triplet state is reached by internal conversion to T1 state, and
of glass (~1.5) to allow efficient transmission of scintilla- then it drops to T0 in a complex manner with emission of
tion light to the photocathode. phosphorescence light (slow, >10−4 s).
The absorption and emission spectra have only a small
There are two types of scintillating materials: organic overlap area (corresponding to transitions between levels S00
scintillators and inorganic scintillators. and S10):
Fig. 5.4 Diagram of energy Singlet Triplet

levels of organic scintillators
(https://www.researchgate. S30
net/publication/278332922_
Radiation_hardness_of_ Radiationless internal
plastic_scintillators_for_the_
conversion
Tile_Calorimeter_of_the_
ATLAS_detector/
figures?lo=1, accessed 9th S21 T30
July, 2017) S20
S13
T20
S12
S11
S10 Inte
r-sy
cro ste
ssin m
g
T10
Fluorescence
Absorption
Phosphorescence
S03
S02
S01
S00
126 M. E. Fantacci
Organic scintillators can be: accompanied by radiation emission (fluorescence or phos-

phorescence), or the energy can be dissipated in non-radiative
• Organic crystals: The most common are anthracene and mode (thermal energy or vibration of the lattice,
naphthalene. The anthracene is relatively slow (30 ns), but quenching).
it gives enough light. Naphthalene is quicker but gives The effect of scintillation can also be obtained from noble
little light. liquid elements such as argon, xenon, and krypton (Ar, Xe,
• Liquid and plastic scintillators (the most used ones): Kr), whose sparkling light is however difficult to reveal as it
Liquids: Solutions of one or more organic scintillators is emitted in the ultraviolet (Ar, 130 nm; Kr, 150 nm; Xe,
dissolved in an organic solvent. The energy released by 175 nm); even high-pressure noble gases can scintillate, with
the particle is generally absorbed by the solvent and then the same wavelengths of the corresponding liquids.
released to the solute (rapid and efficient transfer). The
solute (or solutes) work as wavelength-shifter actuators or
absorb, for example, in the ultraviolet frequency band and 5.2.3 P
hysical Principles of Semiconductor
emit in the visible. Detectors
Plastics: They are all analogous to liquid scintillators for
operation and composition (solvents and solutes) but are Materials which have in their bend structure a prohibited
solid. range of energies that separates a lower band (valence
band) and a superior (conduction band) are semiconduc-
The response times of liquid and plastic scintillators are tors, if the energy bandwidth is lower than a certain value,
short: for liquids 3–4 ns and for solid 2–3 ns. Organic scintil- usually fixed at 2 eV. These high-purity materials also show
lation scales have low Z (being H and C) and therefore the property of electrical conductivity, called intrinsic, if
exhibit a low detection efficiency for photons. carried to a certain temperature sufficient to cause the pres-
ence of charge carriers for thermal excitation. A pure semi-
5.2.2.2 Inorganic Scintillators conductor, which is not subject to doping phenomenon
Inorganic scintillators consist principally of salt crystals, (also called intrinsic semiconductor), has two types of
basically alkali halides, containing small amounts of impuri- charge carriers, electrons and holes, and its conductivity
ties having the function of activating the luminescence will be due to two different mechanisms. These carriers are
process. generated by the thermal excitation of electrons in the con-
In crystals, the scintillation mechanism depends on the duction band, which leave an equal number of holes in the
energy states determined by the crystalline lattice of the valence band. We note that the hole is actually a “dummy”
material. In the valence band, electrons are bound to the lat- charge carrier, which does not correspond to any particle.
tice sites, while in the conduction band, electrons are free to Intrinsic semiconductors (e.g., Ge, Si) have a cubic struc-
migrate into the crystal. Intermediate energy levels are for- ture of a diamond. In this structure, the bonds are predomi-
bidden in a pure crystal, but small amounts of impurities nantly covalent (and therefore strongly directional). In the
added to inorganic scintillators change the band structure, as lattice of the semiconductor material, doping atoms can be
shown in Fig. 5.5. introduced, which, having dimensions similar to those of
The interaction of ionizing radiation in the crystal can the element constituting the material, can partially replace
cause the passage of electrons from the valence band to the these. For example, in silicon lattice the silicon atoms can
conduction band. An electron in the conduction band is in an be exchanged atoms of phosphorus, arsenic, boron, germa-
excited state; when it returns to the valence band (fundamen- nium, etc. If these atoms possess a different number of
tal state), the energy is released. An excited electron can external electrons (valence electrons), lower or higher than
move through the crystal until it comes near an impurity. In those of the host atom (e.g., in the case of silicon, such
this case it occupies an energy level associated with it. From number is 4), they can provide electrons to the conducting
this state there may be a transition to the valence band band or holes to the valence band. In this case the conduc-
tivity is mainly due to holes (type “p”) or to electrons (type
“n”), and the semiconductor is called extrinsic. For exam-
pure crystal conduction band ple, the addition of a pentavalent impurity such as phospho-
rus (P), arsenic (As), or antimony (Sb) in a tetravalent
Activator excited states atomic network (Si, Ge) provides an electron, as only four
Band Gap of the five electrons of the pentavalent atom can participate
in the formation of the four lattice bonds. Thus, in the
Activator ground state valence band, four states are available, and the fifth elec-
pure crystal valence band
tron does not enter a bond state but does not even enter the
Fig. 5.5 Energy levels of inorganic scintillators (https://inspirehep. conduction band, remaining confined in a region, positively
net/record/1279895/plots, accessed 9th July, 2017) charged, around the impurity atom. To this additional elec-
Fig. 5.6 Energy levels in

extrinsic semiconductors Conduction Band Conduction Band
Extra
(http://hyperphysics.phy-astr.
gsu.edu/hbase/Solids/dsem. hole
Fermi energy
html, accessed 9th July, 2017) level levels
Extra
Fermi
electron
level
energy
levels Valence Band Valence Band
N-Type P-Type
tron can be assigned a series of quantum states, analogous signal provides information on the radiation under examina-
to those of the electron in the hydrogen atom; the binding tion. Given the low energy required to produce such pairs
energy is much smaller, of the order of 0.01 eV, very close (e.g., 3.6 eV in silicon, to be compared, e.g., with more than
to the conduction band, being exactly 0.01 eV (binding 30 eV needed for the production of an electron-ion in the
energy). air), these materials are indicated for spectroscopic applica-
Remember the relationship between absolute temperature tions. In fact, when an electric field is applied to a semicon-
(T) and thermal drift speed (V): ductor, both electrons and holes move. Their speed is the
sum of the thermal velocity, of random direction, and of a
1 2 3
mv = kT (5.30) drift speed, parallel to that of the electric field:
2 2
vh = µ h E (5.31)
where m is the mass of the electron and k is the Boltzmann
constant. ve = µe E (5.32)
At room temperature kT ≈ 0.025 eV, so the free electrons
can easily be excited by the thermal energy in the conduction where vh is the speed of the holes, ve is the velocity of elec-
band. An impurity of this type is called donor because it trons, μh is the mobility of the holes, μe is the mobility of the
gives the conducting electrons without producing gaps in the electrons, and E is the electric field.
valence band. In a semiconductor of this type, called extrin- For high electrical fields, saturation speeds of 107 cm/s
sic type n, electrons become then the major carriers and are achieved. An obstacle to this process is due to the fact
holes the minority carriers. that the average life span of charge carriers is influenced by
Trivalent impurities such as boron (B), aluminum (Al), the presence of impurities, consisting of metal atoms (gold,
gallium (Ga), and indium (In), substituted in the tetravalent zinc, cadmium) that occupy crystalline lattice sites in the
lattice, have an opposite effect, one of the four bonds sur- manufacturing process of the semiconductor and introduce
rounding the trivalent atom lacks an electron, so there is a energy levels at around half of the forbidden energy gap
gap near the valence band. The gap is attracted by the nega- (deep levels). These levels act as “traps,” capturing elec-
tive charge established on the trivalent atom and sets a set of trons or holes. Some deep impurities act as recombination
quantum states similar to those of the electron previously centers, capturing both electrons and holes. Structural
illustrated. The basic state of the gap is about 0.01 eV above defects in the lattice can also act as trapping or recombina-
the valence band, so in perfect analogy, a hole can be formed tion centers.
in the valence band simply by thermal excitation. In an
extrinsic semiconductor of this type (p), the major carriers
are therefore the holes.
Figure 5.6 summarizes the characteristics of energy levels Key Learning Points
in type n and type p extrinsic semiconductors. Main Kinds of Radiation Detectors.
Thin semiconductor layers with very high concentrations The main kinds of radiation detectors are ionization
of impurities are indicated as n+ or p+. They have very high detectors, semiconductor detectors, and scintillation
conductivity and are used to make the contacts of semicon- (both organic and inorganic) detectors.
ductor devices, as their very low density of minority carriers Physical Principles of Radiation Detectors.
allows their use as “blocking contacts.” The physical principles most commonly used in radia-
The possibility of using semiconductors for radioprotection detectors are the ionization of a gas (ionization
tion purposes derives from the fact that the energy trans- detectors), the scintillation in an organic or inorganic
ferred to semiconductor materials from ionizing radiation material (scintillation detectors), and the production of
results in the formation of electron-gap pairs which, in the electron-hole pairs in a semiconductor (semiconductor
presence of suitable electric fields, migrate to the electrodes, detectors).
giving rise to an electrical signal. Measuring this electrical
128 M. E. Fantacci
5.3 Detectors for Dosimetry lows: C 78.0%, H2 10.0%; O2 5.2%, N2 5.2%, Ca 1.8%, F

1.7% per wall; CH4 64.4%, CO2 32.4%; N2 3.2% for gas. We
5.3.1 Ionization Detectors for Dosimetry note that, with respect to the chemical composition of a fab-
ric, there is a defect of oxygen and an excess of carbon. So,
Ionization detectors have been among the first detectors this detector is equivalent to the tissue only until such differ-
used for dosimetric and radioprotection purposes and are ences can be neglected, that is, only in a certain range of
currently among the most common. Regarding the practical energy. This interval, in the case of X-rays and gamma,
arrangements to be used for their proper use in this field, extends up to a few megaelectronvolts. Tissue-equivalent
remember that the voltage between the electrodes must chambers can be made in very small sizes with sensitive
allow complete collection of all the produced ions. It is also volumes of the order of 0.1 cm3, as is required, for example,
necessary to take into account the thickness of the walls. If for dose measurements at various points of a human body
these can be considered thin in the sense that their thickness simulator exposed to an X-ray beam.
is small compared to the range of secondary charged parti-
cles set in motion by indirect ionizing primary incident radi- 5.3.1.1 Free-Air Ionization Chamber
ation, the charge provides a measure of ionization produced It is used for direct exposure measurement. Photons pene-
in the gas by the secondary charged particles generated in trate the sensitive volume through a collimator or diaphragm
the material surrounding the walls of the detector. Since the made up of a material with a high atomic number, so as to
path of low-energy electrons in the materials is very short, it control the input geometry of the photon beam by absorp-
makes sense to use a thin-walled room only in the case of tion. The chamber is equipped with flat and parallel elec-
high-energy radiation. If the detector has thick walls, the trodes; one of them is connected to the high voltage, while
collected charge provides a measure of the ionization pro- the other is connected to the ground by means of a measuring
duced in the gas by the secondary charged particles set in instrument that collects the charge. The measurement vol-
motion in the material that forms the walls. In this case, it is ume is located at the intersection of the collection volume
necessary that these have the same characteristics as those with the radiation cone delimited by the collimator (Fig. 5.7).
of the material in which the dose is to be measured. In par- The ratio between the charge collected and the air mass
ticular, special materials known as “tissue-equivalent” are contained in the measuring volume provides the exposure
used for measurements in tissues. These detectors have value, if the CPE conditions are satisfied. Verification of this
walls made of plastic whose characteristics with respect to condition is indispensable since the electrons which are pro-
radiation are very similar to those of organic tissues, and duced outside the measurement volume penetrate into it and
also the filling gas used presents the same characteristics. cause ionization inside it and must be compensated by others
The chemical compositions of walls and gases are as fol- with similar characteristics which are produced within the
Fig. 5.7 Schematic diagram High-voltage electrode

of a parallel plate free-air
Shielding box
ionization chamber (https://
www.researchgate.net/
publication/264978394_
Conception_and_realization_
of_a_parallel-plate_free_air_
ionization_chamber_for_the_ Diaphragm Collection
absolute_dosimetry_of_an_ volume
ultrasoft_X-ray_beam/
figures?lo=1, accessed 9th
Measurement
July, 2017)
volume
d
Measurement
Reference plane
plane
Guard plate
X-ray beam
Guard electrode Collector electrode Guard electrode

electric discharge which, suitably amplified, is registered by

the instrument. The Geiger-Müller counter can then detect
the number of particles. In the Geiger-Müller counters, the
potential difference applied to the electrodes is quite high
(generally over 800 V), higher than that used for the propor-
tional counter. For the higher voltage used, the feature of
this instrument is that the electric impulse detected at the
passage of an ionizing radiation is independent of the ion-
ization produced. This does not permit distinguishing the
type of radiation or measuring its energy. For this character-
istic, however, it is the most useful instrument for beta and
gamma ray counting, also because it does not require shield-
ing or isolation as with other ionization counters. A Geiger
detector can also count alpha particles, but in this case it
requires a very thin window that the charged particles can
penetrate; this condition suffers from the pressure of the gas
used, which is about 10-fold less than atmospheric. The
Fig. 5.8 Example of free-air ionization chamber (reproduced with per- Geiger-Müller counter consists of a cylindrical cathode of
mission from Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di 1–10 cm in diameter and a length of two to ten times greater.
Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010) In these devices the charge collection is independent of the
primary ionization: in fact, in addition to ionization, there
measurement volume and exit causing ionization outside it. are phenomena such as excitation followed by visible light
These conditions are satisfied for primary photons with and ultraviolet light. A small portion of these photons gives
energy up to 400 keV. Therefore, in the case of higher-energy rise to photoelectrons that generate new ionization through
photons, other measurement methods are required. An exam- the process of avalanche multiplication. In the operating
ple of ionizing chamber is shown in Fig. 5.8. region of Geiger-Müller counter, only one primary pair is
required to give rise to a complete avalanche discharge, and
5.3.1.2 Cavity Ionization Chamber therefore the amplitude of the output pulse is no longer a
The use of a cavity ionization chamber exploits the relation- measure of primary ionization. It is understood that a Geiger
ship between the dose absorbed at a point and the exposure counter can be used as a radiation counter, not in spectros-
at the same point until the CPE conditions are satisfied. It is copy experiments. Examples of Geiger-Müller counters are
possible to realize a chamber filled with air-equivalent walls, shown in Fig. 5.9.
mixing, for example, carbon and bakelite. We note again that
the thickness of these walls must be such as to meet the con- 5.3.1.4 Dosimetric Pen
ditions for the charged particle equilibrium and to ensure that The dosimetric pen (or stylo dosimeter) is a small pen-shaped
all the ionization produced in the air contained in the cham- ionization chamber that is used for personal dosimetry.
ber is originated by secondary ones set in motion in the walls Sensitive volume is in the order of 2 cm3. The measurement
themselves. If a chamber has to be used for various energies, range can range from a few tens of mR to some R. The opera-
it is equipped with “caps” of different thicknesses of materi- tion mode is as follows: we preliminarily charge the ioniza-
als air equivalent. This is possible for photon energies up to tion chamber and measure the voltage variation between the
3 MeV. Above this energy, in fact, it would no longer be pos- electrodes due to the discharge produced by the passage of
sible to construct cavities that can satisfy the conditions of the ionizing radiation. Reading can be direct or indirect. In
charged particle equilibrium due to the negligible absorption direct reading instruments, the movement of the wire of an
in the walls. electrometer can be observed directly on a translucent quad-
rant included in the chamber without the aid of any external
5.3.1.3 The Geiger-Müller Counter instrument. In indirect reading, instead, the residual stored
The Geiger-Müller (or Geiger) counter is a device able to charge is measured by an external reader. These dosimeters
count charged particles (alpha, beta, protons, and ionizing are particularly suitable for personal control but rarely consti-
particles in general) consisting of a metal cylinder (cathode) tute the only tool assigned to professionally exposed person-
containing a rarefied gas and an isolated conductor wire nel. These instruments do not guarantee high precision when
(collector) to which a high potential difference is applied, used for long periods of time (weeks), not only because of the
close to that of discharge. Under such conditions, when an spontaneous discharge of the chamber, which could be taken
ionizing particle passes through the tube, it causes a short into account, but also because they are affected by shocks and
130 M. E. Fantacci
Fig. 5.9 Examples of

Geiger-Muller counters
(reproduced with permission
from Volterrani D, Erba PA,
Mariani G, Eds. Fondamenti di
Medicina Nucleare – Tecniche e
Applicazioni. Milan: Springer;
2010)
from the reduction, suspended within the gelatin, is respon-

sible for the phenomenon of the blackening. The blackening
level at a point of a photographic film developed and fixed is
measured by means of a densitometer instrument by sending
a narrow beam of light of intensity I0 at that point and mea-
suring the intensity I passing through the film without being
absorbed. The optical density is defined by the base 10 loga-
rithm of the ratio between I0 and I.
I0
Dottica = log10 (5.33)
Fig. 5.10 Example of dosimetric pen (reproduced with permission I
from Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina For example, a film having an optical density equal to 1
Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010)
transmits 1/10 of the incident light. Since optical density at
one point depends on the total energy absorbed at that point
from the photographic emulsion following irradiation, it is
variations of temperature and humidity. An example of a dosi- possible to obtain, through appropriate calibration factors
metric pen is shown in Fig. 5.10: and taking into account the intrinsic optical density of the
film itself the absorbed dose. The possibility of using a film
for a long time (more than a few months) is limited by the
5.3.2 Film Dosimeters phenomenon of spontaneous degradation of the latent image
in the film (fading). Figure 5.11 shows some examples of
In these dosimeters, the blackening of a photographic film film dosimeters.
caused by the absorption of ionizing radiation is related, by a
suitable calibration, to the dose absorbed by the film itself.
Photographic film is made up of a layer of gelatin (emulsion) 5.3.3 Thermoluminescent Dosimeters (TLDs)
in which small granules of AgBr are in suspension, deposited
on a support consisting essentially of cellulose acetate. The Thermoluminescent (TL) dosimeters are based on the proper-
emulsion thickness is between a few micrometers and a few ties of certain substances, typically crystals such as calcium
hundred of micrometers. When an AgBr grain absorbs fluoride (CaF2) and lithium fluoride (LiF). When these mate-
energy from ionizing radiation, metallic silver agglomerates rials are exposed to ionizing radiation the crystal lattice
are formed, which form the latent image. During the chemi- absorbs the radiation. When the material is heated, the stored
cal development process, which takes place in a special energy is released as light. The intensity of the light emitted is
reducing solution, the presence of these agglomerates favors proportional to the absorbed dose. The reading of the dosim-
the metallic silver reduction of the other Ag+ ions present in eter, which is done by heating, destroys the information. The
the grain. The undeveloped AgBr granules are dissolved effect of ionizing radiation in the crystal is to produce free
from the fixation solution in which the film is immersed after electrons, some of which can be captured by impurity atoms
bath in the growth solution. The metallic silver resulting or other defects present in the crystalline solid lattice. These
Fig. 5.11 Examples of film dosimeters (reproduced with permission from Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
190° C • Loss of information at room temperature, due to a long

period of time between irradiation and reading (fading), is
5
negligible in many common TL materials.
• The dosimeter reading system can be automated.
Relative TL Intensity
4
For these reasons, thermoluminescence dosimeters can
be used in the field of personal dosimetry, from all those
3 who work in environments where it is necessary to periodi-
55° C 2 cally check the absorbed dose. Similarly, they are used to
300° C
monitor the environmental dose in the aforementioned envi-
1
ronments. In the clinical context, TLDs can be used as
6
devices for measuring radiation from the machines used in
diagnostic and therapeutic applications, for measuring dose
Phosphor Temperature (°C)
distribution in tumor sites with the help of anthropomorphic
Fig. 5.12 Examples of glow curve (https://www.nukeworker.com/ phantoms for in vivo measurements on the patient, both in
study/hp/tlds/tlds.shtml, accessed 9th July, 2017) the case of external and endocavitary dosimetry. Both in
diagnostics and therapy, TLDs are useful for measuring
defects are commonly called traps, and the electrons captured gonad dose.
in them may last for longer or lesser times. When the crystal Figure 5.13 show an example of TLD and its reading
is heated, trapped electrons acquire enough energy to get out instrumentation.
of traps and return to their original positions; this process pro-
duces a light emission. In the usual technical language, such
light emission is indicated by the term “glow curve.” A glow 5.3.4 Probe for Uptake Measurements
curve is usually composed of a set of peaks, and each peak
corresponds to a certain degree of trapping. An example of An uptake probe typically consists of a lead collimator with
glow curve is reported in Fig. 5.12. a single large hole (flat field collimator), a scintillation
The use of TLDs in the field of radiation protection is detector, and a photomultiplier tube connected to a com-
particularly effective with respect to film-based dosimeters, puter for processing the data. The most commonly used
largely used in the past, for the following reasons: detector for this type of measurement is a NaI(Tl) crystal of
a thickness of 2 in., but other crystals or semiconductor
• They are solid and small materials (e.g., 4 × 4 mm and detectors can also be used. Usually, processing software
less than 1 mm thick). allows, in addition to radiometabolic activity count, a spec-
• Some TL materials are equivalent to the tissue, i.e., they tral analysis of the isotope in use. The latest generation
respond to ionizing radiation in the same way as the bio- uptake probes consist of a real small-field gamma camera
logical tissue. that allows acquisition of images with high resolution and
• TL response is independent of the incidence angle of radi- selection of areas of equal interest for subsequent measure-
ation as the dose rate and is linear in a wide absorbed dose ments. Figure 5.14 shows an example of conventional
area, from a few mGy to dozens of Gy. uptake probe.
132 M. E. Fantacci
Fig. 5.13 Example of TLD

(left) and reading instrument
(right) (http://www.sofimedsrl.
com/dosimetria.html,
accessed 9th July, 2017)
5.3.5 Probes for Intraoperative Use
The probes for intraoperative use can be classified into two

main categories based on the principle of revelation:
1 . Scintillator crystal probes: NaI(Tl), CsI(Na), CsI(Tl)

2. Semiconductor probes: CdTe, CdZnTe
Generally, the scintillators exhibit greater intrinsic effi-

ciency for photon detection due to the higher atomic num-
ber than the semiconductors. This concept, however, does
not automatically result in greater sensitivity as the geom-
etry of the crystal collimator system and the use of addi-
tional collimators determine sensitivity. In addition, the
scintillating crystals can be made with thicknesses of
10 mm or more, while the thickness of the semiconductor
detectors is limited to 6–7 mm due to electron trapping and/
or holes. Because of these factors, semiconductor probes
initially designed for use with 125I (27 and 32 keV peaks)
have low sensitivity for use with 99mTc (140 keV peak) and
are unsuitable for higher energies (171 and 245 keV of
111
In). From the point of view of the energy resolution,
semiconductors are theoretically better, but it does not
always mean that there is a greater ability to discriminate
Compton interactions with respect to photoelectric effect,
since the spectrum is often asymmetrical, with a low-energy
tail caused by the incomplete collection of charges due to
the trapping.
The choice of the probe type is made on the basis of the
application to which it is used. In fact, intraoperative probes
are clinically employed in two main areas of activity that
Fig. 5.14 Example of probe for uptake measurements for the thyroid
(http://www.acn.it/products/Monogamma.html, accessed 9th July, physically present different characteristics: radioimmuno-
2017) guided surgery following systemic injection of radiolabeled
Samples with high counting rates allow short counting

times and allow for good statistical accuracy. When mea-
sures are to be taken on a limited number of such sam-
ples, the use of manual loading well counters is indicated.
Instead, for cases which require long counting times and/
or measurements on many samples, loading and changing
the samples is automated. Typically, these instruments
can hold 100 or more standard size samples, and each
sample is loaded in the counter sequentially. Most of
these systems use a “through-hole” geometry, where the
sample hole crosses the entire crystal. The main advan-
tage of this configuration is that the sample can always be
positioned at the center of the crystal regardless of its
volume. Furthermore, with these devices it is possible to
Fig. 5.15 Example of intraoperative probe for sentinel lymph node search
(http://www.medicalexpo.com/prod/capintec/product-70709-561595.html, perform automatic counting of background counts by
accessed 14th July, 2017) simply alternating the samples to be measured with
empty vials. One disadvantage is instead that no lead ver-
tical shield can be inserted at the sample housing.
monoclonal antibodies and radio-guided surgery for the The NaI(Tl) cockpit counters are used for radionuclide
detection of sentinel lymph nodes following local injection counts emitting X- or γ-rays. For β-emission counting mea-
of radiolabeled colloids. The substantial difference in the sures with these devices can be made by detecting brems-
modality of administration of the radiopharmaceutical means strahlung radiation, but the count rate is very low due to the
that in the first case the lesion is immersed in a radiation low production efficiency of that radiation.
background from the whole patient, with a very low (typi-
cally 2:1) uptake contrast between tumor lesions and sur-
rounding tissues, while in the second case the lymph node is 5.3.7 Liquid Scintillation Beta Counters
usually a few centimeters away from the inoculum site (in
the case of breast cancer) or tens of centimeters (in the case Liquid scintillators are used for the count measurements to
of skin melanoma) and exhibits a large uptake difference be made on radionuclides emitting β− radiation (3H, 14C) that
(50–200:1) with respect to the surrounding tissues. It is clear would be heavily absorbed into the glass or plastic that forms
therefore that probe localization is much more difficult in the the walls of the tubes and vials containing the samples on
first case, so its performance and settings have a more critical which the measurements are carried out through the well
role. We also note that the enormous recent diffusion of these counters. The range of β particles in these materials is, in
devices is mostly linked to the search for sentinel lymph fact, very low. In liquid scintillation counters, the radioactive
node (Fig. 5.15). sample is dissolved in a scintillating solution contained in
the counting vial. The emitted scintillation light is collected
directly from two photomultiplier tubes. The scintillating
5.3.6 Well Counters solution is usually made of low atomic number material (Z
between 6 and 8), but it is sufficient to detect both beta par-
A well counter consists of a single NaI(Tl) crystal in which ticles than low-energy X- and γ-rays with high efficiency.
an aperture has been made for inserting the sample to be
measured. In a standard well counter, the crystal has a diam-
eter of 4.5 cm and a height of 5 cm and contains a well of 5.3.8 Dose Calibrators
1.6 cm in diameter and 3.8 cm in height. Larger detectors (up
to 13 cm in diameter and 25 cm in height, with wells up to Dose calibrators consist of small well-type ionization cham-
3.8 cm in diameter and 7.0 cm in height) are used for particu- bers that are used to determine activity levels in syringes,
lar uses, such as low volume counts of high energy emissions tubes and other small containers intended to contain materi-
(40K and 137Cs). The well counters are equipped with lead als that are administered to patients. They are mainly suitable
screens of thicknesses of at least 5 cm to reduce counts due for measuring high levels (of the order of MBq) of gamma
to the background. Figure 5.16 shows a scheme and an emitters activity. Generally, in these devices the chamber in
example of a well counter. which ionization occurs is sealed to eliminate the effects
134 M. E. Fantacci
a b
Test tube
containing
sample
Nal(TI)
crystal
Lead
shielding
PM tube
Fig. 5.16 Scheme (a) and example (b) of a well counter (https://radiologykey.com/counting-systems/, accessed 14th July, 2017)
caused by the variation in atmospheric pressure on the mea-

surement. Dose calibrators are typically calibrated to read
directly in units of activity (Becquerel) and have selectors to
adjust the display in relationships to the various radionu-
clides for which they can be used. The most commonly used
gas for such devices is argon, which is sealed and pressurized
to avoid fluctuations in the response caused by variations in
the pressure in which the measurement is carried out. An
example of dose calibrator is given in Fig. 5.17.
5.3.9 Hand-Foot Contamination Monitors
The HF (hand-foot) contamination monitors use scintillation

detectors for the measurement of surface contamination
(typically in Bq/cm2) by alpha, beta, and gamma emitting
radionuclides on hands, feet, and clothing. The most com-
monly used materials in these monitors are plastic scintilla-
tors, such as ZnS(Ag), or their combinations. In their typical
configuration, they use two scintillators for hand contamina-
tion measurements and two for foot contamination. Usually
the monitor measures and evaluates the background level
throughout its standard operation. As soon as any detector Fig. 5.17 Example of ionization chamber dose calibrator (reproduced
(hand or foot detector) detects that a person is present, the with permission from Volterrani D, Erba PA, Mariani G, Eds.
Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
background measurement is interrupted and after reaching Springer; 2010)
Key Learning Points

Examples of Devices Used for Personal Dosimetry.
For personal dosimetry film dosimeters and thermolu-
minescent dosimeters (TLDs) are very frequently
used.
Examples of Instruments Based on Ionization Detectors
Free-air ionization chamber, cavity ionization cham-
ber, Geiger-Müller counter, dosimetric pen, and dose
calibrators are based on ionization detectors.
Examples of Instruments Based on Scintillation
Detectors and Semiconductor Detectors.
Hand-foot contamination monitors, well counters, and
liquid scintillation beta counters are usually based on
scintillation detectors. Probes for uptake measure-
ments and probes for intraoperative use can make use
of both scintillation and semiconductor detectors.
Further Reading
Bailey DL, Humm JL, Todd-Pokropek A, van Aswegeb A. Nuclear
medicine physics: a handbook for teachers and students. Vienna:
International Atomic Energy Agency; 2016.
Cantone MC, Hoeschen C, editors. Radiation physics for nuclear medi-
cine. Berlin: Springer; 2011.
Cherry SR, Soreson JA, Phelps ME, editors. Physics in nuclear medi-
cine. Philadelphia, PA: Saunders (Elsevier); 2003.
Knoll GF, editor. Radiation detection and measurement. New York,
NY: John Wiley & Sons, Inc; 2000.
Volterrani D, Erba PA, Mariani G, editors. Fondamenti di medicina
nucleare – tecniche e applicazioni. Milan: Springer Italy; 2010.
Fig. 5.18 Example of hand-foot contamination monitor (http://www.
vf.eu/data/files/b-14-a0001en-170205-hf-series-382-en.pdf, accessed
14th July 2017)
all measuring positions, the contamination measurement is

started. Information about the potentially wrong positioning
of any hand or foot is indicated on the display. The monitor
can be equipped with an acoustic warning alarm which will
sound if the contamination detected results higher than a pre-
set threshold. An example of HF monitor is shown in
Fig. 5.18.
General-Purpose Gamma Cameras,
Dedicated Gamma Cameras, 6
and Gamma-Probes for Radioguided
Surgery
Roberto Pani, Federica Guidoccio, Raffaele Scafè,

Pat Zanzonico, and Giuliano Mariani
Contents
6.2 Standard General-Purpose Gamma Cameras 138
6.2.1 Photodetection 139
6.2.2 Signal Processing 140
6.2.3 Energy Resolution 141
6.2.4 Gamma Cameras Based on Solid-State Detectors 141
6.2.5 Performance 141
6.3 Single-Photon Emission Computed Tomography (SPECT) 142
6.4 Collimators 144
6.4.1 Materials 144
6.4.2 Geometry 144
6.5 Multimodality Imaging: SPECT/CT 147
6.6 Dedicated Gamma Cameras 149
6.7 Basic Components of Dedicated Gamma Cameras and Small FOV Imaging Probes 150
6.7.1 Scintillation Materials 150
6.7.2 Other γ-Ray Detectors 150
6.7.3 Arrays of Scintillation Detectors 150
6.7.4 Position-Sensitive Photomultiplier Tubes (PSPMTs) 151
6.7.5 Geiger-Mode Avalanche Photodiodes (G-APD) 154
6.8 Single-Photon Emission Mammography (SPEM) 155
6.9 Cardiac-Dedicated Gamma Cameras 157
6.10 Handheld Gamma-Probes for Radioguided Surgery 161
6.10.1 Scintillation Probes 161
6.10.2 Semiconductor Probes 161
6.11 Imaging Probes 162
6.11.1 Generalities on the Use of Portable Probes for Radioguided Surgery 162
References 169
R. Pani R. Scafè
Department of Medical and Surgical Sciences and Department of Molecular Medicine, Sapienza University of Rome,
Biotechnologies, Sapienza University of Rome, Rome, Italy Rome, Italy
F. Guidoccio · G. Mariani (*) P. Zanzonico
Regional Center of Nuclear Medicine, Department of Translational Memorial Hospital Research Laboratories, Department of Medical
Research and Advanced Technologies in Medicine and Surgery, Physics, Memorial Sloan Kettering Cancer Center,
University of Pisa, Pisa, Italy New York, NY, USA
e-mail: giuliano.mariani@med.unipi.it

138 R. Pani et al.
Learning Objectives maceutical distribution in the body. In clinical practice, this

• Introduce the fundamental concepts of radionuclide imag- technique is referred to as scintigraphy. Tomography records
ing based on single-photon emission. data from multiple angles around the body, allowing visualiza-
• Describe the basic components and working operations of tion of radiopharmaceutical distribution in three dimensions,
standard gamma cameras: collimator, scintillation crystal, i.e., to display tracer distribution in various sections (slices),
photomultiplier tube, position logic circuitry, signal pro- from a set of planar scans acquired by each detector at various
cessing, sensitivity, energy resolution, and spatial rotation angles around the patient, a technique known as sin-
resolution. gle-photon emission computed tomography, or SPECT.
• Describe the working principles of solid-state semicon- The gamma camera consists of a collimator (see further
ductor detectors for dedicated gamma cameras. below), a slab of a scintillation crystal that is optically cou-
• Understand the structural and operational differences pled with a set of photomultiplier tubes (PMTs), and an elec-
between gamma cameras based on scintillation crystals tronic reading and data acquisition system. These components
and gamma cameras based on solid-state semiconductor create what is usually called the “revelation head”; a gamma
detectors. camera may be constituted by a single revelation head or by
• Describe the main modalities of scintigraphic acquisition multiple revelation heads (most frequently two heads, more
for planar imaging and for tomographic imaging with rarely three heads).
single-photon emission computed tomography. The scintillator crystal is the active part of the γ-ray detec-
• Introduce the main principles of multimodality imaging tion process. When crossing the scintillator, some γ-rays
with hybrid instrumentation. pass through without interacting, while other γ-rays interact
• Describe the main features of dedicated gamma cameras one or more times in the crystal by depositing all or part of
for cardiac imaging and for breast imaging. their energy. The probability that a γ-ray of given energy E
• Illustrate the main principles and components of small interacts with a scintillator depends on the thickness of the
field-of-view imaging probes. crystal and on the linear attenuation coefficient μ(E). This
• Illustrate the main principles and components of hand- probability is given by Eq. (6.1) expressing the number N(x)
held gamma-probes for intraoperative use during radiogu- of photons that can cross a thickness x of the scintillator as a
ided surgery. function of the number N0 of incident photons:
N ( x ) = N 0 e- m ( E ) x (6.1)

6.1 Introductory Background The value of μ(E) depends on the atomic number Z and
density ρ of the material and on the energy E of the photon.
Since its introduction in 1958 [1], the scintillation camera devel- This value is usually tabulated in terms of μ(E)/ρ (mass
oped by Hal Anger remains the most widely used system for sin- attenuation coefficient, expressed in cm2/g).
gle photon radionuclide imaging. The success of this instrument As scintillator material, NaI(Tl) is typically used in stan-
is largely due to the simplicity of the principle of localizing the dard gamma cameras. Since NaI(Tl) is highly hygroscopic,
position of the γ-ray interaction with the scintillation crystal. the crystal is protected from external moisture by sealing
Furthermore, continuing technological advances in growing within a thin aluminum housing. The inner surface of the crys-
large-volume NaI(Tl) scintillation crystals and in producing tal is coupled with the PMTs through a glass or quartz window
hexagonal photomultiplier tubes (PMTs) have led to the (called “light guide”), while the outer surface is shielded from
development of large size detectors, with an area sufficiently large ambient light by an opaque cover. The mass attenuation coef-
for human studies. The intrinsic spatial resolution is optimized to ficient of NaI(Tl) for 140 keV γ-rays is 0.654 cm2/g, corre-
image large and deep organs, where spatial resolution is dictated sponding to a length of attenuation of about 4.2 mm.
by the collimator properties. On the other hand, the large size and A γ-ray can interact with the crystal scintillator through
weight and the extension of a “dead” zone at the edges of the the photoelectric effect or through the Compton effect. In the
Anger camera make it difficult to position the camera for some photoelectric effect, the incident photon interacts with the
studies, e.g., when imaging a small organ [2]. atom as a whole; the photon is completely absorbed, causing
emission of an electron with energy equal to that of the inci-
dent photon less the binding energy of the electron itself.
6.2 tandard General-Purpose Gamma
S Instead, in a Compton interaction, the γ-ray interacts
Cameras with an orbital electron. After transferring only part of its
energy to the electron, the photon is diverted from its origi-
There are two acquisition modes for single-photon imaging: nal flight direction. The residual γ-ray energy depends on its
planar and tomographic [3–5]. Through planar imaging, a dual initial energy E0 and on the deviation angle θ according to
head gamma camera can acquire two projections of radiophar- Eq. (6.2):
6 General-Purpose Gamma Cameras, Dedicated Gamma Cameras, and Gamma-Probes for Radioguided Surgery 139
E0 non lies in the fact that the amount of light produced,

E= (6.2) expressed as a whole number of photons emitted, is (almost)
1+
(1 - cos q ) E0
proportional to the energy lost by the incident photon in the
me c 2
interaction.
where me is the rest mass of the electron. The residual energy
of the γ-ray varies between E0 (value corresponding to a null
angle deviation) and a minimum value Emin (corresponding to 6.2.1 Photodetection
a deviation of 180°), as expressed by Eq. (6.3):
Once the scintillator has produced the light flash, this must be
E0 revealed in such a way so as to measure its total energy (which
Emin = (6.3)
2 E0 is assumed to be equal to the energy lost by the incident pho-
1+
me c 2 ton) and exact position of the interaction. The system used for

this purpose was invented by Hal Anger in 1958. Although
In the Compton interaction, the energy transferred to the since then significant progress has been made in refining this
orbital electron (and therefore “measured by the scintilla- method, the operating principle is still the same, so much so
tor”) varies between 0 and a maximum value Eβmax, called that this detector is still called the Anger camera. The Anger
Compton edge, as expressed by Eq. (6.4): camera photodetection system consists of a set of PMTs
arranged on a regular hexagonal grid and optically coupled to
2 E02 the scintillator crystal through an appropriate light guide. In
Eb max = (6.4)
me c 2 + 2 E0 the original Anger camera, photomultipliers were circular,

while in current gamma cameras, PMTs are hexagonal, to
The fundamental characteristic property of scintillating maximize light collection (Fig. 6.1).
materials is to emit light in the visible (or near ultraviolet) Each PMT provides an electrical signal that is proportional
range when a radiation γ (or X) promotes an electron from to the amount of light that has reached the PMT (a vacuum tube
the valence band to the conduction band. In the subsequent with a transparent input window). A material that releases elec-
de-energization, the electron returns to the valence band. In trons upon interaction with light photons is deposited on this
the case of NaI(Tl), light photons are produced for each 3 eV window, which is called the photocathode. The electron emit-
of energy deposited in the crystal. The light has a 415 nm ted at the photocathode is called photoelectron (Fig. 6.2). The
wavelength. As mentioned in the previous paragraph, when a efficiency of this conversion (i.e., the fraction of electrons pro-
photon interacts with the material, an electron is emitted duced versus the number of photons that have reached the pho-
which passes from the valence band to the conduction band. tocathode) is called “quantum efficiency” and is about 20–25%
While repositioning the same electron in a state of minimal in modern PMTs with a typical diameter of 3″. Electrons pro-
energy, this goes through intermediate states that are induced duced during this photoconversion phenomenon are acceler-
in the forbidden band by suitable activating ions (or dop- ated by a potential difference applied within the PMT. Here, a
ants). In the NaI(Tl) crystal, this role is played by thallium. number of metal plates, called “dynodes,” provide for multipli-
When crossing these levels, a short-lived light flash (nano- cation (amplification) of the number of electrons. Dynodes
second to microseconds) is produced that can be measured emit multiple secondary electrons when they are hit by an elec-
by suitable photodetectors. The key point of this phenome- tron that has been accelerated by the voltage gradient (usually
Fig. 6.1 Diagrammatic representation of the possible arrangement of involving the PMT directly over the scintillation as well as the adjacent
hexagonal-based PMTs in a gamma camera: view from the PMT side PMTs (reproduced with permission from Volterrani D, Erba PA,
(left) and from the scintillator side (right). Position of the collimator is Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
also shown in the left panel. In the right panel, a scintillation event Applicazioni. Milan: Springer; 2010)
within the crystal is represented, with the resulting illumination area
140 R. Pani et al.
100 V) between each dynode and the next one; the first dynode and thus to the light collected. By adding the signals from the
has a potential higher than that of the photocathode, and each PMTs involved in the single event, it is possible to measure an
subsequent dynode has a greater potential (or less negative) electrical signal that is proportional to the energy deposited by
than the previous one. Thus, in the propagation of electrons the incident photon.
between the dynodes induced by the potential difference, there
are cascade multiplication effects and therefore amplification
of the signal. Typically, between anode and cathode, a negative 6.2.2 Signal Processing
potential difference of 600–1000 V is applied, and there can be
about 8–12 dynodes in a PMT. In this way, it is possible to Signals from individual PMTs are processed by an electronic
obtain a measurable signal whose intensity is proportional to system and transferred to a computer for subsequent analy-
the initial number of electrons produced by the photocathode sis. The energy deposited by the beam of γ rays can be
obtained from the sum of the electrical signals coming from
visible- the single PMTs. This information also serves as a consensus
light signal for the acquisition, as only signals having intensity
Y photon greater than a certain threshold value are considered signifi-
photo-e-
cant and therefore acquired (Fig. 6.3).
The coordinates x,y of the radius impact point of the γ-ray
in the crystal can be determined by the spatial distribution of
output the recorded signals. Such definition of the position in a pla-
glass signal
nar space is obtained by weighing the position of the PMT
window dynode anode
photocathode involved in each event with the intensity of the signal pro-
duced by it and calculating the mean of the values obtained.
Fig. 6.2 Diagrammatic representation of the operations of a PMT
(reproduced with permission from Volterrani D, Erba PA, Mariani G,
In practice, this corresponds to calculating the barycenter of
Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. the distribution of light in the crystal. In fact, the intensity of
Milan: Springer; 2010) light that “illuminates” a two-dimensional array of PMTs
Output Pre-amplifier
signal
+ 1200 V
Anode
Photomultiplier tube (PMT)
+ 1100 V
Magnetic Amplifier
+ 1000 V
shielding High
+ 600 V voltage
supply x106
Dynodes + 500 V
Energy
+ 400 V discriminator
Focusing + 300 V
grid
Photocathode
Timer/scaler
Light
Entrance Light pipe
photon
window 1 2
Scintillator
Reflective (crystal)
Display
Inner surface
of crystal housing
X - or γ-Ray Energy Eγ 2
Fig. 6.3 Diagrammatic representation of a scintillation detector cou- other two photons have pulse heights (therefore energies) outside the
pled with the PMT connected with the counting/discrimination system. photopeak energy window and are therefore not counted or included in
Only two of the four pulses have energy levels lying between the preset image formation (reproduced with permission from Zanzonico
pulse-height range (indicated by the two dash horizontal lines below the P. Instrumentation for positron emission imaging. In: Strauss HW,
energy discriminators in the right part of the figure). Thus, only two Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
photons correspond to the energies within the preset photopeak energy Pathophysiology to Clinical Applications. New York: Springer;
window and are considered for counting or for image formation. The 2017:217-50)
varies inversely with the distance between the scintillation improves the ability to reject the diffused γ-rays originated in
event within the crystal and the corresponding PMT(s). The Compton interactions.
Anger position logic circuitry is based on this inverse rela-
tionship for precisely localizing in the x,y matrix the scintil-
lation event within the gamma camera crystal. 6.2.4 G
amma Cameras Based on Solid-State
The sum of the signals (energy) and the center of gravity Detectors
(coordinates x,y) are obtained through statistical computing tech-
niques that use a combination of hardware, software, and the Gamma cameras based on solid-state ionization utilizing
implementation of information from experimental calibration semiconductor cadmium-zinc telluride (or CdZnTe) detectors
measurements. Although this technique implies a certain degree are now commercially available. Although CdZnTe detectors
of complexity, its main advantage consists in the possibility of are more expensive than NaI(Tl) crystals coupled with PMTs,
correcting any spatial nonlinearity when calculating the position as in the standard conventional Anger cameras, the solid state
of the γ-ray/crystal interaction as well as any local deviation cameras offer several important advantages, including
from the expected value when estimating the deposited energy, improved energy resolution (typically resulting in a 30%
based on prior appropriate calibration procedures. These are reduction of the scatter counts in the photopeak energy win-
obtained through the acquisition of the γ-rays emitted by radio- dow) and superior intrinsic spatial resolution [6, 7]. Solid-
active sources with regular geometry typically placed on a regu- state detectors allow direct conversion of an incident γ-photon
lar grid. Correspondence between the known positions of the into a certain number of electron holes that are transported to
sources and the measured position is stored in a correction table the respective electrode, thus producing a measurable electri-
(lookup table) and used for the calibration of spatial linearity. cal charge with an intensity which is proportional to the
energy of the incident γ-ray.
This detection process is at variance with the detector sys-
6.2.3 Energy Resolution tems based on a scintillation crystal coupled with a PMT,
where the measurable electrical signal is generated through
The sum of signals from PMTs is used to select only photons the production of secondary low-energy electrons. In partic-
with energy that is included within a certain predefined range, ular, in the scintillation detectors, intrinsic spatial resolution
i.e., within the energy window. The other detected photons are is degraded by the enlargement of the light spot in the crys-
discarded and therefore not used for formation of the final scin- tal, whereas, in solid-state detectors, lateral enlargement of
tigraphic image. Definition of the energy window serves to the electron cloud is much more limited, thus resulting in
select only those photons emitted by the γ-ray source of interest better spatial resolution.
and that have not undergone the Compton interaction in the sub- The mass density of CdZnTe is higher than that of NaI(Tl)
ject/patient, as these would lead to a degradation of image qual- (5.8 g/cm3 versus 3.7 g/cm3); however, due to higher cost and in
ity. Usually the optimal energy window is selected prior to the order to ensure excellent energy resolution, CdZnTe detectors
beginning acquisition, based on the characteristics of the radio- are thinner (~5 mm) than the typical 3/8″NaI(Tl) crystals (about
isotope used and the detection system. To this end, the charac- 9.5 mm). Therefore, the intrinsic sensitivities of CdZnTe-based
teristic parameter of a gamma camera is its energy resolution and NaI(Tl)-based detectors are comparable. However, the
In fact, the value provided by the PMT, intensity of the light compact form factor of CdZnTe detectors permits novel detec-
signal due to interaction with the radius γ-rays, does not tor geometries and, in turn, high-sensitivity collimation.
exactly match the value of the energy released in the scintilla-
tor, but there is a certain degree of fluctuation. This uncertainty
on the energy of the incident γ-ray is measured by a quantity 6.2.5 Performance
called energy resolution. This uncertainty stems from the fluc-
tuation in the number of luminous photons emitted during the The NEMA NU-1 document “Performance Measurements
scintillation phenomenon, from the statistical conversion proof Scintillation Cameras” (2001) provides guidelines for
cess of photons into electrons in the photocathode, and from measuring the performance of a gamma camera. This docu-
additional fluctuation of the signal produced by the PMT dur- ment describes the procedures to be followed for measuring
ing multiplication of the electrons. Energy resolution is usu- (1) intrinsic spatial resolution of the detector, (2) spatial lin-
ally expressed as ΔE/E for a given energy, i.e., as fluctuation earity, and (3) energy resolution and uniformity (differential
of the energy signal, measured as the full width at half maxi- and integral), along with the procedures for measuring per-
mum (or FWHM) of the peak in the energy spectrum, divided formance of the combined detector-collimator system. In
by the energy value itself. In a typical clinical gamma camera particular, the intrinsic spatial resolution of the detector (Rγ)
based on the NaI(Tl) crystal and PMT system, energy resolu- is defined as the semi-FWHM of the image reconstructed
tion is approximately 10% at 140 keV. Energy resolution is a from a point source. The value of Rγ depends on the charac-
key parameter for image quality, since a better value of ΔE/E teristics of the scintillator and of the photodetector.
142 R. Pani et al.
In general, Rγ is approximately 3–5 mm, and it depends

on the thickness of the crystal, the number of PMTs per unit
of surface, the positioning algorithm used, and the radius
energy incident γ-rays. It should be emphasized that spatial
resolution improves with increasing energy of the incident
γ-rays and therefore depends on the radioisotope used.
Spatial resolution worsens with increasing thickness of y
the crystal, because of the greater enlargement of the light
spot that reaches the PMTs. On the other hand, thickness of z
the crystal cannot be too small, as it directly affects the
detection efficiency, as described in Eq. (6.1).
x
6.3 ingle-Photon Emission Computed

S
Tomography (SPECT)
The detectors described above are used to obtain projection

images of radionuclide distribution within the source of
interest (body of the patient), each acquisition yielding a
two-dimensional representation on the x,y axes, i.e., a planar Fig. 6.4 Each row of the matrix records views/projections that always
view. A SPECT gamma camera system allows acquisition of come from the same section of the source object (reproduced with per-
multiple planar, angular views of the source of interest yield- mission from Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di
Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010)
ing scintigraphic images that represent radionuclide distribu-
tion in three dimensions.
In planar emission imaging, the intensity observed in a Data from each section of the object is therefore separated
given area of the
image is a function of the amount of radioac- and can be reconstructed independently of each other.
tivity present along a given propagation direction, which is Figure 6.5 summarizes the acquisition of data from a sin-
selected by using an appropriate collimator (see further below). gle slice. For each angular position of the detector, a histo-
The resulting image is therefore a two-dimensional projection gram (profile) is obtained that shows, for each detector
of the actual distribution of activity in the body, which is intrin- element, the intensity of the measured radiation. This will
sically three-dimensional. By changing the angle of the detec- represent the value of the line integer for each flight loop
tor equipped with, e.g., a parallel-hole collimator with respect defined by the detector element. Starting from such intensity
to the patient, it is possible to obtain an image where some profiles, a suitable tomographic reconstruction software
details, which in the previous image were covered by other yields the reconstructed 2D image of the slice itself.
structures, become more obvious. The second image will still Combination of the various slices therefore allows recon-
be planar, i.e., two-dimensional, but from a different angle. struction of a 3D image of the unknown distribution (ρ(x,y,z))
With tomography, the information obtained by measuring at of the radiotracer within the body.
various angles the intensity of the radiations emitted by a cer- A tomographic acquisition is considered “complete”
tain source are suitably combined, by means of a computer, to when the projections have been acquired from a number of
reconstruct the image of a section (or “τομος” in Greek) of the angular positions over an arc of at least 180°. It should be
patient’s body. In a tomographic section, a given area within noted that, after a 180° rotation, the flight directions of the
the image has an intensity that is directly related to the local γ-rays that are acquired are the same as the acquisition at 0°,
radioactivity concentration, and each structure in the section is while what changes is the direction of propagation.
visible because in certain angle(s) it is not covered by other A SPECT tomograph is usually made up of one or more
structures. Therefore, the section (or slice) is not affected by gamma camera detectors (usually two, sometimes three detec-
“overlap” problems, which are typical of planar imaging. tors), each coupled with its own collimator and capable of
In the case of SPECT with parallel-hole collimators, rotating around the patient’s body. The system is connected to
reconstruction of 3D images can be reduced to a 2D problem a computer that must control both data acquisition and mechan-
considering how images of individual projections planar are ical rotation of the detectors. There are two rotation modes
recorded. In fact, for each angular position radioactivity dis- commonly used in SPECT, the so-called “continuous rotation”
tribution is projected onto the detector that “acquires” an mode and the “step-and-shoot” mode. In the continuous rota-
image that typically has a 64 × 64 or 128 × 128 pixel size. tion mode, the detectors are rotated at constant speed around
Thus, each line of the matrix records projections originating the patient, and the data so acquired is then grouped into a finite
always from the same section of the source object (Fig. 6.4). number (typically 64 or 128) of angular values (or projections).
Fig. 6.5 Diagrammatic
representation of the matrix 100
data acquired simultaneously
50
with a three-head gamma
camera for SPECT 0
(reproduced with permission 20 40 60 80 100 120
from Volterrani D, Erba PA,
Mariani G, Eds. Fondamenti
di Medicina Nucleare –
Tecniche e Applicazioni.
Milan: Springer; 2010)
100
50
0
20
40
60
80
100
120
10
0
50
0
20
40
60
80
10
0
12
0
a b c d
Fig. 6.6 Possible configurations of the revelation gamma camera (reproduced with permission from Volterrani D, Erba PA, Mariani G,
heads for acquisition of SPECT data. (a) single head; (b) two opposing Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
heads; (c) two heads at a 90° angle one to another; (d) triple head Milan: Springer; 2010)
In the step-and-shoot mode, the rotational motion stops at vari- increases linearly with the number of detectors used. In par-
ous angular positions (corresponding to the angles for which ticular, assuming a certain number of total events to be
you want to acquire a projection) and remains stationary until acquired during tomography, a three-head gamma camera
the data have not been acquired, after which it moves on to a will complete acquisition in exactly one-third the time
next angle; 3° or 6° angular steps are usually employed. required by acquisition with a single-head gamma camera.
SPECT systems differ from one another according to the Figure 6.6 shows the most common configurations for
number of gamma camera revelation heads, which typically SPECT detection heads: (a) single head, (b) two opposing
can be one to three. The higher the number of detectors used, heads, (c) two heads at 90°, and (d) triple head. Usually, two-
the greater the overall sensitivity of the instrument, which head gamma cameras can vary the position of the two heads
144 R. Pani et al.
in configurations (b) or (c). This is useful when using SPECT especially gold alloys are less frequently used due to the diffi-
as a planar imaging system to simultaneously obtain anterior culty of tungsten processing and the high cost of gold.
and posterior (b) or anterior and lateral (c) views. In case of
the (a) and (b) configurations, a complete tomographic acqui-
sition can only be obtained by rotating the detectors by 180°, 6.4.2 Geometry
while with the (c) configuration, a 90° rotation is sufficient,
and with the (d) configuration, a 60° rotation is sufficient. In Collimator geometry defines the geometric condition of
order to achieve better sampling of the entire field of view, it “acceptance” for incident photons on the gamma camera
is preferable to acquire the same direction of flight from head. Various types of collimators are used, including parallel-
γ-rays along both sides, so as to mediate the effects of attenu- hole collimators, the converging-hole or diverging-hole col-
ation that γ-rays undergo while crossing the body (see further limators, and pinhole collimators. Collimators with specific,
below). For full coverage of the entire field of view, the dedicated applications include the slanted-hole collimator.
detector(s) must rotate by 360°(configuration a), 180° (con-
figurations b and c), or 120° (configuration d). 6.4.2.1 Parallel-Hole Collimators
A parallel-hole collimator consists of a block of γ-ray-
absorbing material where parallel holes are perforated, per-
6.4 Collimators pendicular to the detector. This is the geometry most
commonly used for gamma camera imaging, as it provides
The most important issue regarding intrinsic spatial resolution wide-field views and yields a real-size projection of radioac-
of SPECT is the need for reaching an optimal trade-off tivity distribution within the patient’s body.
between spatial resolution and sensitivity that both depend on Ideally, a parallel-hole collimator allows only the γ-rays
the characteristics of the collimator(s) used. Therefore, that travel in an orthogonal direction to the detector. In prac-
optimization and choice of the collimator(s), also in relation to tice, due to finite size of the section of the holes, the collima-
the specific clinical application, are the crucial factors affecting tor accepts photons with an angle of incidence included
the whole image quality in single-photon emission imaging. within a certain range of values. Therefore, increasing the
Collimators are usually classified according to construction diameter of the holes, as well as reducing their length, wors-
material, geometry, and shape and size of holes. ens spatial resolution. Each hole of a parallel-hole collimator
defines a cone whose angular opening is given by Eq. (6.5):
D
6.4.1 Materials JTot = (6.5)
L
Ideally, a collimator works like a binary filter. When a γ-ray where D and L are the diameter (considered in this case cir-
photon hits the collimator, it can reach the detector only if its cular) and length of the holes, respectively. Therefore, the
flight direction respects the geometric conditions defined by the geometric spatial resolution Rc (estimated as FWHM of the
collimator holes. This “selection” is based on the absorption of point spread function or PSF) is given by Eq. (6.6):
the γ-rays whose flight directions do not meet the same condi-
tions but also on secondary events taking place upon the γ-ray/
(d + L)
Rc = D (6.6)
L
collimator interaction. In fact, some γ-rays that propagate in a
direction not parallel to the collimator’s holes will pass through As a consequence, increasing the distance of a source
the collimator material without interacting (this phenomenon is object from the collimator by increasing the diameter of the
called “penetration”), while others may undergo Compton inter- underlying cone worsens spatial resolution, which instead
actions; in the latter instance, some Compton-diverted photons improves by keeping the source object closer to the collima-
reach the scintillator in spots that do not exactly mirror the origi- tor. Optimal positioning of the parallel-hole collimator holes
nal flight direction. In both instances (penetration and Compton is the position that minimizes the distance of the source
scattering), unintended photons can contribute to formation of object from the collimator surface.
the final image, thus degrading its quality. If the detector behind the collimator has a spatial resolu-
To reduce these phenomena, collimators are usually made tion, Rr, the system will therefore have a FWHM of spatial
of materials with high-absorption power for γ-rays, especially resolution given by Eq. (6.7):
materials with a probability of photoelectric interaction greater
2
than the probability of Compton scatter. In practice, the materi- æ Dö
Rs = Rc2 + Rr2 = ç D + d ÷ + Rr2 (6.7)
als chosen have a high atomic number, Z, and high density, ρ. è Lø
The most commonly used materials are lead alloys (Z = 82,
ρ = 11.3 g/cm3), followed by tungsten alloys (Z = 71, ρ = 19.1 g/ On the other hand, geometric efficiency of a parallel-hole
cm3) or gold (Z = 79, ρ = 19.3 g/cm3). However, tungsten and collimator, ϵc, measured as the fraction of photons that are
Fig. 6.7 Most common shapes of the holes in parallel-hole collima- h = thickness of the septa (reproduced with permission from Volterrani
tors: round holes (left), hexagonal holes (middle), square holes (right). D, Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare –
D = diameter of each hole; S = space between two adjacent holes; Tecniche e Applicazioni. Milan: Springer; 2010)
accepted by the collimator versus the total number of pho- application energies. This simplifies their choice depending on
tons emitted by the source, is given by Eq. (6.8): the use required for a specific diagnostic application, regardless
of the different parameters that a manufacturer can choose to
2
æDö D2 achieve the same or similar characteristics. This classification
c = K 2 ç ÷ (6.8)
èLø ( D + h)
2
refers to the range of γ-rays energy (therefore of radionuclides)
for which the collimator may be used and the best trade-off
where h is the minimum thickness of the septum between the between spatial resolution and sensitivity. The most common
holes and K is a factor that depends on the geometry of the holes. types of parallel-hole collimators are classified as:
It is important to note that the efficiency of a parallel-hole
collimator does not depend on the distance between the • Low-energy, high-resolution (LEHR) collimators are
source object and the collimator. In fact, ignoring the attenu- used with radionuclides that emit γ-rays of low energy, in
ation effects on the emitted γ-rays, all areas within a certain order to achieve high spatial resolution at the expense of
field of view will be measured with the same efficiency. sensitivity.
Typical orders of magnitude of the efficiency values of paral- • Low-energy, all-/general-purpose (LEAP or LEGP) col-
−4
lel hole collimators may vary between 10 and 10 . −6
limators are used with radionuclides that emit γ-rays of
The holes in the collimator are usually made in three low energy, in order to achieve an optimal trade-off
shapes: hexagonal holes in a hexagonal grid, circular holes between spatial resolution and efficiency.
on a hexagonal grid, and square holes in a square grid • Medium-energy, all-/general-purpose (MEAP or MEGP)
(Fig. 6.7). Since the collimator structure is periodic, the col- collimators are used with radionuclides that emit γ-rays
limator’s characteristics are uniquely determined by thick- of intermediate energy, in order to achieve optimal trade-
ness of the holes but also by their shape, size, and separation. off between spatial resolution and efficiency.
Size of the hole is usually indicated by the diameter, D. • High-energy, all-/general- purpose (HEAP or HEGP) col-
While in the case of circular holes, the meaning of D is limators are used with radionuclides that emit γ-rays of
immediate, in the case of hexagonal or square holes, D high energy, in order to achieve optimal trade-off between
expresses the value of the diameter of the circle having the spatial resolution and efficiency.
same area as the hole that is being considered.
As for separation between the holes, the value S indicates Collimators classified as LE (low energy) are usually rec-
the distance between the centers of two adjacent holes. In ommended for 57Co (122 keV), 123I (159 keV), 99mTc
many cases, however, separation between the holes is (140 keV), and 201Tl (69–81 keV); ME collimators are used
expressed in terms of thickness of the septum (h), as this for 67Ga (93, 184, and 296 keV) and 111In (172 and 247 keV);
parameter provides more direct information on the collima- HE collimators are commonly used for 131I (284 and 364 keV).
tor ability to stop the photons traveling in directions that are
oblique with respect to the axis of the holes. In the case of 6.4.2.2 Converging- or Diverging-Hole
hexagonal and square holes, h simply indicates the thickness Collimators
of the material that separates two adjacent holes, while in the For some particular applications, it may be useful to use colli-
case of circular holes, it indicates the minimum thickness. mators that have holes that, instead of being mutually parallel,
For added convenience, parallel-hole collimators for clini- are plotted along convergent or divergent directions.
cal applications are not classified according to shape/size of the Converging-hole collimators magnify the image of the source
holes and thickness of the septa, but simply according to their object on the detector; they can therefore be used only with
spatial resolution and relative efficiency and their range of source objects that have size smaller than the detector. Thanks
146 R. Pani et al.
to their particular geometry, these collimators are inherently size of the object. Although use of these collimators yields
more efficient than parallel-hole collimators. Because of these high spatial resolution (e.g., by exploiting to the maximum
properties (magnification of the image and increased counting magnification and using a pinhole with a very small hole),
efficiency), these collimators may be used to image relatively counting efficiency drops very quickly with increasing dis-
small source objects with increased sensitivity or with better tance between the hole and the source object.
spatial resolution—depending on the design of the collimator. In fact, geometric efficiency is given by the fraction of
Instead, diverging-hole collimators are used in those cases solid angle subtended from the hole from a certain point of
where one wishes to image a source object larger than the the field of view. The geometric efficiency of a pinhole col-
detector. For either converging- or diverging-hole collima- limator εp is given by Eq. (6.9):
tors, the main drawback consists in providing distorted
2
images of the distribution of activity in the field of view, æ Dö 3
since the magnification factor depends on the distance of the
p = ç ÷ sin J (6.9)
è 4d ø
object source from the collimator. This fact is important
especially for planar imaging; a SPECT acquisition can be where D is diameter of the hole, d is the distance between the
corrected through accurate modeling of the collimator to be object source and the hole, and ϑ is the angle that the incident
applied in the algorithm of image reconstruction. photon forms with the plane of the detector, as shown in
Fig. 6.8. Therefore, efficiency is inversely related to the
6.4.2.3 Pinhole Collimators square of the distance (αd−2), thus decreasing as the distance
Especially for high-resolution imaging, pinhole collimators d increases, as well as with increasing from the axis of the
may be used. These consist of a single hole with a double pinhole (as a proportion of sin3ϑ). Figure 6.9a plots the per-
cone shape (Fig. 6.8). The rationale of image acquisition formance efficiency of a pinhole collimator as a function of
with such collimators is similar to that of the darkroom used the distance d of the object from the hole. It should be noted
in photography. In fact, these collimators yield on the detec- that, unlike a parallel-hole collimator, efficiency varies with
tor an image of the source object that is inverted and magni- d and the pinhole collimator is more efficient only at small
fied (or downsized, depending on the object distance of the distances, whereas it drops dramatically for large values of d.
source object from the collimator) with respect to the actual Besides efficiency, distance d from the collimator affects
also the size of the field of view. In fact, the field of view of
a pinhole collimator is defined as the cone projected from the
detector through the pinhole. In particular, the field of view
increases in a linear fashion with distance d, but on the other
hand efficiency is greater for small values of d, where the
field of view is minimal (Fig. 6.9b).
For a pinhole collimator, the FWHM of spatial resolution,
Rp, is given by Eq. (6.10):
(d + L)
Rp = De (6.10)
L

ϑ α Therefore, for values of d much smaller than L, spatial
resolution is practically given by the value of De, which is the
effective diameter of the hole. This value is greater than the
nominal diameter of the hole, because the radiation imping-
ing in the pinhole can penetrate into the material in the vicin-
ity of the edge of the hole, where thickness is a minimum. De
depends not only on the geometric diameter D but also on the
linear attenuation coefficient of the material, μE, constituting
the collimator (and hence the energy E of the incident radia-
tion), and on its defined aperture angle α. The value of De can
be estimated using Eq. (6.11):
L d æ 2 æa öö
De = d ç d + tan ç ÷ ÷ (6.11)
è m è 2 øø
Fig. 6.8 Diagrammatic representation of a pinhole collimator (see
text) (reproduced with permission from Volterrani D, Erba PA, Mariani
G, Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Overall, the FWHM value for spatial resolution of a
Milan: Springer; 2010) system equipped with a pinhole collimator with spatial

Fig. 6.9 Comparison of a b

performances between a
parallel-hole collimator (gray
FOV
continuous line) and a pinhole
collimator (dash black line)
with respect to counting (a),
size of the field of view (b),
and spatial resolution (c) as a
function of the distance “d” of
the source object from the
collimator (reproduced with
permission from Volterrani D,
Erba PA, Mariani G, Eds.
Fondamenti di Medicina
Nucleare – Tecniche e
Applicazioni. Milan:
Springer; 2010) d d
c
R
Parallel-hole
Pin-hole
resolution Rp and coupled with a detection system for γ-rays Pinhole collimators are usually made of lead or tungsten.
with intrinsic spatial resolution Rr is given by Eq. (6.12): However, in order to limit the value of De, use of high atomic
number materials, such as gold or platinum, is preferred. In
2 order to minimize costs, it is sufficient to use such precious
æ (d + L) ö æ d ö
2 2
æ dö
2 metals to make only the central part of the pinhole collimator
Rs = R + ç Rr ÷ = ç De
p ÷ + ç Rr ÷ (6.12)
è Lø è L ø è Lø that contains the hole—called “insert.”

The pinhole collimator is suitable for a wide range of
Figure 6.9c compares the intrinsic spatial resolution of a γ-ray energies, ranging from 20–30 keV (125I) to single pho-
pinhole collimator with a parallel-hole collimator. The pinhole ton imaging of positron annihilation photons (511 keV).
collimator has better resolution for small values of d, where the
efficiency is at a maximum but the field of view is minimal.
For this reason, pinhole collimators are often used for
high-resolution imaging of small source objects, in particu- 6.5 Multimodality Imaging: SPECT/CT
lar for preclinical imaging of small animals. In fact, by
exploiting the magnification effect, high-resolution images The γ-rays emitted from the body undergo variable attenua-
of small objects (small FOV) compared to the size of the tion while traveling from the point of origin to the revelation
detector can be obtained, while maintaining the sensitivity head of the gamma camera. Attenuation is indeed a crucial
that would be obtained with parallel-hole collimators. factor in diagnostic imaging with standard gamma cameras,
Further increased magnification can be achieved with spa- as γ-rays are attenuated according to the exponential law
tial resolutions of the order of a few hundred micrometers, expressed by Eq. (6.1). For example, energy of the γ-rays
even with clinical gamma cameras. By using more holes emitted by 99mTc, the most widely used radionuclide for
(multi pinhole collimator), it is possible to obtain wider single-photon emission imaging, is 140 keV. The linear atten-
fields of view and at the same time to increase the overall uation coefficient in the water (μH2O) for photons with this
sensitivity of the instrument. By reducing the magnifica- energy is approximately 1.53 × 10−1 cm−1 corresponding to a
tion, it is thus possible to obtain high-resolution images of half-thickness (λ½, i.e., the thickness that halves the number
fields of view sufficient to display “total body” of small of emerging photons compared to the original number of pho-
rodents. tons) of approximately 4.5 cm (note that λ½ = ln2/μ).
148 R. Pani et al.
It is therefore obvious that, especially in the case of sort of a tomographic image (similar to CT) that would pro-
SPECT acquisitions, the effect of attenuation is not negligi- vide information on the distribution of attenuation coeffi-
ble if one wants to quantitate absolute radioactivity concen- cients (the so-called μmap); alternatively, theoretical
trations within certain volumes of interest. For example, attenuation coefficients based on phantom data mimicking
while crossing the human body (approximately 40 cm in composition of the body were utilized. In modern SPECT
diameter), 140 keV γ-rays would be attenuated by a factor of gamma cameras correction for attenuation is obtained by
about 450. Considering attenuation, the integral equation coupling the SPECT system with a CT scanner, through the
along a certain line corresponds to Eq. (6.13): use of a multimodal imaging instrument called SPECT/CT.
In this way, the μmap to correct for attenuation can be
ò
- m ( x ,z ,y ¢,E ) dy ¢
obtained directly from the CT data, by suitably rescaling the
Ng ( x,z ) = òr ( x,z ,y ) e L¢
dy (6.13)
values of μ(E) obtained with the X-rays energy of CT to the
L
energy values of the γ-rays used for SPECT.
where L is the line along which the activity ρ(x,y,z) is mea- As an added value, the SPECT/CT systems provide,
sured (integrated) and L′ is the actually traversed line by a together with the functional images of SPECT, the possibil-
photon emitted from point P(x,y,z) and directed along the line ity to overlap each reconstructed γ-ray emission slice onto
L, while μ(x,y,z) is the local value of the linear attenuation the corresponding anatomical structures depicted by the
coefficient for photons with given energy photons E (Fig. 6.10). X-ray transmission scan. This property can greatly assist
Clearly, in order to obtain true distribution of the activity at diagnostic interpretation, by providing complementary mor-
various depths within the body or to be able to provide a quan- phological information that, considering the relatively low
titative estimate, it is necessary to know a priori the distribu- spatial resolution and inherently functional nature of the
tion of the linear attenuation coefficients for the given energy. SPECT image, were not present in the emission image alone.
In the stand-alone SPECT instrumentation, this measure- It should, however, be emphasized that, unlike PET/CT, the
ment was based on acquisitions “in transmission” where a CT component integrated in the SPECT/CT systems usually
radioactive source (linear or planar) was used to obtain some has physical performances inferior to those of a true clinical
CT (i.e., it yields a lower diagnostic value), because the system
is optimized for complementary use of CT with SPECT. These
y systems are therefore usually characterized by poorer spatial
R resolution than clinical CT scanners; furthermore, the CT
component of the scan is optimized so as to confer to the
patient a radiation dose lower than that of clinical CT, there-
fore not significantly increasing the overall radiation dose to
the patient over the radiation burden caused by radiopharma-
ceutical administration for the SPECT scan per se.
L
L
P(x,z)
Key Learning Points
X • The two fundamental modes of single-photon imag-
ing are planar and tomographic.
• The revelation head of a standard gamma camera
consists of a collimator and of a scintillation crystal
optically coupled with a set of photomultiplier
tubes.
• Gamma-rays interact with the crystal by depositing
all or part of their energy, which is transformed into
release of a scintillation photon, through either the
photoelectric effect and through the Compton
Fig. 6.10 Diagrammatic representation of the acquisition process of effect.
γ-ray emission data along a certain projection line in the presence • The light flash produced by interaction of γ-rays in
of attenuation. The emission point P(x,z) contributes to the integral of the scintillation crystal is converted into an electri-
counts along line L, subject to an attenuation factor due to crossing
of the emitted photons along line L′ (reproduced with permission from
cal signal by a photodetection system composed of
Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina a set of photomultiplier tubes.
size, easy handling, and close positioning are required.

• Signals from individual photomultipliers are pro- Furthermore, the standard Anger camera suffers from intrin-
cessed by an electronic system and transferred to a sic poor spatial and energy resolution.
computer for subsequent analysis allowing defini- Interest has recently emerged in developing compact
tion of the x,y coordinates of the point of the scintil- gamma cameras with improved imaging performances. This
lation crystal where interaction of the incident γ-ray has led to design new gamma cameras mainly falling into
with the crystal matrix takes place. three categories: (1) semiconductor detectors, such as cad-
• The fundamental parameters that characterize the mium telluride (CdTe) or cadmium-zinc telluride (CdZnTe),
imaging performance of a standard gamma camera in which the γ-rays interact directly with a position-sensitive
are spatial resolution and energy resolution. solid-state detector; (2) scintillation crystals coupled to an
• Gamma cameras based on solid-state ionization use array of solid-state photodetectors (photodiodes or silicon-
semiconductor detectors that achieve better energy avalanche photodiodes); or (3) scintillation crystals coupled
resolution and better intrinsic spatial resolution with a position-sensitive PMT (PSPMT).
than scintillation detectors. By using scintillation crystal arrays coupled with PSPMT,
• The parameters that define the imaging performance compact gamma cameras with high spatial resolution and
of gamma cameras include spatial resolution, spatial relatively low cost can be assembled. Finally, smaller
linearity, energy resolution, and uniformity. devices, based on the same principles, but scaled down to
• SPECT images provide information on radionu- intraoperative dimensions, have been successfully developed
clide distribution within the body based on acquisi- as imaging probes for use during radioguided surgery [2]; in
tion of multiple planar views at variable angles fact, conventional gamma cameras (even single-head—but
around the patient’s body. especially multiple-head gamma cameras) cannot be housed
• SPECT images can be produced using either single- in the surgical suite.
head, two-head, or triple-head gamma cameras. The above considerations have started the idea of using a
• Collimators are made of high-absorption power for different specific device for each diagnostic application,
γ-rays and are designed to select γ-rays with direc- instead of only one general-purpose instrument—not unlike
tions defined by certain geometric conditions. the variety of diagnostic equipment of a modern radiology
• According to geometry of the holes, collimators are division [8].
defined as parallel-hole collimators, converging-hole Intrinsic spatial resolution of an imaging system such as
or diverging-hole collimators, and pinhole collimators. the Anger camera, which is based on a single large planar
• Collimators with specific, dedicated applications detector, is worse than that of collimator for a short source-
include the slanted-hole collimator. to-collimator-face distance. Furthermore, the coarse image
• The most common types of parallel-hole collimators digitization (usually 3.2 mm) places additional limits to the
include low-energy, high-resolution collimators; low- resolution and contrast values that can be achieved.
energy, all-/general-purpose collimators; medium- On the other hand, advancing knowledge in biology at the
energy, all-/general-purpose collimators; and molecular level and the development of new radiopharma-
high-energy, all-/general- purpose collimators. ceuticals have generated demands for imaging devices as
• Hybrid SPECT/CT allow multimodality imaging additional or alternative tools to the Anger camera, in order
that has the advantages of providing morphologic to optimize radionuclide imaging for specific biological and
anatomic correlates for the sites of radiopharma- clinical requirements. Technological advances in the fields of
ceutical uptake and of allowing correction for actual photodetectors and of scintillators, as well as in electronics,
attenuation of γ-rays based on the transmission data encouraged the development of dedicated imagers for new
obtained by the X-ray beam. diagnostic techniques in quite all the fields of functional
radionuclide imaging. The main feature of such new detec-
tors is a compact design that allows optimized anatomical
fitting. Furthermore, such detectors are suitable for com-
6.6 Dedicated Gamma Cameras bined use with other diagnostic imaging modalities, so as to
provide co-registered and fused images where intrinsic spa-
Conventional Anger cameras using NaI(Tl) scintillation tial resolution of the γ-ray emission image becomes compa-
crystals coupled to a bulky array of PMTs are precluded rable with the values achieved with radiologic imaging
from use in all the applications where light weight, small (around 0.4 mm).
150 R. Pani et al.
6.7 asic Components of Dedicated

B higher stopping power for γ-rays than the NaI(Tl) crystals.
Gamma Cameras and Small FOV Nevertheless, BGO crystals have lower light output and
Imaging Probes higher refractive index at emission wavelength than NaI(Tl)
crystals, thus being less suitable for scintillation light read-
6.7.1 Scintillation Materials out. After having been originally proposed for applications
such as image intensifier screens, BGO-based detectors have
The basic components of general-purpose gamma cam- actually become a standard for positron emission tomogra-
eras, dedicated gamma cameras, and probes for radiogu- phy (PET).
ided surgery intrinsically include a detector system to
convert into an electrical pulse the charge collected in the 6.7.1.4 YAP(Ce)
interaction between the γ-ray emitted by the source of Although spectral data on YAP(Ce) were also published in
interest and the exploring probe. After preliminary selec- 1973, initial studies were really geared to laser applications.
tion of the incoming γ-rays through suitable collimation, It was only in 1991 that the first results on the use of YAP(Ce)
the first material that interacts with γ-rays and initiates the for the detection of ionizing radiation were published [12].
whole detection process is either a scintillation crystal The light efficiency of YAP(Ce) crystals (that are not hygro-
(usually inorganic) or a semiconductor material. scopic) is about 40% the efficiency of NaI(Tl) crystals.
As mentioned above, NaI(Tl) is the first in a long series of Because of the yttrium atomic number (Z = 39) and the rela-
scintillation materials suitable for standard gamma cameras; tively high density (5.37 g/cm3), YAP(Ce) has a good γ-ray
on the other hand, its use is not suitable for small-area pix- absorption.
elated cameras because of drawbacks due to temperature-
depending cracking effect which hampers the light output
from small-sized crystals. Among others, thallium-doped 6.7.2 Other γ-Ray Detectors
cesium iodide [CsI(Tl)], bismuth germanate (BGO), and
cerium-activated yttrium aluminum perovskite [YAP(Ce)] Solid-state detectors, also known as semiconductor ioniza-
have been introduced into clinical use, as they do not suffer tion detectors, utilize CdZnTe as the material that interacts
from the cracking effect. with the incoming γ-rays to produce an electrical pulse.
Semiconductor detectors of clinical interest include cad- Although these detectors are more expensive than the most
mium telluride (CdTe) and cadmium-zinc telluride (CdZnTe), common NaI(Tl) scintillation crystal used in standard
currently used in gamma-probes for radioguided surgery and in gamma cameras, they offer several important advantages. In
dedicated gamma cameras, e.g., for cardiac imaging or for particular, their density is higher than that of NaI(Tl) (5.8 g/
small field-of-view (FOV) imaging. cm3 versus 3.7 g/cm3), thus resulting in greater stopping
power for the incoming γ-rays. Furthermore, they are charac-
6.7.1.1 NaI(Tl) terized by improved energy resolution (thus reducing by
Robert Hofstadter (Nobel Laureate of 1961) described in 30% the scatter counts in the selected energy window) and
1948 the luminescence from a small NaI(Tl) crystal under superior intrinsic spatial resolution. However, in order to
γ-ray irradiation; this observation started up a long and bright keep the overall cost of such detectors comparable to the cost
career of the scintillator which continues up until today [9]. of NaI(Tl)-based detectors, CdZnTe detectors are usually
In fact, this discovery constituted the milestone marking the kept thinner (around 5 mm) than conventional NaI(Tl) crys-
beginning of new horizons for nuclear medicine, leading to tals (around 9–10 mm); therefore, with these parameters, the
the development in 1956, only 8 years later, of the Anger intrinsic sensitivities of the two types of detectors are
gamma camera [1]. comparable.
6.7.1.2 CsI(Tl)
A couple of years after describing the scintillation properties 6.7.3 Arrays of Scintillation Detectors
of NaI(Tl) crystals, Robert Hofstadter reported similar obser-
vations with other luminescent materials, such as CsI(Tl), Clusters of individual NaI(Tl) scintillation detectors arranged
which has subsequently been used particularly for pixelated around a target for detecting the γ-rays produced by (α,n)
imaging probes or for small FOV gamma cameras but also as reactions under irradiation of a target by a beam of α parti-
a detector in nuclear physics [10]. cles have been utilized in nuclear physics since 1963 [13]. A
similar approach was adopted by Hal O. Anger, who in 1953
6.7.1.3 BGO developed a whole-body rectilinear scanner made of a linear
The luminescence properties of BGO under X-ray irradia- array of individual NaI(Tl) scintillation detectors supported
tion were first described in 1973 [11], with an emphasis of its by a structure able to move along the body of a patient [14].
The next step was the development of a large-area mono- The PSPMT still works according to the Anger principle,
lithic NaI(Tl) scintillation crystal whose readout is per- with the additional feature of using scintillation arrays with
formed event by event by an array of PMTs. pixel bases less than 1 mm, thus potentially achieving sub-
An approach to improve spatial resolution consists in the millimeter spatial resolution. Moreover, it is possible to
segmentation of the scintillation crystal aiming to strongly choose low-cost geometries with optimized detection fea-
reduce the spread of light distribution following the absorp- tures for specific applications, e.g., for breast or brain
tion of a γ-ray in the scintillator. Studies for upgrading the imaging.
early prototypes of PET scanners were started in the 1980s. Furthermore, the compact design allows a reduction of
In particular, an array of BGO scintillation crystals, readout shielding weight; for instance, for a 140 keV 99mTc energy
by four PMTs located at vertices of the light-output window γ-ray, a 20 × 20 cm FOV camera requires about a 12 kg
of the array, were considered by Casey and Nutt for building weight lead shielding. Table 6.1 compares the main charac-
a single block detector [15]. Setup of PET block detectors in teristics of the three generations of Hamamatsu PSPMTs [2],
the current commercial scanners is nowadays substantially while Table 6.2 summarizes the main physical characteris-
unchanged, even if more effective block designs have been tics and operating features of initial prototype imaging
developed, based on arrays of scintillators coupled to probes based on inorganic scintillators and PSPMTs.
PSPMTs [16]. Composite detectors based on an array of
BGO crystals have been developed also for gamma cameras 6.7.4.1 First-Generation Hamamatsu PSPMTs
with SPECT capabilities [17]. The first 3″ PSPMT (model R2486 developed in 1985) was
based on proximity mesh dynode; thereby, the charge was
multiplied around the position of the light photon striking on
6.7.4 Position-Sensitive Photomultiplier the photocathode. Figure 6.11 schematizes the dynode struc-
Tubes (PSPMTs) ture of such PSPMT with the whole internal multiplication
electrodes setup. In this PSPMT, the charge cascade was
The major steps toward the development of PSPMTs can be affected by a wide intrinsic spread. A number of factors
marked as (1) invention of the photoelectric tube in 1913, (2) affect such spreading, such as the interstage voltage of dyn-
development of the first PMT in 1930, and (3) development odes (in particular between cathode and first dynode), addi-
of the PMT with electrostatic focusing, the basic structure of tional focusing grids between dynodes, and finally the
current PMTs, in 1936. intrinsic spreading of the light spot due to the photocathode
Table 6.1 Main characteristics of the three generations of Hamamatsu PSPMTs

Parameter 3″ R2486 5″ R3292 R5900-M64 R7600-C8/C12 Flat panel
Generation First First Second Second Third
Window thickness (mm) 3.2 6 1.5 0.8 2.8
Dynode structure Proxi mesh Proxi mesh Metal channel Metal channel Metal channel
Dynode number of stages 12 12 12 11 12
Anodes structure Crossed wire Crossed wire Multi-anode Crossed plates Multi-anode
Number of anodes 16 + 16 28 + 28 64 (8 × 8) 4 + 4/6 + 6 64 (8 × 8)a
256 (16 × 16)b
Pixel size/pitch (mm) 3.75 3.75 2/2.25 5/5.5/3.6 5.6/6.0a
2.8/3.0b
Active area (mm) 50 Ø 100 Ø 18.1 × 18.1 22 × 22 49 × 49
Outer area (mm) 76 Ø 132 Ø 26 × 26 26 × 26 51.7 × 51.7
PMT thickness (mm) 55 113 20.1 20.1 15.5
Intr. charge spread (FWHM, mm) 7 11 <1 <1 <1
Packing density (%)c 43 57 48 71.6 90
Cathode luminous sensitivity (mA/Lm) 80 80 70 70 80
Gain 105 105 3 × 105 7 × 105 3 × 106
Anode dark current (nA) 20 40 12.8 2 96
Transit time (ns)d 17 – 5 5 5
Uniformity among anodes 1:4 1:4 1:5 1:4 1:3
a
H8500 model
b
H9500 model
c
Effective area/outer size
d
Time response
152 R. Pani et al.
Table 6.2 Main characteristics and applications of prototype imaging probes based on inorganic scintillators and PSPMTs
Spatial resolutionb
Scintillator and size (mm) PSPMT Lead collimator, specificationsa (FWHM) Imaging test
NaI(Tl) R2487 LEHR, 25.4 mm thick, Ø 0.1 mm holes, 2.0 mm Guinea pig lungs and
60 × 60 × 4 thick 76.2 square 0.02 mm septa 140 keV heart
NaI(Tl) R2487 Diverging holes, 13 mm thick, Ø 1.3 mm, 2.5 mm for 99mTc Rat bone scan
3 thick 76.2 square 18° half angle 4.0 mm for 111In
CsI(Tl) R2486 Pinhole 0.5 mm 1.0 mm for 99mTc Rat cerebellum
Ø 50 × 0.5 Ø 76.2
NaI(Tl) R3292 Parallel-hole LEHR 2.5 mm Thyroid phantom
76.2 × 76.2 × 5 Ø 125 122 keV
YAP(Ce) R2486 Pinhole 0.5 mm 0.7 mm for 99mTc 0.4 mm collimated
11 × 22 array Ø 76.2 source
0.6 × 0.6 × 7 thick
CsI(Tl) R2486 Pinhole 0.5 mm 1.0 mm for 99mTc Guinea pig brain
Ø 50 × 0.5 Ø 76.2
NaI(Tl) R2487 Square parallel holes, 21.5 mm high, 2.4–3.8 mm for 99mTc 500 g rat,
73 × 73 × 8 76.2 square 1.2 mm pitch, 1.02 mm opening child hand
CsI(Tl) R2487 Flood fields: 22.1 4.5 mm Commercial baggage
16 crystal linear array 76.2 square (keV) 32.1 3.5 mm scanning
1.25 × 1.25 59.5 1.9 mm
0.25 septa 122 1.3 mm
NaI(Tl) R2486 – 1.0 mm Green plant
Ø 50 × 1.0 Ø 76.2
LEHR Low-energy high-resolution
a
FWHM Full width at half maximum

b
Fig. 6.11 Diagrammatic PHOTON Scintillation array

representation of the charge PHOTON Scintillation array
hν Photocathode
multiplication setup used by Photocathode
Hamamatsu Photonics for
first-generation PSPMTs 1st dynode
Focusing mesh
Focusing mesh and
and grid type dynodes
grid type dynodes
X anode
X anode Y anode
Y anode last dynode
optical glass window. Typical charge spreads are 9 mm and small organs such as the thyroid, breast, or brain.
14 mm FWHM for a 3″ and a 5″ PSPMT, respectively [18]. Unfortunately, because of the quite large dead zone at the
Such intrinsic wide charge spreading favors the use of edge of this PSPMT (1 cm or more), it is not possible to
crossed-wire anodes and the centroid method for position assemble arrays of tubes as in the Anger camera.
determination, rather than application as hodoscope where Figure 6.12 compares first- and second-generation
pixelated anodes with minimum crosstalk are required. Hamamatsu PSPMTs. Models R6870 (8″), R3292 (5″),
Furthermore, a wide intrinsic charge distribution can limit R5900 (1″ square), and R2486 (3″) are shown from left to
the spatial resolution values of small light spots. right. A 3.5” floppy disk and a BNC adapter are shown in
The first-generation PSPMT (which is housed in glass) is foreground and on the right of the R5900 PSPMT, respec-
characterized by a large active area (5″ in diameter), which is tively, for size comparisons (the R2486 PSPMT is repre-
useful for applications in physics, but is not large enough for sented with a reduced scale). The major limitation of the
medical imaging, where a 20 cm FOV is required even for large-area, 8″ PSPMT (about 20 cm) was thickness of the
Fig. 6.12 Hamamatsu
PSPMT models belonging to
first and second generations
(reproduced from: Pani R,
Pellegrini R, Soluri A, De
Vincentis G, Scafè R, Pergola
A. Single photon emission
imaging by position sensitive
PMT. Nucl Instr & Meth A
1998;409:524-8, Figure 2)
photocathode glass window (7.5 mm), which produces fur- photocathode glass window thickness (0.8 mm) and the
ther broadening of the light spot on the photocathode and, as extensible detection area with regular or irregular shapes dur-
a consequence, on the intrinsic spread of charge. ing the camera design. On the other hand, the relevant frac-
tion of non-active area (30%) requires the use of an additional
6.7.4.2 Second-Generation Hamamatsu PSPMTs light guide (usually 3 mm thick) to allow detection in dead
The second-generation PSPMT was based on metal channel zones between neighboring photodetectors and reduction of
dynode technology for charge multiplication, combined with light loss. This partially reduces the high imaging perfor-
a photocathode glass window less than 1 mm thick. The mance of individual PSPMTs.
intrinsic light spread was then reduced down to 0.5 mm Figure 6.13 shows two prototypes of detectors for small
FWHM. The new dynode structure consists of electron mul- FOV gamma cameras based on assemblies of 4 × 4 (left) and
tiplier layers that channel the charge, thus saving the position 6 × 7 (right) Hamamatsu second-generation R7600-C12
of the light spot interacting on the photocathode. The charge PSPMTs. The corresponding FOVs are about 10 × 10 cm
is collected by a multi-anodic structure made of an array of (100 cm2) and 15 × 18 cm (270 cm2), respectively, suitable
anodes lying on the same plane or a crossed-plate lying on for optimized anatomical fitting for specific organ imaging.
two parallel planes for X and Y charge collection. Contrary
to the first generation, the narrower charge distribution 6.7.4.3 Third-Generation Hamamatsu PSPMTs
potentially increases the required number of anodes for opti- The third generation started with the H8500–H9500 Flat
mum light sampling [19]. The Philips XP1700 PSPMT series Panel PSPMT families [21]. The major advance is repre-
was never commercialized with optimized peripheral dead sented by housing that compacts a 2″ active area with the
zone, and Hamamatsu Photonics has remained the only cor- narrowest dead boundary (<1 mm), thus assuring a fraction
poration actively working in this field. of active area >80%. These Hamamatsu models use the same
The intrinsic charge spread produced during the multipli- metal channel dynode technology and collect the charge with
cation process slightly amplifies the scintillation light distri- 8 × 8 or 16 × 16 anode array, respectively. With a photocath-
bution at the photocathode, and a 1-mm-thick anode could be ode glass window thickness of 2 mm, the H8500 family
redundant and thus involve an oversampling of light distribu- model would correctly sample the light spread produced by
tion and useless additional readout electronics. Hamamatsu a 2 × 2 mm scintillation array. In order to clearly identify the
R7600-C12 series (6X + 6Y crossed-plate anodes) and pixels of a scintillation array with 1 × 1 mm crystals, the
R5900-M64 series (8X × 8Y anode array) have anode granu- optimum anode size would be 3–4 mm, as in the case of
larity well optimized, to 3 mm and 2.25 mm, respectively. the H9500 Flat Panel PMT family. This would compensate
Such light sampling size allows good identification of scin- the worse spatial resolution due to the glass window thickness
tillation crystal pixels with side as small as 1 mm (C12) or with a better position linearity.
less (M64), without relevant image distortions for 140 keV Figure 6.14 shows the miniaturization steps from a large
γ-rays. Finally, the adoption of a metal housing allows very detection area first-generation Hamamatsu 5″ PSPMT
compact size (about 1 in.3) and reduction of the peripheral (model R3292 on the right) to a third-generation H8500 Flat
dead zones down to 2 mm. Panel (left, an array of 2 × 2 tubes is shown). The multiplica-
A novel gamma camera, based on multiple assembly of 1″ tion structure is visible through the broken glass of R3292.
PSPMTs of second generation, was proposed in 1997 [20]. Overall, glass window areas are of about 127 cm2 (R3292)
Advantages of this novel camera include a reduced overall and 103 cm2 (array of 2 × 2 H8500 tubes). The flat panel
154 R. Pani et al.
Fig. 6.13 Two prototypes of

multiple-crystal small FOV
gamma cameras based on
assemblies of 4 × 4 (left) and
6 × 7 (right) Hamamatsu
second-generation PSPMT
model R7600-C12
6.7.5 G
eiger-Mode Avalanche Photodiodes
(G-APD)
The G-APD can be defined as a solid-state device able to

detect a single light photon emitted under illumination [23].
It should be noted that the properties of such solid-state
detectors have mainly been explored in the field of telecom-
munications, because PMTs used for fiber-optic coupling
transmission have been burdened by important drawbacks
for applications in medical imaging, e.g., gain instability and
high cost. Nevertheless, the high-cost issue and other limita-
tions have been addressed with the latest generation G-APDs,
which can currently be utilized for compact gamma cameras
as well as for PET/MR instrumentation, with some advan-
tage with respect to conventional detection devices.
Fig. 6.14 Miniaturization steps for large detection area Hamamatsu
PSPMT from first-generation 5″ Ø R3292 model (right) to third-generation
H8500 Flat Panel (left, an array of 2 × 2 tubes is shown). The multiplica-
tion structure is visible through the broken glass of R3292. Overall glass
Key Learning Points
window areas are of about 127 cm2 (R3292) and 103 cm2 (array of 2 × 2
H8500 tubes). HV-printed circuit boards are shown at respective bottoms. • Dedicated gamma cameras developed for specific
A 2-EUR coin is included in the picture for size reference applications fall into three categories: (1) semicon-
ductor detectors in which the γ-rays interact directly
with a position-sensitive solid-state detector; (2)
PMT is designed to overcome the limitation affecting the scintillation crystals coupled to an array of solid-
prior PSPMT generations. In fact, it can be assembled in state photodetectors; or (3) scintillation crystals
arrays with an improved effective area up to 97% and a pho- coupled with a position-sensitive photomultiplier
tocathode glass window thickness down to 2.8 mm. tube.
An additional anode configuration made of 16X + 16Y • Compact gamma cameras with high spatial resolu-
crossed-plate, similar to the former model R5900-C8, has tion can be assembled.
recently been proposed. Among the advantages of the • Smaller devices scaled down to intraoperative
crossed-plate layout with respect to the individual anodes, dimensions have been successfully developed as
one can note a dramatic lowering of readout electronics (32 imaging probes for use during radioguided surgery.
channels for 16X + 16Y versus 256 channels for 16X × 16Y) • Different materials are used for either scintillation
and a relevant reduction in the spatial density of pins in the detectors or semiconductor detectors.
PMT base (1.6 pin/cm2 for 16X + 16Y versus 9.8 pin/cm2 for • Small field-of-view imaging probes based on posi-
16X × 16Y). This configuration overcomes the limitations to tion-sensitive photomultiplier tubes have been
the improvement of anodic granularity due to the electrical developed.
connections constraints [22].
6.8 Single-Photon Emission remained unsatisfactory for smaller lesions (T1a). It should
Mammography (SPEM) be emphasized that SPEM with the new small FOV detectors
allows acquisition of scintigraphic images in similar views
Diagnosis and treatment of cancer at an early stage of develop- and operating conditions (breast compression) as during
ment improve long-term outcome [2]. For breast cancer, X-ray X-ray mammography.
mammography currently constitutes the main method of early Over the last two decades, several prototypes for SPEM
cancer detection; however, its diagnostic accuracy is subopti- imaging based on PSPMT technology and pixelated CsI(Tl)
mal, and many patients undergo unnecessary biopsies. Some of scintillation crystals have been developed. Cameras with
the equivocal/inconclusive findings of mammography can be about 11 cm in diameter useful FOV have been described,
clarified by exploiting the tumor-seeking properties of scinti- with spatial resolution around 2–3 mm when the camera is in
mammography (SM) with 99mTc-Sestamibi [24, 25]. As origi- contact with the source object—as it occurs for imaging dur-
nally described by I. Khalkhali in 1993 [24], prone ing breast compression [29] (Fig. 6.16). Furthermore, mild
scintimammography (PSM) consists in positioning the detec- compression of the breast allows separation of overlapped
tion head of a standard gamma camera in vertical position, lat- objects that would otherwise not be distinguished during
eral and as close as possible to the body, with the patient in PSM with a standard gamma camera (Fig. 6.17). The effi-
prone position and the breast hanging. Although the initial cacy of breast compression optimal imaging parameters is
results were encouraging, there was a critical detection limit for shown in Fig. 6.18, where the respective spectra are plotted.
breast lesions smaller than 1 cm in diameter. Reduced sensitivity The large tumor size combined with the large variations in
for smaller cancers is mainly due to the inadequate positioning breast thickness (4.5 cm) shows an impressive increase of
capability of a standard gamma camera head, which is too counts in the specific energy peak with respect to the broad
bulky for optimal breast imaging. In particular, inadequate Compton peak.
spatial resolution (typically 3 mm) and the fact that major In early prototypes, the relatively small FOV limited opti-
portions of the breast are positioned at a distance >5 cm (and up mal imaging of the breast. To overcome this limitation, novel
to 15 cm) from the collimator surface dramatically reduce gamma camera has been proposed based on the Anger prin-
breast lesion contrast in PSM [26]. ciple, but replacing conventional PMTs with compact
On the other hand, as shown in Fig. 6.15, a dedicated PSPMTs. Larger prototypical modular gamma cameras,
imager can easily reduce the breast-to-collimator distance, made of up to a 7 × 8 array of 1″ PSPMTs, have been devel-
thus increasing the sensitivity of SM for lesions much smaller oped worldwide by different research groups. In this way, it
than 1 cm in size, such as T1a and T1b cancers. In this regard, was possible to exploit the narrow scintillation light distribu-
it was shown that use of a small FOV gamma camera with tions, recovering at the same time the position information
very high intrinsic spatial resolution (1.7 mm), located closer for events falling in the dead space between neighboring
to the tumor by breast compression, results in increased sen- PMTs. The main advantages of such modular design are (1)
sitivity (up to 80%) for tumors ≤1 cm in size, without appar- minimum detector thickness (<3 cm), (2) light weight, (3)
ently reducing specificity [27, 28]; nonetheless, the results small dead boundary zone (3 mm), and (4) no limitation in
the active area shape and size.
Moderate compression 70
60
ANGER CAMERA
SPEM
50
40
SNR
3 cm 30
ANGER
20
10 Lower visibility
0 imit
SPEM 0 1 2 3 4 5 6
Tumor size (cm)
10 – 15 cm
Object – collimator distance Fig. 6.16 Plot of the signal-to-noise ratio (SNR) values as a function
of tumor size, obtained from clinical measurements made with a SPEM
Fig. 6.15 Main geometric arrangements for prone scintimammogra- camera in cranio-caudal view and with an Anger camera (lateral view of
phy (PSM, left) and single photon emission mammography (SPEM, the body, with the patient in prone position and the breast hanging),
right) (reproduced from: Garibaldi F, Cisbani E, Cusanno F, Iommi R, respectively (reproduced from: Pani R, Pellegrini R, De Vincentis G,
Urciuoli GM, Pani R, Pellegrini R, Scafè R, Indovina L, Cinti MN, Cinti MN, Weinberg IN, Soluri A, et al. Factors affecting cancer detect-
Trotta G. Optimization of compact gamma cameras for breast imaging. ability in Tc-99m MIBI scintimammography. Nucl Instr & Meth A.
Nucl Instr & Meth A. 2003;471:222-8, Figure 1 (a,b)) 2003;497:90-7, Figure 2)
156 R. Pani et al.
a b
Fig. 6.17 Right breast carcinoma, 7 mm in size: 99mTc-sestamibi circle) (reproduced from: Pani R, Pellegrini R, De Vincentis G, Cinti
images obtained with a SPEM camera without breast compression (a) MN, Weinberg IN, Soluri A, et al. Factors affecting cancer detectability
and with mild breast compression (b). Mild compression results in bet- in Tc-99m MIBI scintimammography. Nucl Instr & Meth A.
ter definition of the lesion as an area with increased tracer uptake (red 2003;497:90-7, Figure 4)
1200
1000
Compressed breast
800
Counts
600
400
Uncompressed breast
200
0
0 20 40 60 80 100 120
Channels
Fig. 6.18 Energy spectra for the same tumor (4.5 cm in size) acquired
with and without mild breast compression (reproduced from: Pani R, Fig. 6.19 Prototype of a breast-dedicated SPECT/CT scanner (repro-
Pellegrini R, De Vincentis G, Cinti MN, Weinberg IN, Soluri A, et al. duced from: Gambaccini M, Fantini A, Bollini D, Castelli E, Del
Factors affecting cancer detectability in Tc-99m MIBI scintimammog- Guerra A, Di Domenico G, et al. Development of a quasi-monochro-
raphy. Nucl Instr & Meth A. 2003;497:90-7, Figure 5) matic CT system for breast cancer study with combined emission-trans-
mission tomography. IEEE Trans Nucl Sci. 2001;48:703-6, Figure 2)
A compact-design combined SPECT/CT imaging camera Commercial cameras for SM, currently better identified as
has also been proposed to overcome the limitations of poor “breast-specific gamma imaging” (BSGI) or “molecular breast
specificity for stand-alone X-ray mammography and poor imaging” (MBI), are mostly based on solid-state detector tech-
sensitivity of SM [30] (Fig. 6.19). The reconstructed images nology. Both single-head and dual-head BSGI are available,
strongly enhanced scintigraphic contrast for tumors <1 cm in with useful FOV up to 16 × 20 cm. Intrinsic spatial resolution
size with respect to planar geometry. These results demon- using tungsten collimators is generally better than 2 mm, with
strate the feasibility of exact 3D localization of tumor lesions energy resolution values around 5% (FWHM) for 140 keV pho-
using single-photon compact ring tomography combined tons (the typical γ-ray emission of 99mTc). Mild breast compression
with X-ray transmission tomography. is performed during scintigraphic acquisition, and the detector(s)
right anterior oblique to left posterior oblique. By position-

ing the two detectors at a 90° angle, the two detectors can
simultaneously acquire usable data over the 180° orbit; if the
two detectors were in the usual opposed position (180°),
only one of the two detectors would actually acquire usable
data. Dedicated cardiac SPECT scanners with two scintilla-
tion detectors fixed at a 90° angle have been built by Philips
(CardioMDTM) and by Siemens (CCamTM).
Siemens has also introduced the so-called “cardio-focal”
(SmartZoom™) field-upgradeable collimators for high-
sensitivity cardiac imaging with conventional dual-head
gamma cameras [6, 7]. Also known as “astigmatic” collima-
Fig. 6.20 Setup of a commercial breast-dedicated gamma camera that tion, in such converging/diverging collimator, the converging
can be equipped with either one or two revelation heads based on
CdZnTe detectors. Mild breast compression is performed by adjusting collimation magnifies the central FOV in both the axial and
the distance between the two detectors (or between the single detector transaxial directions, while the diverging collimation mini-
and the opposing breast holder) according to size of the breast in the fies the periphery of the FOV to provide coverage of the
individual patient. The detector(s) can be positioned in any position to entire body (Fig. 6.21), thereby avoiding truncation artifacts
acquire scintigraphic views mimicking the projections adopted for
X-ray mammography, from orthogonal to oblique; an oblique view is commonly seen with fan-beam collimators when imaging
shown here (image downloaded from “www.cmr-naviscan.com/ the torso. Such cardio-focal collimators increase sensitivity
lumagem/”) for cardiac imaging by a factor of approximately two com-
pared to high-resolution parallel-hole collimators.
can be rotated to r eplicate the standard orthogonal and oblique The Digirad Cardius 3 XPO™ cardiac scanner employs
views/projections as during X-ray mammography, thus three pixelated CsI(Tl) scintillation detectors and images the
facilitating comparisons between the two sets of images patient in an upright seated position (Fig. 6.22); one- and
(conventional mammography and BSGI/MBI) (see Fig. 6.20). two-detector Cardius systems are available as well. The sys-
tem uses fan-beam collimators in the three-detector configu-
ration coupled to photodiode detectors. The two outer
detectors are positioned at a 67.5° angle relative to the cen-
Key Learning Points tral detector. Data is acquired by rotating the chair (rather
• The signal-to-noise ratios obtained by SPEM than detectors) by 67.5°, yielding an overall acquisition arc
acquired during mild breast compression are greater of 202.5°. Its SPECT spatial resolution (FWHM) is 11.0 mm
of about a factor two versus the corresponding value (radius of rotation: 20 cm) and sensitivity 160 cpm/μCi for
obtained during scintimammography. 99m
Tc using LEHR collimation.
• SPEM with breast-dedicated gamma cameras can
efficiently visualize tumors less than 1 cm in size,
with signal-to-noise ratios ranging from 7 to 20.
• Mild breast compression allows discrimination of
overlapped lesions.
6.9 Cardiac-Dedicated Gamma Cameras

diverging
When employed for cardiac SPECT studies, conventional
gamma cameras generally utilize two scintillation detectors
fixed at a 90° angle to one another; some (older generation) converging
gamma cameras are also equipped with adequate radioactive
sources for transmission imaging-based, patient-specific Fig. 6.21 Data acquisition geometry with the Siemens SmartZoom™
attenuation correction. Due to the left anterior position of the cardio-focal (diverging/converging) collimator. It provides higher sen-
sitivity in the heart region by selective maximal magnification of that
heart in the chest, cardiac SPECT imaging typically involves
region (reproduced from: Slomka PJ, Berman DS, Germano G. New
acquisition of projection data over only a 180° orbit (instead cardiac cameras: single-photon emission CT and PET. Semin Nucl
of the 360° orbit used for general SPECT imaging)—from Med. 2014;44:232–51)
158 R. Pani et al.
Fig. 6.22 The Digirad Cardius 3 XPO™ system. The patient is seated upright, and the chair rotates to achieve the required angular sampling (repro-
duced from: Slomka PJ, Berman DS, Germano G. New cardiac cameras: single-photon emission CT and PET. Semin Nucl Med. 2014;44:232–51)
The Digirad Cardius X-ACT™ system augments the (2.46 × 2.46 × 5 mm thick) arranged in a 16 × 64 element
Cardius 3 XPO™ features by adding a fan-beam X-ray array with a size of 40 × 160 mm, achieving 16 cm of cov-
source for CT imaging. The CardiaArc HD-SPECT™ [31] erage of the thorax. Each detector array is fitted with
also images patients in an upright seated position. It is con- square-aperture, high-sensitivity, parallel-hole collimators
stituted of three curved stationary NaI(Tl) crystals backed by such that the dimensions of each hole are matched to the
three rows of 20 PMTs, with a thin curved-lead collimator size of a single detector element. Spectrum Dynamics also
(the so-called aperture arc) having six vertical slots for hori- recently introduced a less expensive six-detector column
zontal collimation and rotating slowly back and forth every system. In another CdZnTe-based camera design devel-
10 s. Spatial resolution of CardiaArc™ is 3.6 mm at an oped by General Electric (Discovery NM 530c™ and
82 mm source-to-detector distance and 7.8 mm at a 337 mm Discovery NM/CT 570c™), a curved array of 19 5-mm-
source-to-detector distance. thick CdZnTe pixelated detector arrays is employed, each
Two vendors have introduced CdZnTe-based cardiac with a focused pinhole collimator (Fig. 6.24). Each detec-
SPECT systems. The Spectrum Dynamics D-SPECT™ tor array consists of four detectors, each with 246 CdZnTe
utilizes pixelated CdZnTe detector arrays mounted in nine detector elements (2.46 × 2.46 × 5 mm thick) arranged in a
vertical columns, with four detectors in each column, posi- 16 × 16 element array. In these two systems, the detector
tioned in a 90° gantry geometry (Fig. 6.23). Each column gantry does not rotate or otherwise move. The remarkably
consists of an array of 1024 CdZnTe elements high sensitivity of these systems is not related to the intrin-
a b
c d
Fig. 6.23 Spectrum Dynamics D-SPECT™ system. (a) Detector con- nine detector blocks. (d) One of the CdZnTe detectors (reproduced
figuration (top view) with the nine angularly directed detector blocks. from: Slomka PJ, Berman DS, Germano G. New cardiac cameras: sin-
(b) Photograph of the system. The patient may be seated upright, or the gle-photon emission CT and PET. Semin Nucl Med. 2014;44:232–51)
chair reclines so that the patient may lie nearly supine. (c) One of the
sic sensitivity of the CdZnTe detectors, but rather to the

high-sensitivity geometric arrangement and collimation of Key Learning Points
the detectors. • Cardiac-dedicated gamma cameras utilize special
The above cardiac-dedicated cardiac scanners provide geometry of the detector head(s) to obtain tomo-
higher sensitivity (up to sevenfold) and better spatial resolu- graphic images (SPECT) without having to perform
tion than conventional gamma cameras, as illustrated in detector head rotation.
Fig. 6.25. Imaging times with these scanners thus range from • Cardiac-dedicated scanners provide higher sensitiv-
2 to 3 min for systems with solid-state detectors and 4 min ity (up to sevenfold) and better spatial resolution
for scintillation-detector systems with cardio-focal than conventional gamma cameras.
collimation.
160 R. Pani et al.
a b
Top view Side view
c d
m
6m
2.4
Fig. 6.24 General Electric NM 530c™ system with nine detector limation. (d) One of the CdZnTe detectors (adapted from: Slomka PJ,
blocks arrays. (a) Detector configuration (top and side views). (b) Berman DS, Germano G. New cardiac cameras: single-photon emis-
Photograph of the system. The patient lies supine. (c) The pinhole col- sion CT and PET. Semin Nucl Med. 2014;44:232–51)
Sensitivity Resolution
0.10% 18.00
0.09% 0.09% 15.0 15.3

15.00
0.08%
0.07%
Resolution [mm]
12.00
Sensitivity [%]
0.06%
9.2*
0.05% 8.6 9.00
0.05%
6.7 0.04%
0.04%
0.03%
6.00
0.03%
0.02%
0.01% 3.00
0.01%
0.00% 0.00
T
T
us
l
M
na
C
EC
di
E
io
SP
SP
ar
nt
C
IQ
ve
D
on
C
Fig. 6.25 Sensitivity in % (left axis) and reconstructed image resolu- Digirad X∙ACT™ system with fan-beam collimators; conventional,
tion in mm (right axis) of dedicated cardiac SPECT systems compared conventional dual-headed SPECT with low-energy high-resolution col-
to those of a conventional dual-head system. D-SPECT, Spectrum limators (reproduced from: Slomka PJ, Berman DS, Germano G. New
Dynamics D-SPECT™ system; NM, GE 530c/570c™ system; cardiac cameras: single-photon emission CT and PET. Semin Nucl
IQSPECT, Symbia system with dual-head Siemens SmartZoom™ Med. 2014;44:232–51)
cardio-focal (diverging/converging) collimators; Cardius, Cardius
6.10 Handheld Gamma-Probes

for Radioguided Surgery Counting
system
Handheld gamma-probes for radioguided surgery have been Scintillator crystal or

Collimator semiconductor detector
used since the late 1940s, constituted by Selverston-Robinson
Electronics system:
needle-shaped detectors (a Geiger-Müller-type device) PMT or amplifier
employed to assess completeness of tumor resection after
administration of 32P-orthophosphate to patients with brain Handle
FOV
tumors [32].
At that time, the primary device available to count radio-
activity in the earliest nuclear medicine applications was the
Fig. 6.27 Schematic of a traditional scintillation or semiconductor
Geiger-Müller (GM) tube, the basic radiation detector gamma-probe
(together with the ionization chamber) used in nuclear plants.
GM tubes were commercially available, particularly in min-
iaturized models, well suitable for ionizing radiation probes 6.10.1 Scintillation Probes
(Fig. 6.26).
After several years of apparent oblivion, new-design An early scintillation probe was developed in the late 1950s
handheld gamma-probes for radioguided surgery became [33], based on the use of a CsI(Tl) crystal. The scintillation
commercially available in the mid-1980s, initially operating light output was read out remotely by using optical fibers or
as single-channel counting devices yielding a counting value light pipes, because the available PMTs had a minimum
(or range) of event rate and the corresponding audible/visible diameter of 1” [34]. The light losses were reduced when reli-
signals, whose intensity/color was modulated according to able 10-mm-diameter PMTs became commercially available
the count rate. for more efficient gamma-probe assemblies, which also
The family of single-channel, or counting, probes includes allowed better energy resolution and background rejection
devices based on a plurality of physical principles for detec- by pulse-height windowing [35].
tion ionizing radiation, all of them basically providing a count- The term “background” used during radioguided sur-
rate value in the examined region of interest, that can be gery does not refer to environmental radiation (as com-
explored in the open surgical field after administration of a monly referred to for conventional radionuclide imaging),
diagnostic radiopharmaceutical suitable for accumulation/ but rather to the quantity of radioactivity present in the
concentration and retention in the tumor lesion to be resected. healthy tissues surrounding the target tissue to be resected.
The region to be explored may be delimited by the short range Design of the distal, radiation-sensitive portion of handheld
of radiations in the tissues, as in the case of the β− particles gamma-probes for radioguided surgery is critical, as it must
emitted by 32P detected by a GM-type tube (the earlier probe), constitute the best design trade-off between absorption of
or by a suitable collimator absorbing γ-rays, as in the currently background γ-rays by the collimator (depending on mate-
available scintillation or semiconductors probes (Fig. 6.27). rial and thickness) and its aperture that defines the probe
sensitivity.
Furthermore, appropriate electronics must assure a coarse
spectrometry (due to the poor energy resolution of the probe)
for rejecting gammas scattered in the body and connects the
selected signals to a rate meter, which produces audible sig-
nals and digital readouts.
Sensitive
Region
6.10.2 Semiconductor Probes
Solid-state semiconductor detectors quickly replaced

GM-like tubes in the few applications where miniaturization
Fig. 6.26 Photograph of the earliest Geiger-Müller (GM) tubes used was required [36]. These detectors are characterized by good
as intraoperative probes (Courtesy of Melvin Griem, MD, Professor energy resolution in a very small size [35]. As an example of
Emeritus, University of Chicago, assistant to C. Robinson in 1948). The extreme miniaturization, Lauber developed a single detector
sensitive region is about 3 mm in diameter, and only about 1–1.5 cm at
the tip is sensitive to radiation (reproduced from: Hoffman EJ, Tornai
placed inside a needle of 1.1 mm in diameter, as well as an
MP, Janecek M, Patt BE, Iwanczyk JS. Intraoperative probes and imag- array of multiple detectors arranged inside a needle with
ing probes. Eur J Nucl Med. 1999;26:1913-35, Figure 1) 1.5 mm diameter [37].
162 R. Pani et al.
6.10.2.1 Cadmium Telluride Detectors sized 0.6 × 0.6 mm, 7 mm high [40]. The array setup includes
Due to the higher atomic number values of cadmium and tel- optical insulation between the crystals with 5-μm-thick
lurium (Z = 48 and Z = 52, respectively) with respect to sili- reflective layers and is optically coupled with the R2486
con (Z = 14), CdTe detectors are better suited for low-energy Hamamatsu PSPMT. The final spatial resolution was about
γ-rays (e.g., 140 keV from 99mTc) than Si detectors, which 0.7 mm at 140 keV. Instead, Truman et al. [41] used CsI(Tl)
are instead more appropriate for β− detection [43]. crystal arrays to develop a small FOV SPECT device with
Nevertheless, if interactions occur deeply in the detector, spatial resolution of 1.3 mm at 122 keV (extrapolated to
a fraction of ionization charge is not recorded, thus produc- 1.1 mm at 140 keV).
ing long tails at low energy of the photo peak with a relevant
loss of energy resolution. Notwithstanding these limitations,
Key Learning Point
CdTe (as well as CdZnTe and CdHgTe) are mostly used for
• Imaging probes are mostly used for small field-of-
probes at lower-energy photon detection, like those emitted
view imaging of the surgical bed, but not for actual
by 125I (27–37 keV, the ideal energy range for CdTe detec-
intraoperative guide.
tors). In these conditions, a very thin collimator can be highly
effective and cheap.
6.11.1 Generalities on the Use of Portable

Key Learning Points Probes for Radioguided Surgery
• Single-channel (counting) probes include a wide
category of devices based for detection of ionizing The use of portable probes (either single-channel counting
radiation, all substantially providing a count-rate probes or dedicated small FOV imaging probes) for
value in the examined region/volume of interest. radionuclide-based detection and localization of tumors,
• Handheld gamma-probes are employed intraopera- especially small tumors, has several well-known limitations
tively for radioguided surgery. [42], as follows:
• The absolute tumor uptake of tumor-seeking radiophar-

maceuticals is generally quite low, typically ~0.1% or less
6.11 Imaging Probes of the administered activity per gram.
• The overall sensitivity of radiation detection in vivo is
Several applications of radioguided surgery, typically also quite low, ranging from about 0.1% for gamma cam-
sentinel lymph node biopsy—especially if performed in
era imaging (including SPECT) to ~10% for PET.
intracavitary environments (e.g., during laparoscopic sur- • Such low counting efficiency is further exacerbated by the
gery)—are considering the use of dedicated imagers instead signal-degrading effect of attenuation of emitted radiation
of a standard Anger camera or single-channel gamma-probes by overlying tissue.
that only detect or count radioactivity. For image-guided sur- • A significant fraction of the counts apparently emanating
gery, a combined approach is generally adopted: (1) preop- from a tumor or other targeted tissue may actually include
erative scintigraphy with a standard gamma camera, followed counts originating elsewhere (i.e., from background activ-
by (2) intraoperative use of a handheld counting gamma- ity in adjacent tissues) because of contrast- and resolution-
probe. Several prototypes of miniaturized gamma cameras degrading Compton scatter.
have been developed, based on the use of PSPMTs such as
the second-generation Hamamatsu R5900-C8; these devices The above limitations can be mitigated, at least in part,
are called “imaging probes” (IPs) and are characterized by through the use of intraoperative probes and, potentially, intra-
high resolution (around 2–3 mm FWHM, with 20% window operative gamma cameras. In fact, because of the close prox-
centered on the 140 keV peak energy of 99mTc), full portabil- imity that can be achieved at surgery between a collimated
ity with light weight for intraoperative use, and a useful FOV detector and the target tissue to be surgically resected (e.g., a
of at least 1 in.2 [38, 39]. tumor or a sentinel lymph node following their adequate pre-
Early small FOV gamma cameras were based on inor- operative radiolabeling), radionuclide detection of such struc-
ganic scintillators coupled to 3″ or 5″ crossed-wire anode tures can be enhanced using easy-to-handle intraoperative
Hamamatsu PSPMT for scintillation light-output readout gamma-probes or small FOV, dedicated gamma cameras.
(Table 6.2). The setups generally involved monolithic crys-
tals, although systems based on 2-D and 1-D [40, 41] 6.11.1.1 Gamma-Probes
achieved spatial resolution values of around 1 mm. In par- The most widely used type of intraoperative probe (e.g., for
ticular, Pani et al. used an array of 11 × 22 YAP(Ce) crystals, sentinel lymph node detection) is the general-purpose gamma-
Side View
Detector
Detector Thickness
Detector Width Collimator Length
Collimator
End View
Fig. 6.28 The general design and operating principles of an intraopera- probe itself is connected to a control unit (lower left panel), which typi-
tive gamma-probe. The handheld probe (upper left panel) is comprised cally provides both a visual readout of the count rate and an audible
of a collimated, small-area (typically ~1 cm in diameter) scintillation or signal related to the count rate, with the frequency of the latter signal
solid-state ionization (i.e., semiconductor) detector (right panel). The increasing or decreasing in relation to the detected count rate
probe (Fig. 6.28), designed for counting local activity of radio- Collimated and
nuclides emitting X- and/or γ-rays; such single- photon shielded detector
emitters include 99mTc, 111In, 123I, and 131I [43]. Gamma-probes
generally use inorganic (i.e., non-plastic) scintillation detec- a
tors or solid-state (i.e., semiconductor) ionization detectors
and lead or tungsten for shielding and collimation. Scintillators
used in such probes include NaI(Tl), CsI(Tl), CsI(Na), and
cerium-doped lutetium ortho- oxysilicate [LSO(Ce)]. b
Semiconductors used for intraoperative gamma-probes
include CdTe, CdZnTe, and mercuric iodide (HgI2).
Although scintillation detectors are characterized by high
c
sensitivity, while the solid-state semiconductor detector
allows better energy resolution (and therefore better scatter Minimally
rejection), in the clinical routine use, the two types of probes Collimated and
generally provide comparable performance. Positron emitters shielded detector
such as fluorine-18 (18F) may also be counted with such
Fig. 6.29 Comparative thickness of gamma-probe collimation and
probes by single-photon (i.e., non-coincidence) counting of shielding required for low- to medium-energy X- and γ-rays of single-
the 511 keV annihilation photons. However, as shown in photon emitters (a), for the high-energy (511-keV) annihilation γ-rays of
Fig. 6.29, this requires thicker collimation and shielding to positron emitters (b), and for negatrons and positrons of beta-particle
minimize penetration of such high-energy photons emitted emitters (c). Note the much thicker collimation and shielding required for
counting of the annihilation photons and the minimal collimation and
from outside of the FOV (as defined by the collimator aper- shielding for counting of beta-particles. Courtesy of IntraMedical
ture) from reaching the detector and thereby degrading spatial Imaging, Los Angeles, CA (reproduced from: Heller S, Zanzonico
resolution as well as target-to-background contrast [44, 45]. P. Nuclear probes and intraoperative gamma cameras. Semin Nucl Med.
2011;41:166–81)
164 R. Pani et al.
Beta-Probes ratio of the measured sensitivities for 511 keV photons is the

Since X- and γ-rays penetrate relatively long distances (of weighting factor by which the Detector-2 count rate is multi-
the order of 10 cm) of soft tissue, a major limitation of the plied and then subtracted from the Detector-1 count rate to
use of gamma-probes to specifically identify the target tissue yield the Detector-1 positron-only count rate. This probe was
during radioguided surgery is the presence of variable, gen- evaluated using the phantom setup shown in the figure, with
erally high levels of background activity in surrounding nor- a small 18F-containing capsule simulating a tumor positioned
mal tissues. Thus, even with a gamma-probe positioned over within a uniform 18F-filled cylindrical container simulating
the skin projection of a tumor, the contribution of counts underlying normal tissue activity. The probe was then
originating from activity in normal tissue underlying the scanned across the phantom, and the Detector-1 and
tumor and even outside the field of view (due to a certain Detector-2 count rates at lateral positions relative to the
degree of penetration of the collimation and shielding) may “tumor” (i.e., capsule) were recorded; the results obtained in
degrade the tumor-to-normal tissue contrast (e.g., reduce the terms of the measured count rates and target-to-background
tumor-to-normal tissue count ratios to less than 1.5:1 [44, ratios (particularly the dramatic improvement from ~2 to
46] and thus tumor detectability to the point where lesions ~10 in the target-to-background ratios) clearly demonstrate
may be missed). the feasibility of this dual-detector design and weighted-
A potential solution to this limitation of radioguided sur- subtraction algorithm for beta-probes in general and for pos-
gery is the use of so-called beta-probes, that is, intraoperative itron probes in particular.
probes which specifically yield counts of negatrons or posi-
trons. Since they have very short ranges in soft tissues (typi- 6.11.1.2 Intraoperative Gamma Cameras
cally of the order of 1 mm or less), the β particles emitted by Although the sensitivity and specificity for sentinel lymph
activity outside the probe’s FOV or underlying the surface node detection using current approaches such as preopera-
tissue do not reach the detector and are not counted. By the tive gamma camera imaging, intraoperative handheld
same token, minimal if any collimation and shielding is gamma-probes, and the “blue dye” technique are quite high,
required (Fig. 6.29c). Therefore, the discrimination between there remains a need to develop techniques to improving sen-
higher-activity tumor and lower-activity normal tissues is sitivity and reducing the false-negative rates of sentinel
enhanced, and the tumor-to-normal tissue count ratios are lymph node detection. Intraoperative small FOV gamma
increased. Of course, the very short range of beta-particles in camera imaging may provide the improvement required to
water/tissues restricts the use of such probes to surface satisfy such requirements. A gamma camera system having a
lesions; beta-probes could not be used, for example, for (per- spatial resolution of 3 mm or better at a distance (depth) of
cutaneous) detection of sentinel lymph nodes. Beta-probes the order of 1 cm would likely visualize such problematic
generally utilize either semiconductor or plastic scintillator nodes intraoperatively. Such an imaging system would offer
detectors, since such detectors have lower effective atomic other practical advantages over the handheld, non-imaging
numbers and mass densities than inorganic scintillators such gamma-probes, as follows:
as NaI(Tl) and lower intrinsic efficiencies for X- and γ-rays,
thus minimizing the potentially confounding count contribu- • The signal is provided in the familiar format of a scinti-
tion of any such radiations accompanying beta-particle emis- graphic image rather than a numerical display or a sound
sion [46, 47]. with variable frequency tone/pitch.
Daghighian et al. [48] have developed and evaluated a • The larger FOV of even small dedicated gamma cameras
plastic scintillator-based positron probe whose design is rep- (several cm) than that of gamma-probes (<1 cm) allows
resented in Fig. 6.30. The basic design of this dual-detector more rapid exploration of large areas and/or acquisition
probe includes two scintillator detectors, a central solid- of more counts and therefore reduction in statistical
cylinder detector (designated “Detector-1”) and a uncertainty (noise).
hollow-cylinder detector (designated “Detector-2”) in 1-mm- • More straightforward re-examination of the surgical site
thick stainless steel cladding; the outputs of the two detectors post-lymph node excision to verify removal of foci of
are transmitted by fiber-optic cabling to separate PMTs activity.
(PMT 1 and PMT 2, respectively). The Detector-1 counts • Less reliance on potentially obliterated and otherwise
result from both positrons and the 511-keV annihilation pho- ambiguous preoperative skin markings directing where
tons associated with positron emission, while the stainless measurements are to be performed intraoperatively [49].
steel cladding of Detector-2 completely attenuates the posi-
trons and allows only the high-energy annihilation photons Thus, intraoperative small FOV gamma camera systems
to enter the detector and generate counts. Because of the dif- merit development and evaluation. A number of small FOV
ferent shielding, sensitivities of Detectors 1 and 2 for the intraoperative gamma camera systems have been developed
511 keV photons are different. The Detector-1-to-Detector-2 [43, 50, 51]. The earliest systems were handheld devices
a To PMT 2 c
2800
To PMT 1
2400
Detector 1 count rate
Fiberoptic 2000
(counts/5 sec)
Count Rate
cables
1600
1200 Detector 2 count rate
800
Metallic Weighted difference
jacket 400 of Detector-1 and -2 count rates
Detector 2
0
–40 –30 –20 –10 0 10 20 30 40
Lateral Position of Probe Relative to Capsule
(mm)
Detector 1
b d
10
9 For weighted difference
Plastic film of Detector-1 and -2 count rates
Beta (50 µm thick) 8
Capsule-to Background
Scanning direction
probe
Count-Rate Ratio
7
6
18
F-containing capsule: 5
0.5-cm diameter x 0.5-cm long 4
(0.5 µCi/mL) For Detector 1 count rate
3
Uniform 18F 2
background activity 1
(0.05 µCi/mL)
0
in 20-cm diameter x 3-cm high cylinder –40 –30 –20 –10 0 10 20 30 40
Lateral Position of Probe Relative to Capsule
(mm)
Fig. 6.30 (a) Basic design of a dual-detector beta-probe. (b) Detector-1 and Detector-2 count rates (see text) as a function of the
Experimental phantom setup for evaluation of the performance of the lateral position of the probe relative to the capsule. (d) The capsule-to-
probe shown in (a). The phantom consisted of a small 18F-containing background-background count-rate ratios for Detector-1 with and with-
capsule simulating a tumor and a uniform 18F-filled cylindrical source out weighted subtraction (reproduced from: Daghighian F, Mazziotta
simulating underlying normal tissue activity. The probe was then JC, Hoffman EJ, Shenderov P, Eshaghian B, Siegel S, et al. Intraoperative
scanned across the phantom. Note that the capsule-to-background beta probe: a device for detecting tissue labeled with positron or elec-
activity concentration ratio was 10:1. (c) The measured Detector-1 and tron emitting isotopes during surgery. Med Phys. 1994;21:153–7)
Detector-2 count rates and the calculated weighted difference of the
having FOVs of only 1.5–2.5 cm in diameter and used con- device is light enough (820 g) to hold for a short time (up to
ventional NaI(Tl) or CsI(Tl) scintillation detectors. ~1 min). The CdZnTe detector has a 3.2 × 3.2 cm FOV and
Subsequent systems used two-dimensional arrays (mosaics) is 5 mm thick, with an efficiency of 87% and energy resolu-
of scintillation crystals connected to a PSPMT and, more tion of 9% for 99mTc γ-rays. Its collimators are easily
recently, semiconductors such as CdTe or CdZnTe detectors. exchanged. The CZT crystal is divided into a 16 × 16 array
The main problems with these early units were their very of 2 × 2 mm pixels, with integral and differential uniformi-
small FOV and the resulting large number of images required ties of 1.6% and 1.3%, respectively, with low-energy high-
to explore the surgical field, associated with the difficulty in resolution (LEHR) collimation. The system spatial resolution
holding the device sufficiently still for the duration (up to with the LEHR collimation is 2.3 mm, 8.0 mm, and 15 mm
1 min) of image acquisition. More recently, larger FOV FWHM at source-to-collimator distances of 1 cm, 5 cm, and
devices have developed which are attached to an articulating 10 cm, respectively. As shown in Fig. 6.31, this camera is
arm for convenient and stable positioning. These systems able to clearly image sentinel lymph nodes as well as lym-
remain nonetheless fully portable and small enough overall phatic vessels. However, the small 3.2 × 3.2 cm FOV remains
to be accommodated in typical surgical suites. limiting. For example, a single lymph node occupies nearly
The eZSCOPE (Anzai Medical, Tokyo, Japan) is a hand- half of the FOV, and thus searching the surgical field can be
held CdZnTe-based semiconductor gamma camera. The time-consuming.
166 R. Pani et al.
Sentind Lymph Node Lymphatic vessed

a b
c
(eZ-SCOPE Image (eZ-SCOPE Image
Fig. 6.31 (a) Photograph of the eZSCOPE intraoperative gamma cam- (top left) and lymphatic vessel in a patient (top right) and conventional
era (Anzai Medical, Tokyo, Japan). (b) Sample eZSCOPE image of a gamma camera image of the same patient obtained preoperatively (bot-
lymph node in a patient. (c) Sample eZSCOPE images of a lymph node tom). See text for additional details
General Electric (Haifa, Israel) developed a CdZnTe- FWHM at 1 cm, 2 cm, and 5 cm, respectively. Images are
based semiconductor gamma camera with a FOVof 4 × 4 cm, acquired in a 50 × 50 pixel matrix. However, energy resolution
larger than that of the eZSCOPE. The 4 × 4 cm pixelated of the POCI is rather poor (28%), and the wide energy win-
detector consists of a 16 × 16 array of 2.5 × 2.5 mm pixels. dows thus required result in inclusion of substantial amounts
Using parallel-hole collimation, the spatial resolution is of scatter in the image, a particular disadvantage when a lymph
5 mm FWHM at a distance of 5 cm, with a sensitivity of node is close to the injection site. In a preliminary clinical
100 cps/MBq. Energy resolution for 99mTc is 8.0%, some- study, lymph nodes in all three patients were identified with
what better than the ~10% value typically quoted for conven- the POCI gamma camera, including one in whom two deep
tional gamma cameras. In phantom experiments, this system nodes were missed with a gamma-probe (most likely due to
could clearly resolve 99mTc-filled spheres 1 cm in diameter in depth-related loss of sensitivity and proximity of the nodes to
contact with one another at distances of up to 6 cm; a gamma- the injection site). The total imaging times depended upon the
probe could only distinguish the two sources at a 1 cm depth scan area and varied from 15 s to 3 min.
and only when separated by at least 2 cm. A potential advan- Tsuchimochi and colleagues in Japan [53] developed
tage of gamma camera imaging is the ability to resolve another semiconductor gamma camera utilizing a CdTe
sources that may overlap one another in one view by acquir- detector. Their choice of CdTe was based on its superior uni-
ing additional views at different angles, as illustrated by the formity (4.5% integral uniformity) and energy resolution
results of the phantom experiment shown in Fig. 6.32. This (7.8%) than CdZnTe. The gamma camera, referred to as the
feature may be helpful, especially in breast cancer, for “small semiconductor gamma camera (SSGC),” uses an array
localizing a sentinel lymph node at a different depth from the of 32 × 32-mm-thick CdTe elements, with a 1.2 × 1.2 mm
injection site and obscured by the injected activity. pixel matrix and a 4.5 × 4.5 cm FOV. The tungsten collimator
POCI (“Per-operative Compact Imager”) is a small FOV has 1.2 × 1.2 mm square apertures to match the pixel arrange-
gamma camera developed in France. This handheld camera ment. Spatial resolution without scatter is 3.9 mm, 6.3 mm,
utilizes a CsI(Na) scintillation crystal coupled to a focusing and 11.2 mm FWHM at 2.5 cm, 5 cm, and 10 cm, respec-
image intensifier tube and position-sensitive diode [52] tively. Sensitivity for 99mTc at the surface without scatter is
(Fig. 6.33). Its FOV is 4.0 cm in diameter. With high-resolution 300 cps/MBq, comparable to that of the POCI and better than
parallel-hole collimation, its 99mTc sensitivity with scatter is that of a conventional gamma camera with LEHR collimation
250 cps/MBq at 1 cm and 125 cps/MBq at 5 cm; the corre- (~100 cps/MBq). The results of preliminary phantom and
sponding spatial resolution is 3.9 mm, 4.8 mm, and 7.6 mm clinical imaging studies with the SSGC are encouraging.
a Sentinella 102 has been developed by General Equipment

for Medical Imaging, South America. It uses a single 4 × 4 cm
CsI(Na) scintillation crystal and a PSPMT, images being
acquired in a 300 × 300 matrix. Interchangeable pinhole aper-
tures 1.0, 2.5, and 4.0 mm in diameter are available, yielding an
effective FOV of 20 × 20 cm at a distance of 18 cm. The detector
assembly weighs 1 kg and is mounted on an articulating arm
(Fig. 6.34). Sensitivity for 99mTc ranges from 200 to 2000 cps/
1 μCi at 1 cm and 60–160 cps/μCi at 10 cm, depending on the
3
2 pinhole aperture used. The FWHM spatial resolution over the
detector face is 5.4–8.2 mm, 7.3–11 mm, and 10–18 mm at
b 3 cm, 5 cm, and 10 cm, respectively, again depending on the
pinhole aperture used. Beyond ~3 cm, therefore, spatial resolu-
tion is poorer than that of gamma cameras with parallel-hole
1 2 3
collimation. However, despite the coarser resolution, the advan-
tage of a pinhole collimation lies in the larger effective FOV at
such distances. Such a system can therefore be used at large
distances to rapidly survey, with coarser resolution, a large area
Scale and then examine suspicious areas at smaller distances and finer
2 cm
resolution [54–56]. In initial clinical studies, acquisition times
of 20–60 s per image were required. Because the distortion
Fig. 6.32 Setup and results of a 99mTc phantom imaging experiment associated with pinhole collimation varies with position within
with the intraoperative gamma camera developed by General Electric the FOV (i.e., is worse toward the periphery) as well as distance,
(Haifa, Israel). (a) Schematic diagram (side view) of phantom, with two
1 cm spheres at depths of 2 and 4 cm and a third sphere, 1.4 cm in diam- the Sentinella 102 camera is equipped with a laser positioning
eter, at a depth of 2 cm and directly over the 2 cm sphere at a depth of system, two intersecting lines being projected onto the surface
4 cm. The two small spheres and the large sphere had activity concen- of the region being imaged. This allows positioning of suspi-
trations of 1 μCi/mL and 2.5 μCi/mL, respectively. (b) Images (identi- cious foci of activity at the center of the FOV, where image qual-
fied as “1,” “2,” and “3,” respectively) were acquired at angles of −45°,
0°, and +45° relative to an axis perpendicular to the top of the phantom ity is best. The Sentinella 102 camera is also equipped with a
(i.e., an axis in the plane of the diagram). The resulting gamma camera long-lived 153Gd pointer for real-time positioning; the image of
images demonstrate the ability to resolve overlying foci of activity by the 153Gd pointer source is acquired in a separate energy window
acquiring views at multiple angles (reproduced from: Kopelman D, from the 99mTc image and is displayed as a small marker super-
Blevis I, Iosilevsky G, Reznik A, Chaikov A, Weiner N, et al. A newly
developed intra-operative gamma camera: performance characteris- imposed on the 99mTc image.
tics in a laboratory phantom study. Eur J Nucl Med Mol Imaging. CarolIReS [57] is an intraoperative gamma camera devel-
2005;32:1217-24) oped by the Institut Pluridisciplinaire Hubert Curien
a b c
Fig. 6.33 (a) Photograph of the POCI (Per-operative Compact Imager) to two neighboring lymph nodes (reproduced from: Pitre S, Menard L,
intraoperative gamma camera. (b) Intraoperative lymphoscintigraphy, Ricard M, Solal M, Garbay JR, Charon Y. A hand-held imaging probe for
with the POCI in position for imaging of the patient’s left axilla. (c) POCI radio-guided surgery: physical performance and preliminary clinical
image (10-s acquisition time), showing two foci of activity corresponding experience. Eur J Nucl Med Mol Imaging. 2003;30:339–43)
168 R. Pani et al.
a b c
Fig. 6.34 (a) Photograph of the Sentinella 102 small FOV gamma cam- lymphoscintigraphy was performed to identify the popliteal sentinel
era (General Equipment for Medical Imaging, South America). The lymph node. (c) Posterior gamma camera images of the patient’s right
detector assembly is shown mounted on the system’s articulating arm. knee joint before (left) and after (right) surgical excision of the popliteal
(b) Illustration of the device’s laser positioning system, with two inter- node. The pre-excision image (left) clearly shows the node centered in
secting red lines projected onto posterior surface of the patient’s right the field of view, and the post-excision image (right) is notably absent of
knee joint. This patient had malignant melanoma of the right heel, and any such focus of activity, demonstrating complete removal of the node
(Strasbourg, France). This device has a relatively large-area terize suspicious foci of activity. Finally, the scintigraphic
50 × 50 mm cerium-doped gadolinium ortho-oxysilicate image format is familiar to surgeons, thus facilitating clinical
(GSO(Ce)) scintillation crystal and is equipped with a parallel- acceptance and integration of intraoperative imaging.
hole collimator with 2-mm-wide apertures. Its 99mTc spatial Intraoperative PET scanners. Prescient Imaging LLC is
resolution is 10 mm FWHM at 5 cm, sensitivity is 130 cpm/ currently developing a portable compact whole-body PET
kBq, and energy resolution is 45%. A prototype version of this scanner adaptable to intraoperative imaging [59]. It incorpo-
device with a larger 100 × 100 mm FOV has been developed rates a 360° detector system with an axial FOV of 22 cm. The
as well. In a preliminary clinical study with the CarolIReS detectors are assembled into three sets, one planar and the other
camera, Mathelin et al. [58] compared the depth of lymph two circular arcs of 90° each connected with a hinge. One of
nodes estimated by imaging to their actual depth measured at the arcs is fixed, while the other can be rotated about the hinge
surgery and found a generally good correlation except in and thereby opened and closed. The planar detector is fixed
instances where only a portion of the sentinel lymph node was horizontally, such that it can fit under a patient table. By placing
in the camera’s FOV. For 7/11 lymph nodes whose depth could the moveable arc in the closed position, a 360° of detector cov-
be estimated, the image-derived depth was correct. erage about the patient is achieved. Attenuation correction can
Overall, small FOV gamma cameras have demonstrated be performed using a rotating rod transmission source.
detection rates for sentinel lymph nodes equal to or better
than those of non-imaging gamma-probes, despite the fact
that their sensitivities (e.g., expressed in cps/MBq) are typi- Key Learning Points
cally about tenfold lower than those of the handheld on- • Handheld gamma-probes for radioguided surgery
imaging gamma-probes. The ability of such devices to image have been developed for a wide variety of
a surgical field intraoperatively and thus verify complete applications.
excision of lesions is a potentially useful enhancement of the • The detector component of handheld gamma-
radioguided surgery techniques employed in cancer patients. probes for radioguided surgery can be either a scin-
The acquisition times per image are typically well under tillation crystal or a solid-state detector.
1 min, so the overall duration of the surgical procedure is not • Beta-probes designed for special applications based
significantly prolonged. Furthermore, despite having lower on the use of PET radiopharmaceuticals have also
sensitivity than parallel-hole collimation, the use of pinhole been developed.
collimation permits initial imaging at longer distance to • Small-field-of-view gamma cameras for use during
visualize a larger anatomic area of interest, followed by surgery are also commercially available.
imaging at shorter distance to pinpoint and otherwise charac-
Glossary 7. Slomka PJ, Pan T, Berman DS, Germano C. Advances in SPECT

and PET hardware. Prog Cardiovasc Dis. 2015;57:566–78.
8. Fong Y, Giulianotti PC, Lewis J, Koerkamp BG, Reiner T, editors.
APD Avalanche photodiode Imaging and visualization in the modern operating room. A com-
BGO Bismuth germanate prehensive guide for physicians. Berlin: Springer; 2015.. ISBN:
CsI(Na) Sodium-activated cesium iodide 978-1-4939-2325-0.
9. Hofstadter R. Alkali halide scintillation counters. Phys Rev.
CsI(Tl) Thallium-activated cesium iodide 1948;74:100–1.
FOV Field of view 10. Hofstadter R. Scintillation counters; Research reviews. Arlington,
FWHM Full width at half maximum VA: Office of Naval Research.. 4 (September, 1949):4-8. Also in
G-APD Geiger-mode avalanche photodiode Nucleonics. 1950;6:70.
11. Weber MJ, Monchamp RR. Luminescence of Bi4 Ge3 O12: spec-
GM Geiger-Müller tral and decay properties. J Appl Phys. 1973;44:5495–9.
HEAP High-energy all-purpose 12. Baryshevsky VG, Korzhik MV, Moroz VI, Pavlenko VB, Fyodorov
HEGP High-energy general-purpose AA, Smirnova SA, et al. YAlO3:Ce fast-acting scintillators for
IP Imaging probe detection of ionizing radiation. Nucl Instr Meth B. 1991;58:291–3.
13. Morinaga H, Gugelot PC. Gamma rays following (α,xn) reactions.
LEAP Low-energy all-purpose Nucl Physics. 1963;46:210–24.
LEGP Low-energy general-purpose 14. Anger HO. A multiple scintillation counter in vivo scanner. Am J
LEHR Low-energy high-resolution Roentgenol Ther Nucl Med. 1953;70:605–12.
LYSO(Ce) Cerium-doped lutetium yttrium orthosilicate 15. Casey ME, Nutt R. A multicrystal two dimensional BGO detector
system for positron emission tomography. IEEE Trans Nucl Sci.
MEAP Medium-energy all-purpose 1986;33:460–3.
MEGP Medium-energy general-purpose 16. Del Guerra A, Belcari N, Bisogni M. Positron emission tomogra-
NaI(Tl) Thallium-activated sodium iodide phy: its 65 years. Rivista del Nuovo Cimento. 2016;39-4:155–223.
PET Positron emission tomography 17. Blazek KJ, Vrzala VP. Composite detector of bismuth germanate
mono-crystals. Czechoslovakian Patent no. 256143. Registered
PMT Photomultiplier tube 15-04-1986.
PSM Prone scintimammography 18. Pani R, Pellegrini R, Soluri A, De Vincentis G, Scafè R, Pergola
PSPMT Position-sensitive PMT A. Single photon emission imaging by position sensitive PMT. Nucl
RIGS Radioimmunoguided surgery Instr Meth A. 1998;409:524–8.
19. Shao Y, Silverman RW, Cherry SR. Evaluation of Hamamatsu

SM Scintimammography R5900 series PMTs for readout of high-resolution scintillator
SN Sentinel node arrays. Nucl Instr Meth A. 2000;454:379–88.
SNR Signal-to-noise ratio 20. Pani R, Soluri A, Scafè R, Pergola A, Pellegrini R, De Vincentis
SPECT Single-photon emission computed G, et al. Multi-PSPMT scintillating camera. IEEE Trans Nucl Sci.
1999;46:702–8.
tomography 21. Kyushima H, Shimoi H, Atsumi A, Ito M, Oba K, Yoshizawa Y. The
SPECT/CT Single-photon emission computed tomogra- development of flat panel PMT. IEEE Nucl Sci Symp Conf Rec.
phy/computed tomography 2000;1:7/3–7. https://doi.org/10.1109/NSSMIC.2000.949246.
SPEM Single-photon emission mammography 22. Scafè R, Pellegrini R, Cinti MN, Puccini M, Pani R. A novel scintil-
lation imager with charge-spread discrimination.Analytical models
YAlO3(Ce) Cerium-activated yttrium aluminum perovskite suitable for crystal-arrays. Nucl Instr Meth A. 2016;833:110–21.
YAP(Ce) Cerium-activated yttrium aluminum perovskite 23. Renker D. Geiger-mode avalanche photodiodes, history, proper-
ties and problems. Nucl Instr Meth A. 2006;567:48–56. https://doi.
org/10.1016/j.nima.2006.05.060.
24. Khalkali I, Mena I, Jouanna E, Diggles L, Venegas R, Block J, et al.
Prone scintimammography in patients with suspicion of carcinoma
References of the breast. J Am Coll Surg. 1994;178:491–7.
25. Scopinaro F, Schillaci O, Scarpini M, Mingazzini L, Di Macio L,
1. Anger HO. Scintillation camera. Rev Sci Instrum. 1958;29:27–33. Banci M, et al. Technetium-99m sestamibi: an indicator of breast
2. Pani R. Nuclear medicine: special applications in functional imag- cancer invasiveness. Eur J Nucl Med. 1994;21:984–7.
ing. In: Del Guerra A, editor. Ionizing radiation detectors for 26. Garibaldi F, Cisbani E, Cusanno F, Iommi R, Urciuoli GM, Pani R,
medical imaging. Singapore: World Scientific Pub. Co. Inc; 2004. et al. Optimization of compact gamma cameras for breast imaging.
p. 359–83.. ISBN: 978-9812386748. Nucl Instr Meth A. 2003;471:222–8.
3. Zanzonico P, Heller S. Physics, instrumentation, and radiation pro- 27. De Vincentis G, Scopinaro F, Pani R, Pellegrini R, Soluri A, Ierardi
tection. In: Biersack H-J, Freeman LM, editors. Clinical nuclear M, Ballesio L, Weinberg IN, Pergola A. 99mTc MIBI scintimam-
medicine. Heidelberg: Springer; 2007. p. 1–33. mography with a high resolution single tube gamma camera: pre-
4. Zanzonico P. Principles of nuclear medicine imaging: planar, liminary study. Anticancer Res. 1997;17(3B):1627–30.
SPECT, PET, multi-modality, and autoradiography systems. Radiat 28. De Vincentis G, Gianni W, Pani R, Cacciafesta M, Pellegrini R,
Res. 2012;177:349–64. Soluri A, et al. Role of 99mTc-Sestamibi scintimammography by
5. Zanzonico P. Radionuclide imaging. In: Cherry S, Badawy R, Qi J, SPEM camera in the management of breast cancer in the elderly.
editors. Essentials of in vivo biomedical imaging. Boca Raton, FL: Breast Cancer Res Treat. 1998;48:159–63.
CRC Press; 2015. p. 1765–224. 29. Pani R, Pellegrini R, De Vincentis G, Cinti MN, Weinberg IN,
6. Slomka PJ, Berman DS, Germano G. New cardiac cam- Soluri A, et al. Factors affecting cancer detectability in Tc-99m
eras: single-photon emission CT and PET. Semin Nucl Med. MIBI scintimammography. Nucl Instr Meth A. 2003;497:90–7.
2014;44:232–51.
170 R. Pani et al.
30. Gambaccini M, Fantini A, Bollini D, Castelli E, Del Guerra A, 54. Sanchez F, Benlloch JM, Escat B, et al. Design and tests of a por-
Di Domenico G, et al. Development of a quasi-monochromatic table mini gamma camera. Med Phys. 2004;31:1384–97.
CT system for breast cancer study with combined emission- 55. Sanchez F Fernandez MM, Gimenez M, et al. Performance tests of
transmission tomography. IEEE Trans Nucl Sci. 2001;48:703–6. two portable mini gamma cameras for medical applications. Med
31. Travin MI. Cardiac cameras. Semin Nucl Med. 2011;41:182–201. Phys. 2006;33:4210–20.
32. Sweet WH. The uses of nuclear disintegration in the diagnosis and 56. Ortega J, Ferrer-Rebolleda J, Cassinello N, et al. Potential role of
treatment of brain tumors. N Engl J Med. 1951;245:875–8. a new hand-held miniature gamma camera in performing mini-
33. Harris CC, Bigelow RR, Francis JE, Kelley GG, Bell PR. A CsI(Tl)- mally invasive parathyroidectomy. Eur J Nucl Med Mol Imaging.
crystal surgical scintillation probe. Nucleonics. 1956;14:102–8. 2007;34:165–9.
34. Hoffman EJ, Tornai MP, Janecek M, Patt BE, Iwanczyk JS. lntraopera- 57. Mathelin C, Salvador S, Bekaert V, et al. A new intraoperative
tive probes and imaging probes. Eur J Nucl Med. 1999;26:1913–35. gamma camera for the sentinel lymph node procedure in breast
35. Knoll GF, Lieberman LM, Nishiyama H, Beierwaltes WH. A
cancer. Anticancer Res. 2008;28:2859–64.
gamma ray probe for the detection of ocular melanomas. IEEE 58. Mathelin C, Salvador S, Huss D, et al. Precise localization of sen-
Trans Nucl Sci. 1972;19:76–80. tinel lymph nodes and estimation of their depth using a prototype
36. Pircher FJ, Anderson B, Cavanaugh PJ, Sharp KW. Experiences intraoperative mini gamma-camera in patients with breast cancer. J
with a solid state detector for surface counting of phosphorus-32. J Nucl Med. 2007;48:623–9.
Nucl Med. 1967;8:444–50. 59. Daghighian F, Fong Y. Detectors for intraoperative molecular imag-
37. Lauber A. Development of miniaturized solid state detectors for the ing: from probes to scanners. In: Fong Y, Giulianotti PC, Lewis J,
measurement of beta and gamma radiation in superficial and deep Koerkamp BG, Reiner T, editors. Imaging and visualization in the
parts of living tissue. Nucl Instr Meth. 1972;101:545–50. modern operating room: a comprehensive guide for physicians.
38. Soluri A, Scafè R, Scopinaro F, Bruschi R, Casasanta V, Latini New York, NY: Springer; 2015. p. 55–67.
R, et al. The imaging probe: miniaturized gamma camera. In:
Limouris GS, Biersack HJ, Kouros KC, editors. Radionuclides for
lymph node mapping – current status and future aspects. Athens:
Mediterra Publishers; 1999. p. 67–72.. ISBN 960-86437-0-8. Further Reading
39. Pani R, Soluri A, Scafè R, Pellegrini R, Tatì A, Scopinaro F, et al.
A compact gamma ray imager for oncology. Nucl Instr Meth A.
2002;477:509–13.
40. Pani R, de Notaristefani F, Blazek K, Maly P, Pellegrini R, Pergola Books
A, et al. Multi-crystal YAP:Ce detector system for position sensitive
measurements. Nucl Instr Meth A. 1994;348:551–8. Cherry SR, Sorenson JA, Phelps ME. Physics in nuclear medi-
41. Truman A, Bird AJ, Ramsden D, He Z. eCsI(Tl) arrays with position- cine. 4th ed. Philadelphia, PA: Elsevier Saunders; 2012. ISBN:
sensitive PMT readout. Nucl Instr Meth A. 1994;353:375–8. 978-1-4160-5198-5.
42. Woolfenden JM, Barber HB. Intraoperative probes. In: Wagner DeWerd L, Kissick M, editors. The phantoms of medical and health
Jr HN, Szabo Z, editors. Principles of nuclear medicine. 2nd ed. physics. Devices for research and development. New York, NY:
Philadelphia, PA: WB Saunders; 1995. p. 292–7. Springer; 2014. ISBN: 978-1-4614-8303-8.
43. Heller S, Zanzonico P. Nuclear probes and intraoperative gamma Flower MA, editor. Webb’s physics of medical imaging. 2nd ed. Boca
cameras. Semin Nucl Med. 2011;41:166–81. Raton, FL: CRC Press, Taylor & Francis Group; 2012. ISBN:
44. Essner R, Daghighian F, Giuliano AE. Advances in FDG PET 978-1-4665-6895-2.
probes in surgical oncology. Cancer J. 2002;8:100–8. Giussani A, Hoeschen C, editors. Imaging in nuclear medicine.
45. Strong VE, Galanis CJ, Riedl CC, Longo VA, Daghighian F, Humm New York, NY: Springer; 2013. ISBN: 978-3-642-31414-8.
JL, et al. Portable PET probes are a novel tool for intraoperative Godfrey DJ, Das SK, Wolbarst AB, editors. Advances in medical phys-
localization of tumor deposits. Ann Surg Innov Res. 2009;3:2. ics. 5th ed. Madison, WI: Medical physics publishing; 2014. ISBN:
46. Schneebaum S, Essner R, Even-Sapir E. Positron-sensitive probes. 978-1-930524-63-7.
In: Mariani G, Giuliano AE, Strauss HW, editors. Radioguided sur- IAEA. Human health series no. 29. Guided intraoperative scinti-
gery – a comprehensive team approach. New York, NY: Springer; graphic tumour targeting (GOSTT); implementing advanced hybrid
2008. p. 23–8. molecular imaging and non-imaging probes for advanced cancer
47. Raylman RR. Performance of a dual, solid-state intraoperative probe management. Wien: International Atomic Energy Agency; 2014.
system with 18F, 99mTc, and 111In. J Nucl Med. 2001;42:352–60. ISBN:978-92-0-102214-1.
48. Daghighian F, Mazziotta JC, Hoffman EJ, et al. Intraoperative beta Iwanczyk JS, editor. Radiation detectors for medical imaging. Boca
probe: a device for detecting tissue labeled with positron or electron Raton, FL; London; New York, NY: CRC Press, Taylor & Francis
emitting isotopes during surgery. Med Phys. 1994;21:153–7. Group; 2015. ISBN: 978-1-4987-0435-9.
49. Britten AJ. A method to evaluate intra-operative gamma probes for Khalkhali I. In: Maublant JC, Goldsmith SJ, editors. Nuclear oncol-
sentinel lymph node localisation. Eur J Nucl Med. 1999;26:76–83. ogy. Diagnosis and therapy. Philadelphia, PA: Lippincott Williams
50. Hoffman EJ, Tornai MP, Levin CS. Gamma and beta intra-operative & Wilkins; 2000. ISBN 0-7817-1990-9.
imaging probes. Nucl Instr Methods Phys Res. 1997;392:324–32. Knoll GF. Radiation detection and measurement. 3rd ed. New York, NY;
51. Scopinaro F, Soluri A. Gamma ray imaging probes for radiogu- Chichester; Weinheim; Brisbane, QLD; Toronto, ON; Singapore:
ided surgery and site-directed biopsy. In: Mariani G, Giuliano John Wiley & Sons, Inc.; 1999. ISBN 0-471-07338-5.
AE, Strauss HW, editors. Radioguided surgery. New York, NY: Mariani G, Giuliano AE, Strauss HW, editors. Radioguided surgery -
Springer; 2008. p. 29–36. a comprehensive team approach. New York, NY: Springer; 2008.
52. Pitre S, Menard L, Ricard M, et al. A hand-held imaging probe 978-0-387-33684-8.
for radio-guided surgery: physical performance and prelimi- Martin C, Sutton D, editors. Practical radiation protection in health-
nary clinical experience. Eur J Nucl Med Mol Imaging. 2003;30: care. 2nd ed. Oxford: Oxford University Press; 2015. ISBN:
339–43. 9780199655212.
53. Tsuchimochi M, Hayama K, Oda T, et al. Evaluation of the efficacy Perkins A, Lees JE. Gamma cameras for interventional and intraopera-
of a small CdTe gamma-camera for sentinel lymph node biopsy. J tive imaging. Boca Raton, FL; London; New York, NY: CRC Press,
Nucl Med. 2008;49:956–62. Taylor & Francis Group; 2016. ISBN 9781498729284.
Phelps ME, editor. PET: physics, instrumentation and scanners. Povoski SP, Neff RL, Mojzisik CM, O’Malley DM, Hinkle GH, Hall
New York, NY: Springer; 2006. ISBN: 978-0-387-32302-2. NC, et al. A comprehensive overview of radioguided surgery using
Zaidi H, editor. Molecular imaging of small animals. Instrumentation gamma detection probe technology. World J Surg Oncol. 2009;7:11.
and applications. New York, NY: Springer; 2014. ISBN: https://doi.org/10.1186/1477-7819-7-11.
978-1-4939-0893-6. Tsuchimochi M, Hayama K. Intraoperative gamma cameras for
radioguided surgery: technical characteristics, performance
parameters, and clinical applications – review paper. Phys Med.
2013;29:126–38.
Van Audenhaege K, Van Holen R, Vandenberghe S, Vanhove C, Metzler
Review Articles/Chapters SD, Moore SC. Review of SPECT collimator selection, optimiza-
tion, and fabrication for clinical and preclinical imaging. Med Phys.
Bricou A, Duval M-A, Charon Y, Barranger E. Mobile gamma cam- 2015;42:4796–813.
eras in breast cancer care – a review. Eur J Surg Oncol. 2013;39: Zanzonico P. Instrumentation for positron emission imaging. In: Strauss
409–16. HW, Mariani G, Volterrani D, Larson SM, editors. Nuclear oncol-
Fontein DBY, van de Water W, Sven J, Mieog D, Liefers G-J, van de ogy – from pathophysiology to clinical applications. New York, NY:
Velde CJH. Timing of the sentinel lymph node biopsy in breast can- Springer; 2017. p. 217–50.
cer patients receiving neoadjuvant therapy – recommendations for Zanzonico P. Instrumentation for single-photon emission imaging.
clinical guidance. Eur J Surg Oncol. 2013;39:417–24. In: Strauss HW, Mariani G, Volterrani D, Larson SM, editors.
Gullberg GT, Reutter BW, Sitek A, Maltz JS, Budinger TF. Topical Nuclear oncology – from pathophysiology to clinical applications.
review – dynamic single photon emission computed tomogra- New York, NY: Springer; 2017. p. 251–74.
phy - basic principles and cardiac applications. Phys Med Biol. Zanzonico P. Instrumentation for intraoperative detection. In: Strauss
2010;55:R111–91. HW, Mariani G, Volterrani D, Larson SM, editors. Nuclear oncol-
Peterson TE, Furenlid LR. Topical review – SPECT detectors: the ogy – from pathophysiology to clinical applications. New York, NY:
Anger Camera and beyond. Phys Med Biol. 2011;56:R145–82. Springer; 2017. p. 275–304.
Image Acquisition and Processing
with Gamma Cameras Including 7
Integrated SPECT/CT and Dedicated
Gamma Cameras
Duccio Volterrani, Federica Guidoccio, Giulia Puccini,

and Sara Mazzarri
Contents
7.1 Basic Acquisition Parameters 173
7.1.1 Energy Correction Map 173
7.1.2 Linearity Correction Map (Spatial Correction) 174
7.1.3 Uniformity Correction Map 174
7.1.4 Energy Window 174
7.1.5 Acquisition Matrix 175
7.1.6 Zoom Factor 176
7.2 Specific Acquisition Parameters 176
7.2.1 Static Acquisition Mode 176
7.2.2 Dynamic Acquisition Mode 177
7.2.3 Gated or Multigated Acquisition (MUGA) 177
7.2.4 Total Body Acquisition 178
7.2.5 SPECT Acquisition Mode 179
7.2.6 SPECT Data Reconstruction 180
7.3 S PECT/CT 182
7.3.1 C admium-Zinc-Telluride Systems 182
7.3.2 Q uantitative SPECT 184
References 185
Learning Objectives
• Describe the design and operational differences existing
7.1 Basic Acquisition Parameters
among various gamma camera devices.
Although numerous vendors manufacture commercial single
• Understand the different correction maps and key param-
photon gamma cameras, each with their own features, all share
eters commonly applied in planar or SPECT
components such as a scintillation crystal, photomultipliers,
acquisitions.
computer control for detector motion, and electronics to correct
• Describe how a gamma camera can operate with different
the signals for photon energy, field uniformity and linearity. It is
acquisition modes.
also well known that every gamma camera, considered as a
• Understand the methodologies of SPECT data recon-
whole as crystal-photomultiplier-electronic system, invariably
struction, and describe the main reconstruction algorithms
generates distortions that must be minimized and corrected in
and filters.
order to obtain images with good spatial resolution and contrast.
• Describe the clinical use of SPECT/CT imaging.
Several correction maps are commonly included in the system
• Summarize the main features of cadmium-zinc-telluride
as a default preset configuration and then applied in real-time
(CdZnTe) SPECT scanners.
mode when planar or SPECT images are acquired.
• Describe the main principles of quantitative SPECT.
D. Volterrani (*) · F. Guidoccio · G. Puccini · S. Mazzarri 7.1.1 Energy Correction Map

University of Pisa, Pisa, Italy Given a monoenergetic source, the z-value of the centroid of
e-mail: duccio.volterrani@med.unipi.it the energy peak should ideally be the same in every point

174 D. Volterrani et al.
parison between the measured distances on the acquired image

and the known data of the phantom used enables to evaluate
and store deviations from the expected response within a look-
up table to spatially correct each acquisition. Acquisition of
the spatial correction map is part of the tasks of the technician
responsible for periodic maintenance; also in this case,
depending on different manufacturers, a single spatial map
valid for all the isotopes can be acquired, or multiple maps for
the different isotopes can be used, since different energies pro-
duce different patterns of spatial nonlinearity.
7.1.3 Uniformity Correction Map
This parameter corrects local variations in counting efficiency

over the whole FOV. This correction map can be intrinsic
Fig. 7.1 Screen reporting some preset parameters and correction maps
(when performed without a collimator) or refer to the entire
to be used for a scintigraphic acquisition detection system (when performed with the collimator and tak-
ing into account for nonuniformities due also to the collimator
itself); each type of correction can be obtained for each isotope
(x,y) of the whole field of view (FOV). However, since every and for each collimator. It consists in the acquisition of a uni-
photomultiplier interacts with the nearest crystal portion pro- formity map (obtained by uniform irradiation of the detector);
viding an energy spectrum for the detected radiation, possi- by determining the average value of counts within the acquisi-
ble defects in the crystal leading to small variations in the tion matrix and the corresponding corrective factor for each
production of luminous photons or variations in the effi- pixel, a corrective matrix (correction map) can be obtained, to
ciency of light collection and in photomultiplier gain can be subsequently applied in the acquisition phase.
produce differences of the z-value from point to point in the
detector. This problem is solved by using an energy correc-
tion map, periodically acquired by the technical personnel 7.1.4 Energy Window
responsible for maintenance and quality control of the
gamma camera. This map, stored as a look-up table, allows The purpose of any scintigraphic acquisition is to obtain an
corrections of the differences in the detector’s energy image with the highest possible signal-to-noise ratio, mini-
response during the acquisition. Depending on the design of mizing scattered radiation. This is achieved by setting up an
the gamma camera, either a single correction map valid for “acceptance” window (energy window) of the detected pho-
all isotopes or two different maps for low and high energies, tons for each isotope imaged by the system; only the accepted
respectively, or a specific correction map for each isotope is photons will compose the image. The window is generally
preset and loaded for scintigraphic acquisition (Fig. 7.1). identified with the value in keV of the energy peak and with
a window width expressed in percentage value. For example,
for 99mTc, the window is 140 keV, and the amplitude is gener-
7.1.2 L
inearity Correction Map ally 20%. In practice, all those photons that have an energy
(Spatial Correction) between 140 keV ± 10% (±14 keV) will be accepted, i.e.,
those with energy values between 126 and 154 keV. For
Response of the photomultipliers is not isotropic with respect some particular applications, window amplitudes of 15% or
to the light produced in the crystal (i.e., there is a variation in even 20% are recommended. However, by decreasing the
the radial direction starting from the center toward the edges). window amplitude, FOV nonuniformities tend to increase as
Moreover, since photomultipliers do not cover the entire sur- well.
face of the crystal, a scintillation event projected in the space In addition to the values of energy peak and window
between two photomultipliers produces a different signal than width, there is a third parameter concerning the positioning
that generated by an event projected in the center of a photo- of the energy window, called “offset.” An offset of 0 means
multiplier. For this reason, the image can show spatial non- that the center of the window coincides with that of the ener-
uniformities in count distribution. In order to correct these getic peak, while an offset of 10 indicates that the window is
spatial distortions, a correction map is periodically acquired centered on the value of the peak +10%. For example, setting
using a specific phantom (parallel bars or multi-hole); a com- the parameters 140 keV for peak energy and 20% for win-
7 Image Acquisition and Processing with Gamma Cameras Including Integrated SPECT/CT and Dedicated Gamma Cameras 175
dow width, an offset of ten identifies a window centered on

140 + 14 = 154 keV, therefore with actual amplitude between
140 and 168 keV. The offset can be used when the shape of
the energy peak of the radionuclide used is asymmetrical.
Gamma cameras have a set of predefined energy window
values for the different types of γ-emitting isotopes used for
diagnostic imaging.
Since some radionuclides used for diagnostic investiga-
tions emit photons with two or more different energies (e.g.,
201
Tl, 111In, 67Ga, etc.), it is possible to perform a so-called
“multi-peak” acquisition, i.e., an acquisition involving the
use of more than one energy window; the image produced is
given by the sum of the photons acquired in the various
energy windows. For example, 111In emits photons with two
energetic peaks (172 keV and 245 keV, respectively); in this
case, the acquisition is performed by centering two windows
on the corresponding energies: 172 keV ± 10% and
245 keV ± 10%.
Also during multi-isotope acquisition mode, the energy
windows are positioned on different energy peaks; however,
the photons accepted within each individual energy window
will compose separate images (different datasets or matri-
ces). This mode is employed when the biodistribution of two
(or more) different types of radionuclides must be imaged
(Fig. 7.2). For example, in a patient with follicular differenti-
ated thyroid carcinoma, it may be necessary to perform a
scintigraphy with 131I-iodide (to image all iodide-avid metas-
tases throughout the body) combined with 99mTc-MDP (to
identify which of those metastases are located in the skele-
ton); in this case, the parameters set for the two simultaneous
acquisitions (i.e., to constitute two parallel sets of data) are:
Fig. 7.2 Energy windows of multi-peak and multi-isotope
• Set 1 (131I): 364 keV ± 10% acquisitions
• Set 2 (99mTc): 140 keV ± 10%
One of the most frequently used scatter correction meth- 7.1.5 Acquisition Matrix
ods is usually based on multiple energy windows set at dif-
ferent positions across the energy spectrum. In this method, Acquisition matrices generally used in nuclear medicine are
called “dual energy method,” data are simultaneously 64 × 64, 128 × 128, or 256 × 256 for planar acquisitions,
acquired in two different energy windows, a specific photo- while they are 64 × 64 or 128 × 128 for SPECT acquisitions.
peak window (126–154 keV for 99mTc) and a lower-energy However, it is possible to acquire with matrices of different
scatter window (92–125 keV for 99mTc), acquiring two sepa- sizes (32 × 32, 512 × 512, 1024 × 1024) and also with shapes
rate images (emission image and scatter image). The different from the square one (e.g., a rectangular matrix
assumption of this approach is that the events detected in the 128 × 512, 256 × 1024, or 512 × 2048 can be used for total
scatter window are correlated to the scatter component of body studies).
the events detected in the photo-peak window by a constant Why is 128 × 128 the most frequently used acquisition
factor. This approach is usually performed in SPECT (espe- matrix for planar gamma camera imaging? The answer lies
cially SPECT/CT) and is mainly applied on the projection in the gamma camera spatial resolution and in the sampling
images before image reconstruction. Preset energy and scat- theorem (Nyquist-Shannon), which states that for a finite and
ter windows to perform this dual energy method scatter cor- known bandwidth signal, the minimum sampling frequency
rection are available also for isotopes other than 99mTc (123I, of this signal must be at least twice its maximum frequency.
I, In, etc.).
131 111
In a scintigraphic image, the signal is the response of the
gamma camera, and its maximum frequency is the limit of

spatial resolution, while sampling is given by the pixels that Key Learning Point
compose the digital matrix of the image. Assuming 8 mm as • Gamma cameras invariably generate distortions
the spatial resolution limit of a gamma camera for low ener- that must be corrected in order to obtain images
gies and with a LEHR collimator, it will be necessary for with optimal spatial resolution and contrast.
each pixel to measure around 4 mm for the sampling theorem • Preset correction maps, such as the energy correc-
to be respected. With a gamma camera having field of view tion, linearity correction, and uniformity correction
of 58 cm, using a 128 × 128 matrix results in pixels with a maps, are applied in real-time mode when planar or
size of about 4.5 mm (58/128 = 0.453 cm), which is suffi- SPECT images are acquired.
cient to satisfy the sampling theorem and to do not loose • Gamma cameras have a set of predefined energy
information. window values for the different isotopes.
The practical application of these concepts justifies the • In multi-peak and multi-isotope acquisition modes,
general principle that the limited resolution power of the the energy windows are positioned on different
gamma camera does not justify the use of the very fine matri- energy peaks.
ces (512 × 512) used in radiodiagnostics, where spatial reso- • Scatter correction can be performed acquiring data
lution is much higher. Therefore, using very fine matrices for with a dual energy method.
scintigraphic images not only is useless but can even be • The choice of an acquisition matrix and zoom fac-
counterproductive; in fact, keeping the acquisition times tor should satisfy the sampling theorem in order to
constant, as the number of pixels that compose the image not lose information.
increases, the number of counts/pixels is reduced, thus caus- • The use of preset protocols is useful to correctly
ing an increase in background and, therefore, a reduction in choose energy settings, matrix of acquisitions, and
the contrast of the image. zoom factors.
7.1.6 Zoom Factor 7.2 Specific Acquisition Parameters
This parameter defines the dimensions of the field of interest A gamma camera can operate with different acquisition
which is sampled with the preset acquisition matrix. A zoom modes, each one requiring specific key parameters.
factor of 1 means that no magnification factor is applied,
since the preset field of interest has the dimensions of the
maximum FOV. The acquisition size can be obtained by 7.2.1 Static Acquisition Mode
dividing the maximum FOV for the zoom factor; since the
FOV varies with the zoom factor, they vary in the pixel For static planar (2D) acquisition, the detector is stationary
dimensions. For example, in the case of a gamma camera with respect to the patient’s body, and the accepted photons
with FOV of 58 cm and setting an acquisition matrix of are recorded producing a single two-dimensional projection
128 × 128, the following pixel dimensions are obtained by image. It is often necessary to acquire multiple views (front,
changing the zoom factor: rear, right and left side, front and back oblique, right and
left), through which it is possible to offset the different planes
• Zoom 1, acquisition FOV: 58/1 = 58 cm → pixel size to which the anatomical structures being explored belong.
4.5 mm Typical examples of static planar acquisitions are thyroid
• Zoom 1.33, acquisition FOV: 58/1.33 = 43.6 cm → pixel scintigraphy, pulmonary (perfusion and ventilation), spot
size 3.4 mm views during bone scintigraphy, etc.
• Zoom 1.5, acquisition FOV: 58/1.5 = 30.6 cm → pixel The key parameters that define a static planar acquisition are:
size 3.0 mm
• Zoom 2, acquisition FOV: 58/2 = 29 cm → pixel size • Start: it can be manual, based on time (defining how many
2.25 mm minutes or seconds after tracer administration the acquisi-
tion should start), or defined on the basis of reaching a
The zoom factor setting is associated with the key called specific counting rate (collecting events when the count
“pan” that specifies the position where the zoomed FOV is rate is sufficient, e.g., 1 kcps).
centered. Usually the offsets of the spatial coordinates X and • Stop: it can be manual, predefined on reaching a certain time
Y are also required to be set. (e.g., 5 min) and/or a total image count (e.g., 500 kcounts).
evaluation, for example, 60 s in total (with 1/40 s per frame)

for a first-pass cardiac transit study but 60 min in total (with
1 min per frame) for a hepatobiliary study. Using this type of
acquisition, it is possible to perform processing procedures
that provide activity/time curves, from which it is possible to
derive various semiquantitative/quantitative parameters; a
typical example of dynamic acquisition is represented by
sequential renal scintigraphy.
The parameters that define a planar dynamic static are:
• Start: it can be manual, based on time (defined after how

many minutes or seconds must the acquisition starts), or
defined on the basis of reaching a specific counting rate
(e.g., 1 kcps).
• Stop: it is predefined based on the number and duration of
Fig. 7.3 Screen reporting information on patient position on the imag- the frames; several phases are possible, each character-
ing table ized by a certain number of frames with a specific dura-
tion (e.g., first phase of 60 s, 60 frames each with a
duration of 1 s; second phase of 14 min, 84 frames each
• The count-based stop can also be predefined within a cer- with a duration of 10 s).
tain area of interest (ROI). • Mirror (see above).
• Mirror: if selected, the image is mirrored on the Y- or • Rotation (see above).
X-axis; to be selected in case of acquisitions with pinhole
collimator, where the image is upside down and specu-
lated with respect to the source object. 7.2.3 Gated or Multigated Acquisition (MUGA)
• Rotation: it specifies the degrees of rotation (0°, 90°,
180°, 270°) to be applied to the image. This is mainly In a MUGA the detector is stationary with respect to the
used when the patient is not positioned in a standard way patient’s body. This mode is used typically to generate a
(head facing the gantry); a 180° rotation is selected when sequence of images (frames) that represent an average car-
acquiring with a pinhole collimator where the image is diac cycle. A three-lead EKG trace is used to synchronize
inverted and speculated with respect to the source object. the acquisition with the cardiac cycle. In particular, the R
wave of the EKG (which is the trigger for ventricular sys-
In recent gamma cameras, the rotation function has been tole) is used to synchronize the acquisition. In a MUGA
complemented with the possibility to set position of the study, data are recorded from multiple (300–500 depending
patient on the table (supine or prone) and orientation with on heart rate) cardiac cycles, generating a single “average”
respect to the gantry (head first or feet first) (Fig. 7.3), in cardiac cycle—usually composed of at least 16 frames. As
order to correctly display the images (head on top, feet on the R wave is detected, the acquisition of each cycle begins
bottom, left side on the right for anterior view, left side on the by acquiring frames 1, 2, 3, … n, until the next R wave
left for posterior view). occurs, at which time the acquisition is redirected to frames
1, 2, 3, … n (gating), and so on (Fig. 7.4). MUGA continues
until the desired counts or time has been reached. The most
7.2.2 Dynamic Acquisition Mode commonly used acquisition mode is defined as “frame-
mode” thereby the number of frames in which the cardiac
Similarly as during spot static acquisition, the detector is sta- cycle is divided (16, 24, or 32) and the duration is estab-
tionary with respect to the patient’s body, and all the photons lished a priori by calculating the average duration of a car-
that reach the plane of detection are collected. However, this diac cycle based on the acquisition of a certain number of
acquisition mode enables to display how distribution of a beats. Since changes in heart rate can occur during acquisi-
radiopharmaceutical within the stationary FOV varies as a tion, the classic frame-mode results in loss of counts and in
function of time; this effect is obtained by acquiring a sequence a poor definition of the diastolic phase, which is more sub-
of planar images, each one of which is called a frame. ject to variations in relation to heart rate. While in the past
The duration of the whole acquisition and of each frame the only alternative was represented by the acquisition mode
varies according to the biological process/function under called “list-mode,” currently there are variations in the
11 12 13 14 15 16 1 2 3 5 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 5 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10
Fig. 7.4 Diagrammatic representation scheme of EKG-gated acquisition
frame-mode mode that make it much more versatile and at 7.2.4 Total Body Acquisition
the same time simple and quick application. While in the
classic frame-mode the 10 beats for calculation of the aver- This acquisition mode is used when a body scan that exceeds
age cardiac duration were acquired only before beginning the axial dimensions of a gamma camera FOV is performed
the acquisition, currently the average heart rate (and there- without interruption. In most current gamma cameras, this
fore the duration of the single frame) can be dynamically type of acquisition is possible since the imaging table moves
updated during the acquisition. with a predefined constant speed during acquisition. With
This type of acquisition is used in gated equilibrium double detector gamma cameras, a single scan acquiring
angioscintigraphy, thereby functional data such as the (left) front and back views is possible. Total body acquisitions are
ventricular ejection fraction can be derived. The parameters frequently used for bone scintigraphy, for 131I-iodide scintig-
specifically defining MUGA are: raphy in patients treated for differentiated thyroid cancers,
for somatostatin receptor scintigraphy, for scintigraphy with
• Trigger window of acceptance of the variation of cardiac radiolabelled autologous leukocytes, etc. The use of laser
cycle length (generally expressed as a percentage with systems allows body contouring to perform a scan adapted to
respect to the average value, e.g., ±10%). the external profile of the patient’s body, thus optimizing the
• Start: it is generally manual. detector-patient distance from head to feet.
• Stop: it can be manual, predefined on reaching a certain The parameters that define a total body acquisition are:
time (e.g., 5 min), and/or counts and/or a predetermined
number of acquired cardiac cycles. • Start: generally manual.
• Mirror (see above). • Stop: predefined based on the length of the scan.
• Rotation (see above). • Scanning speed: expressed in cm/min or in min/meter.
Thus, raw data are acquired as a series of discrete planar

images at multiple angles about the longitudinal axis of the
patient. The number of counts recorded in each projection
image pixel represents the integral of sampling from the
source object through the detector.
Besides the use of the same correction maps commonly
employed in all the acquisition modes described above, for
SPECT a dedicated map correcting for center of rotation
(COR) errors is applied. During SPECT imaging, COR
errors, related to an alignment error between the mechanical
center of the rotational gantry and the center of the electronic
matrix, lead to misalignment of the projection data and can
potentially degrade quality of the reconstructed images. In
modern SPECT systems, the acceptable limit for COR offset
is ±1 mm, and COR should be checked frequently. Most
manufacturers include software designed to determine if the
COR is within acceptable limits. Moreover, COR offset cor-
rection maps are periodically acquired for each detector and
collimator, to be applied for SPECT acquisitions.
During SPECT acquisition, a sampling of the body region
under examination is performed, acquiring a predefined
number of views separated by a certain rotation angle or
step: from 60/64 to 120/128 projection images from 6° to 3°
angular increments, respectively; an angular increment
Fig. 7.5 Screen of a preset whole-body acquisition with a dual- greater than 6° between successive projection images will
detector gamma camera using five partially overlapping FOVs result in undersampling artifacts in the reconstructed images.
A 180° or 360° rotation arc is employed for cardiac or non-
• Limits of the scan: the imaging table generally has a mea- cardiac studies, respectively.
suring guide in cm, from which it is possible to set the The scanning orbit can be circular or elliptical; the avail-
starting point and the end of the scan. ability of gamma cameras equipped with a laser system that
• Mirror (see above). defines body contouring enables recording a scan adapted to
• Rotation (see above). the external profile of the patient’s body, optimizing the
detector-patient distance in all the projection angles.
In modern gamma cameras, a total body acquisition can Because of the relatively long time required to acquire a
also be obtained from a predefined and automatic sequence scan, dynamic SPECT remains largely impractical with most
of acquisitions of several FOVs along the patient body (seg- commercially available conventional SPECT systems. On
ments or beds), partially overlapping one onto each other, the other hand, whole-body SPECT (and SPECT/CT) is fea-
from which dedicated software rebuilds the whole-body sible with state-of-the-art gamma cameras that allow sequen-
image (Fig. 7.5). tial acquisition of multiple body segments and perform data
reconstruction to yield final whole-body SPECT or SPECT/
CT dataset.
7.2.5 SPECT Acquisition Mode Although PET offers important advantages over SPECT
(i.e., generally better spatial resolution, higher sensitivity,
Although there are many possible combinations of detector and more accurate activity quantitation), SPECT offers the
number, geometry, and motion that can acquire the necessary capability of simultaneous multi-isotope imaging using dis-
projection data, two-detector rotating gamma camera-based tinct, isotope-specific peak energy windows.
SPECT systems are the most common configuration [1]. The It is also possible to perform gated SPECT acquisitions,
basic SPECT imaging paradigm includes (1) acquisition of e.g., a combination of SPECT and MUGA, thereby the
planar images (views—or projections in analogy with radio- acquisition for each view of a series of frames (bins) span-
logic imaging) from multiple angles around the subject, (2) ning the length of the cardiac cycle at equal intervals during
correction of the acquired data for nonuniform scanner an EKG-gated acquisition is recorded. Image data for each
responses and other signal-degrading effects, and (3) mathe- frame are acquired repeatedly over many cardiac cycles and
matical reconstruction of transverse tissue section images [2]. stored. During processing, all data of a particular frame are
added together in order to reconstruct a specific phase of the A higher fidelity approach uses iterative algorithms.
cardiac cycle. If all temporal frames are summed together, Iterative reconstruction tends to be less noisy, has fewer
reconstruction is equivalent to a standard set of ungated streak artifacts, and often allows incorporation of physical
images. factors associated with the data acquisition into the recon-
Specific parameters defining a SPECT and gated SPECT struction process, thus yielding more accurate results. It may
acquisition are: include assumptions regarding the data acquisition process,
such as estimates of the device’s spatial resolution, that vary
• Start: usually manual with position within the FOV; for example, the variation of
• Stop: based on a predefined number of views and duration collimator spatial resolution as a function of the distance
(usually 20–30 min for data acquisition) between the object and the collimator can be incorporated
• Orbit: circular, elliptical, or programmed with a body into the reconstruction process. In addition, a model of the
contouring system imaging process may incorporate assumptions regarding
• Scanning arc: defines whether on 360° or 180° (in multi- photon attenuation and scatter.
detector gamma cameras, the scanning arc is specified for In a frequently used iterative reconstruction, the pro-
each detector) cess generates an estimate of the object that has the highest
• Scan direction: specifies whether clockwise (CW) or anti- statistical likelihood to have led to the set of acquired pro-
clockwise (ACW) jection data. A commonly used approach for the recon-
• Number of views: defines the total number of views struction of SPECT (and PET) data is the maximum
acquired likelihood expectation maximization (MLEM) algorithm
• Duration of views: defines the duration of acquisition for [3]. The MLEM approach generates a more accurate
each view reconstruction of the data than that obtained through
• Starting angle: specifies the starting angle of the scan FBP. However, the time required for such processing of
• Type of scan: usually step-and-shoot, but continuous scan data is much longer than reconstruction based on the FBP
can also be selected with some gamma cameras approach.
• Number of bins (gated SPECT): defines the number of One approach for reducing the number of iterations is to
bins (generally from 8 to 16) in which each view image is organize the projection data into a series of ordered subsets
subdivided of evenly spaced projections and update the current esti-
• Trigger window (gated SPECT): specifies the width (%) mate of the object after each subset rather than after the
of the acceptance window of the beats complete set of projections. The most common approach
that uses ordered subsets is referred to as OSEM (ordered
subsets expectation maximization), which allows iterative
7.2.6 SPECT Data Reconstruction reconstruction in an acceptable time for clinical work. In
OSEM, using ten subsets and performing two iterations,
At the end of the acquisition process, data stored in digital data can be reconstructed about ten times faster while gen-
matrices of each image from each detector can be reviewed erating a result of similar image quality, as would be
as a “rotatogram” prior to reconstruction of axial images. obtained with 20 iterations using the complete set. In gen-
The rotatogram allows identification of patient motion. An eral, spatial resolution improves by increasing the number
alternative approach to detect motion is the “sinogram”, a 2D of iterations, despite an increase of high-frequency noise
image in which the horizontal axis represents count location (Fig. 7.6) [2, 5, 6].
on the detector and the vertical axis corresponds the angular When iterative reconstruction is employed, the blurring
position of the detector. The corresponding number of counts effect due to the collimator may be included in the recon-
is assigned to each point of the sinogram. struction process, thus resulting in improved spatial resolu-
There are several methods of reconstructing the projection. In fact, collimator-detector blur is the main factor
tion data into axial images. One early approach, filtered back affecting resolution and noise. Images obtained with com-
projection (FBP), is simple and fast and can be used to recon- pensation of the collimator-detector response exhibit not
struct SPECT as well as PET and CT images. However, the only improved resolution and signal-to-noise ratio but also
FBP algorithm generally produces noisy images. The noise lower noise variability in the reconstructed images. The use
control strategy in a FBP algorithm is usually based on of iterative reconstruction algorithms that compensate for
selecting a low-pass filter’s cutoff frequency, i.e., a filter that collimator-detector response allows imaging with shorter
passes signals with a frequency lower than a certain cutoff acquisition times or with a reduced activity administration of
frequency and attenuates signals with frequencies higher the radiopharmaceutical, yet preserving or improving image
than the cutoff frequency. quality [3–5].
Fig. 7.6 Three examples of iterative reconstruction (OSEM) after spatial resolution, but significantly noisier than image on the left. Right:
acquiring brain SPECT with 123I-FP-CIT. Left: OSEM reconstruction Butterworth filtering with a lower cutoff (0.35 cycles/cm vs. 0.55 cycles/
with two iterations and ten subsets. Middle: OSEM reconstruction with cm) that reduces noise but causes important loss of details
six iterations and ten subsets which yield an image with better intrinsic
Regardless of the type of reconstruction algorithm used, filter, and Shepp-Logan filter. Metz and Wiener are restora-
statistical noise affects SPECT images. A common method to tion filters that enhance the signal with a simultaneous
reduce statistical noise in a SPECT image is the application of reduction of noise without resolution loss. Butterworth filter
low-pass (smoothing) filters, which allow the low frequencies is the most common choice for nuclear medicine imaging
to be retained unaltered and to block the high frequencies. [7]. It is characterized by two parameters: the critical fre-
Low-pass filters are characterized mainly by the “cutoff frequency (which is the point at which the filter starts its roll-off
quency” that defines the frequency above which the noise is to zero) and the order or power (which is the slope of roll-
eliminated. The cutoff frequency is expressed in cycles/pixel off) [7].
or cycles/cm or as a fraction of the Nyquist (Nq) frequency. Selection of the proper filter and determination of filter
The cutoff frequency varies typically from 0.2 to 1.0 times the parameters constitute a major problem in the clinical routine.
Nq frequency. The value of the cutoff frequency determines Literature data do not provide complete information for
how the filter will affect both image noise and resolution. A choosing specific parameters of the imaging filters. In most
high cutoff frequency will improve spatial resolution (there- clinical applications, choice of a filter is done empirically or
fore more detail can be seen), but the image will remain noisy. based on suggestions of the gamma camera manufacturer. It
A low cutoff frequency will increase smoothing but will is advisable to use a limited number of filter types and to
degrade image contrast in the final reconstruction (Fig. 7.6). standardize image-processing approaches, using different
There are several low-pass filters available for SPECT preset reconstruction protocols for better diagnostic accu-
reconstruction: Butterworth filter, Hanning filter, Hamming racy and reproducibility.
the body position must remain unchanged during the SPECT

Key Learning Points (and CT) acquisition, as well as after table translation
• A gamma camera can operate with different acqui- between the SPECT and CT scanning process. However,
sition modes, each one requiring specific key since SPECT image acquisition commonly requires
parameters. 15–60 min, misregistration errors due to respiratory motion
• For tomographic acquisitions, iterative reconstruc- often occur; such misregistration errors can be reduced by
tion provides an alternative to filtered back acquiring the image during quiet tidal respiration, which pro-
projection. vides fairly consistent acquisitions of the SPECT and CT
• OSEM is the most commonly used iterative data. However, since SPECT and CT acquisitions can have
algorithm. misregistrations, it is frequently impossible to simply merge
• A common method to reduce statistical noise in a the data without correcting for gross misregistration errors.
SPECT image is the application of low-pass filters; A method of improving image co-registration relies on image
Butterworth filter is most commonly used in nuclear translation in the image post-processing, to compensate for
medicine. patient motion and mis-calibration between CT and SPECT
• Reconstruction algorithms may influence the final data (Fig. 7.8).
results; therefore, the same algorithm should be
used throughout a certain study.
Key Learning Points
• With SPECT/CT, even low-resolution (and low-
dose) anatomical localization can provide useful
7.3 SPECT/CT information that improves the diagnostic accuracy
of SPECT.
The clinical use of dual-modality imaging for conventional • SPECT/CT procedures should be selected on an
nuclear medicine began in 1999 with the first commercial individual basis and should reflect clinical needs.
SPECT/CT system, the GE Discovery VG Hawkeye, based • Low-dose CT with relatively low radiation doses
on a GE Discovery VG SPECT system (designed and intro- (1–4 mSv) is sufficient when the goal is attenuation
duced originally as the Elscint Varicam), a dual-head correction and anatomic referencing.
variable-
geometry SPECT system. This first-generation
SPECT/CT system was relatively limited in terms of spatial
resolution, scan speed, and signal-to-noise performance of
the CT sub-systems [8]. However, it was well suited in terms 7.3.1 Cadmium-Zinc-Telluride Systems
of cost and performance to addressing the needs of attenua-
tion correction; furthermore, even the low-resolution ana- Conventional gamma cameras consist of a scintillation
tomic localization provided information that improved detector based on NaI(Tl) crystals. The use of the scintilla-
diagnostic accuracy compared to stand-alone SPECT scans tion crystal is based on the principle of luminescence, where
(Fig. 7.7). impact of γ-rays causes a flash of visible light. This light is
In general, rather than being employed as a standard detected by the photomultiplier tube, which amplifies and
imaging protocol, combined SPECT/CT procedures should transforms the flash of light from the γ-photon into an elec-
be selected on an individual basis and should reflect clinical tric signal. The average spatial resolution of collimated
needs. The radiation dose delivered by CT is a major issue in scintillation scanners is about 1–2 cm, and image acquisi-
this regard, because diagnostic CT can increase the overall tion time varies from seconds up to 20 min, depending on
effective radiation dose by up to 14 mSv [9]. administered activity. Cadmium- zinc-telluride (CdZnTe)
Low-dose CT, involving relatively low radiation doses SPECT scanners were introduced in the first decade of the
(1–4 mSv), should be sufficient when the goal of hybrid twenty-first century [11, 12]. The CdZnTe detector works as
imaging is attenuation correction and anatomic referencing a semiconductor that directly converts the incident γ-rays
of SPECT lesions. Furthermore, when SPECT/CT is per- into an electric signal. This process results in better spatial
formed for treatment monitoring and follow-up, low-dose resolution and sensitivity, which translates into the possibil-
CT should be sufficient. Therefore, the use of low-dose, ity of administering lower activities of the radiopharmaceu-
noncontrast-enhanced CT is performed in most cases [10]. tical and/or of setting shorter acquisition times.
In dual-modality imaging, accurate image co-registration Semiconductor detectors are also characterized by improved
is possible if SPECT and CT data are acquired while the energy resolution (typically resulting in a 30% reduction of
patient maintains the same body position during the acquisi- the scatter counts in the peak energy window) with respect
tion of both the emission and transmission data. In particular, to scintillation detectors [13, 14].
Fig. 7.7 [123I]MIBG axial, coronal, and sagittal sections obtained with Despite its low spatial resolution in this first-generation instrumenta-
a first-generation SPECT/CT system (CT component in top row; tion, the CT component provides useful information for localization of
SPECT component in middle row; fused SPECT/CT in bottom row). the SPECT findings, in addition to correction for attenuation
Furthermore, the compact form factor of CdZnTe detectors these detectors allows simultaneous dual isotope imaging
allows for advanced detector geometries and, in turn, high- (e.g., 99mTc and 201Tl for stress and rest myocardial perfusion)
sensitivity collimation. The most frequently used systems are [15]. Because of their high sensitivity and acquisition of
D-SPECT (Spectrum Dynamics, Haifa, Israel), which uses simultaneous views, these systems might lead to re-evaluation
pixelated CdZnTe detector arrays in nine vertical columns of tracers that are rapidly cleared from the myocardium, such
mounted in a 90° gantry geometry, and Discovery 530c (GE as 99mTc-boronic acid teboroxime. Moreover, the high sensi-
Healthcare, Haifa, Israel), which is based on a multi-pinhole tivity of these CdZnTe systems and the possibility of acquiring
collimator system and an array of 19 CdZnTe pixelated detec- dynamic SPECT have renewed interest in first-pass radionu-
tors. These new SPECT systems enable to adopt shorter acqui- clide angiocardiography, according to a novel type of SPECT
sition times and fast myocardial perfusion scan protocols acquisition [15]. Moreover, dynamic SPECT acquisitions also
(significantly shorter than 1 h) while occupying the camera for allow the use of kinetic perfusion tracer modeling to measure
no more than 15 min total. The improved energy resolution of the coronary flow reserve [15].
Fig. 7.8 Correction of misalignment due to motion of the patient modalities. The bottom panel displays the hybrid alignment modified
between the CT and the SPECT acquisitions. The top panel displays the by the user by manually translating and/or rotating the SPECT compo-
preliminary alignment between SPECT and CT components. An auto- nent in all three orthogonal planes, in order to better match the CT com-
matic contour defined on the SPECT component is superimposed onto ponent. Finally, SPECT will be reconstructed and more properly
the CT component, helping to check correct alignment between the two corrected for scatter and attenuation
7.3.2 Quantitative SPECT

Key Learning Points
• The CdZnTe detectors provide better spatial resolu- SPECT images can in principle be quantitative in absolute
tion, energy resolution, and sensitivity than scintil- terms, i.e., provide voxel values representing the local con-
lation detectors. centration of radioactivity. However, in contrast to PET, this
• CdZnTe detectors can have advanced detector is often not the case in the routine practice, because of the
geometries that allow acquisition of dynamic confounding effects due to scatter and attenuation as well as
SPECT. the limited counting statistics and partial volume effect
caused by the finite spatial resolution of a SPECT imaging
Fig. 7.9 SPECT/CT acquisition of a uniform ACR phantom (well counter) containing a well-known concentration of 166
Ho used to obtain the
prerequisite cross calibration factor for quantitative SPECT and dosimetry
system [16]. Although accurate correction for these effects

remains more challenging for SPECT than for PET, informa- Key Learning Points
tion derived from the CT component of a SPECT/CT scanner • Partial volume effect, scatter, and attenuation limit
can help compensate for photon attenuation and scattered the use/reliability of quantitative SPECT in the clin-
radiation using patient-specific images derived from CT ical routine.
[17]. When fused CT data are available, they can also be • Thanks to improvements in reconstruction algo-
used to define the size and shape of target regions, therefore rithms and SPECT/CT systems, a quantitative
to correct SPECT data for partial volume errors, taking into SPECT/CT approach is feasible and particularly
account the “spill-out” of radioactivity from the target into appealing for radiation dose calculation.
the background and the “spill-in” from the background into
the target region.
Thanks to the diffusion of hybrid SPECT/CT imaging
systems and to improved reconstruction algorithms, SPECT References
is increasingly being used to quantify uptake of the radio-
pharmaceutical in a target region by defining volumes of 1. Kohli V, King MA, Pan TS, Glick SJ. Compensation for distance-
dependent resolution in cardiac-perfusion SPECT: impact on uni-
interest around the target region to integrate the number of formity of wall counts and wall thickness. Nucl Sci IEEE Trans.
counts, assumed to be proportional to the activity (Bq). 1998;45:1104–10.
Regional activity can be calculated using a calibration factor 2. Groch MW, Erwin WD. SPECT in the year 2000: basic principles.
obtained by acquiring a uniform ACR phantom containing a J Nucl Med Technol. 2000;28:233–44.
3. Lange K, Carson RE. EM reconstruction algorithms for emis-
known concentration of the same radionuclide used to image sion and transmission tomography. J Comput Assist Tomogr.
the patient, since the calibration factor is specific for every 1984;8:306–16.
radionuclide as for each intrinsic detector sensitivity and col- 4. Hudson HM, Larkin RS. Accelerated image reconstruction using
limator used (Fig. 7.9). Moreover, due to nonlinearities of ordered subsets of projection data. IEEE Trans Med Imaging.
1994;13:601–9.
the detector at different count rates and dead-time effects at 5. Tsui BMW, Zhao XD, Frey EC, McCartney WH. Quantitative
high activities, count rate-dependent calibration factors for SPECT: basics and clinical considerations. Semin Nucl Med.
the same radionuclide should be applied as well, especially 1994;24:38–65.
for high-energy radionuclides. 6. Pretorius PH, King MA, Pan TS, de Vries DJ, Glick SJ, Byrne
CL. Reducing the influence of the partial volume effect on SPECT
The quantitative SPECT/CT approach allows estimation activity quantitation with 3D modelling of spatial resolution in iter-
of the radionuclide content in a certain voxel, information ative reconstruction. Phys Med Biol. 1998;43:407–20.
that can be used to calculate the radiation dose delivered 7. Khalil MM, editor. Basic sciences of nuclear medicine. Berlin:
within a target by applying voxel dosimetry. Springer; 2010.
8. Kuehl H, Veit P, Rosenbaum SJ, Bockisch A, Antoch G. Can PET/ design, system validation and future potential. Eur J Nucl Med Mol
CT replace separate diagnostic CT for cancer imaging? Optimizing Imaging. 2010;37:1887–02.
CT protocols for imaging cancers of the chest and abdomen. J Nucl 14. Duvall WL, Croft LB, Ginsberg LE. Reduced isotope dose and
Med. 2007;48:45S–57S. imaging time with a high-efficiency CZT SPECT camera. J Nucl
9. Buck AK, Nekolla S, Ziegler S, Beer A, Krause BJ, Herrmann K, Cardiol. 2011;18:847–57.
et al. SPECT/CT. J Nucl Med. 2008;49:1305–19. 15. Nkoulou R, Fuchs TA, Pazhenkottil AP, Kuest SM, Ghadri JR,
10. Slomka PJ, Patton JA, Berman DS, Germano G. Advances in tech- Stehli J, et al. Absolute myocardial blood flow and flow reserve
nical aspects of myocardial perfusion SPECT imaging. J Nucl assessed by gated SPECT with cadmium-zinc-telluride detec-
Cardiol. 2009;16:255–76. tors using 99mTc-tetrofosmin: head-to-head comparison with
11. Bocher M, Blevis IM, Tsukerman L, Shrem Y, Kovalski G, Volokh 13
N-ammonia PET. J Nucl Med. 2016;57:1887–92.
L. A fast cardiac gamma camera with dynamic SPECT capabilities: 16. Laurette I, Zeng GL, Welch A, Christian PE, Gullberg GT. A three-
design, system validation and future potential. Eur J Nucl Med Mol dimensional ray-driven attenuation, scatter and geometric response
Imaging. 2010;37:1887–902. correction technique for SPECT in inhomogeneous media. Phys
12. Duvall WL, Croft LB, Godiwala T. Reduced isotope dose with Med Biol. 2000;45:3459–80.
rapid SPECT MPI imaging: initial experience with a CZT SPECT 17. Seo Y, Wong KH, Sun M, Franc BL, Hawkins RA, Hasegawa
camera. J Nucl Cardiol. 2010;17:1009–14. BH. Correction of photon attenuation and collimator response
13. Bocher M, Blevis IM, Tsukerman L, Shrem Y, Kovalski G, Volokh for a body-contouring SPECT/CT imaging system. J Nucl Med.
L. A fast cardiac gamma camera with dynamic SPECT capabilities: 2005;46:868–77.
Principles of CT and MR imaging
8
Christian Bracco, Daniele Regge, Michele Stasi,
Michela Gabelloni, and Emanuele Neri
Contents
8.1 Physical Principles of CT 188
8.1.1 Hounsfield Unit 188
8.2 The CT Imaging System 188
8.2.1 Current Acquisition Configuration 188
8.2.2 Gantry Technology 189
8.2.3 Tube Technology 189
8.2.4 Detector Technology 189
8.2.5 Scintillators for CT 189
8.2.6 Photovoltaic Detector Array (PDA) 189
8.2.7 Anti-Scatter Grids 190
8.2.8 Multi Detector CT 190
8.2.9 Adjustable Parameters for CT Image Acquisition 191
8.2.10 Pitch 191
8.2.11 Image Reconstruction 191
8.3 CT Dose Index 192
8.3.1 Dose in CT, CTDI, CTDIw, DLP, and SSDE 192
8.4 Physical Principles of MRI 193
8.4.1 The Source of the MR Signal 193
8.4.2 Signal Formation 194
8.4.3 Signal Localization and Image Formation 194
8.4.4 Weightings and Image Contrast 195
8.4.5 Sequences 195
8.4.6 MRI Hardware 196
8.4.7 MR Safety 197
References 198
C. Bracco · M. Stasi
Department of Medical Physics, Institute for Cancer research and
Treatment, Candiolo, Turin, Italy Learning Objectives
D. Regge • To understand the physical principles of interaction
Department of Surgical Sciences, University Hospital “Città della between X-rays and tissue
Salute e della Scienza”, University of Turin, Turin, Italy
• To learn the basic principles of image formation in X-ray
Department of Radiology, Institute for Cancer Research and computed tomography
Treatment, Candiolo, Turin, Italy
• To learn the basic principle of radiation dose estimation
M. Gabelloni · E. Neri (*) and measurements
Diagnostic and Interventional Radiology, Department of Translational
• To learn the basic principles of MRI: magnetic field,
University of Pisa, Pisa, Italy sequences, signal parameters, and coils
e-mail: emanuele.neri@med.unipi.it • To learn the safety risks in MRI

188 C. Bracco et al.
8.1 Physical Principles of CT substances except water and air might show variations in HU
values when they are evaluated at different tube voltages.
The formation of CT images involves measurement of the Usually, for each tube voltage, the CT vendors carry out an
X-ray transmission profile through a patient over a large adequate calibration of the HU scale by acquiring images of
number of views. Each profile is acquired by means of a fan- air- and water-filled phantoms.
shaped X-ray beam that penetrates the body. The X-rays
exiting the scanned patient’s body are recorded by a detector
arc, generally consisting of 800–1000 detector elements, Key Learning Points
referred to as a detector row. Each projection of detector’s • The image signal relies on the photoelectric effect,
record (transmission profile) is a view at one position (angle) whereas Compton scatter degrades the image qual-
of the X-ray source. Multiple angular profiles are collected ity by adding noise.
during one complete rotation of the X-ray tube around the • Computed tomography measures the attenuation of
patient, which are used to reconstruct the CT image (basi- X-ray beams passing through sections of the body
cally a square matrix of voxels) of the internal organs and from hundreds of different angles.
tissues for each complete rotation of the X-ray source. • The Hounsfield unit (HU) scale is expressed rela-
Along the pathway through the body, X-rays interact with tive to the linear attenuation coefficient of water at
the matter. In the energy range typically used in a CT scan- room temperature.
ner, there are two principal ways in which X-rays interact
with tissues: photoelectric effect and Compton (incoherent)
scatter. The image signal relies on the photoelectric effect,
whereas Compton scatter degrades the image quality by add- 8.2 The CT Imaging System
ing noise.
The main effect of X-rays’ interaction with matter is 8.2.1 Current Acquisition Configuration
attenuation of the X-ray beam; thereby the detector records
the mean absorption by all the different tissues which the Since its first appearance in the 1970s, CT systems devel-
X-ray beam has penetrated. “Computed tomography, on the oped rapidly as a valuable tool in diagnostic radiology. The
other hand, measures the attenuation of X-ray beams passing geometry of modern CT systems is still similar to scanners
through sections of the body from hundreds of different sold in late 1975: a full X-ray fan beam that covers the com-
angles, and then, from the evidence of these measurements, plete field of view and a simultaneously rotating tube and
a computer is able to reconstruct pictures of the body’s rotating detector array [2]. In 1987, continuously rotating
interior.” [1] systems based on slip-ring technology became available.
In the reconstructed image, each voxel is associated with Typical rotation time was 1 s. Helical CT was introduced in
the linear attenuation coefficient μ (m−1) of the correspond- 1989 and led to a breakthrough in truly three-dimensional
ing tissue. The linear attenuation coefficient μ is a function of (3D) CT imaging. Following the development of larger
photon energy and material density. The X-ray attenuation detector arrays with more detector rows, dual-source CT
through the body is governed by the Beer-Lambert law. (DSCT) was introduced in 2005. Such technology was
intended to further increase scan speed and temporal resolu-
tion, two requirements for cardiac CT.
8.1.1 Hounsfield Unit To date the CT source and detection technologies have
evolved to satisfy three major CT imaging concerns: (1)
Since the earliest days of CT, the attenuation value calcu- increasing the axial field of view, (2) decreasing the rotation
lated for each voxel of the reconstructed matrix is replaced time to improve temporal resolution and scan speed, and (3)
with an integer number called Hounsfield unit, in honor last but not least, reducing the radiation dose to the patients.
of its inventor. Large part of the human body is made up The emergence of multislice CT scanners was possible, thanks
of water and surrounded by air, and this was the biologi- to solid-state detectors segmented into detector elements
cal background for creating the linear Hounsfield scale. arrays. CT sources evolved to support larger coverage per rota-
The Hounsfield unit (HU) scale is expressed relative to tion. The high speed of acquisition required X-ray sources
the linear attenuation coefficient of water at room capable of supporting great accelerations and increased peak
temperature. power, while at the same time, the detection systems were
Distilled water at standard pressure and temperature is designed to have short integration periods. Furthermore, over
defined as 0 HU; the radiodensity of air is defined as the last few years, the availability of iterative reconstruction
−1000 HU. Density of other tissues is related to this range, made the use of low and ultra-low dose acquisitions possible
usually from −1000 HU to +1000 HU. HU = 1 is associated to be used routinely in clinical practice. This has a direct
with 0.1% of the linear attenuation coefficient of water. All impact on the need for lower noise detection systems [3].
8 Principles of CT and MR imaging 189
8.2.2 Gantry Technology tubes ranges from 10,000 to 40,000 h, compared with the
1000 h typical of conventional CT tubes [5].
The introduction of slip-ring technology to transport data
and energy to and from the gantry enabled continuous rota-
tion of the X-ray tube and detector. This was an essential 8.2.4 Detector Technology
prerequisite for the development of spiral CT scanners. Slip
rings are electromechanical devices made up of circular elec- The CT detector converts the X-rays into visible light, which
trical conductive rings concentric to the gantry axis and is in turn converted into an electric current. Typically a CT
brushes that transmit electrical power and all of the control detector is made up of three essential layers: (1) a scintillator
signals from the stationary parts of the gantry to the rotating to convert X-rays into light, (2) a photodiode to convert light
frame. These technological improvements allow the scan into current, and (3) a substrate to provide electrical and
frame to rotate continuously with no need to stop between mechanical infrastructure.
rotations to rewind the system cables [4].
The latest CT systems with gantry rotation times down to
0.25 s use noncontact data and energy transfer between the 8.2.5 Scintillators for CT
rotating and stationary parts of the gantry. Rotating hardware
undergoes huge centrifugal forces on the order of 40 × g. A scintillator is a luminescent material able to convert lin-
This significantly affects the X-ray tube design for mechani- early high-energy photons into visible light. The scintillator
cal reasons. Larger patient body sizes and the need to per- is optically coupled to matching silicon p–i–n photodiode
form whole-body scans additionally require new innovations matrices. The requirements for the scintillator material to be
regarding the patient table. To realize fast and particularly used in X-ray CT can be summarized as follows [6]:
high-pitch scanning (table speeds of up to 74 cm/s are cur-
rently in use), the acceleration and deceleration range of the • High absorption for X-rays in the energy range up to
table must guarantee patient safety and comfort. 150 keV
The CT table has also been adapted to handle increasing • High light output
patient’s body weight. Larger patients require expanded • Radioluminescence spectrum in the visible or near IR range
table capacity limits as well as more noteworthy gantry width to match the spectral sensitivity of silicon photodetectors
(up to 90 cm in dedicated systems), broadened reconstruc- • Decay time in the range of 1–10 μs
tion field of view, and higher tube output. Also, a larger bore • No afterglow or ghosting
width provides more space to CT-guided interventions as • Good radiation hardness for high X-ray fluence
well as uncompromised patient positioning [2]. • Small temperature dependence of the light yield
• Good mechanical properties
• Affordable cost
8.2.3 Tube Technology
These requirements make polycrystalline ceramics and single
X-ray tubes are subjected to far higher thermal loads in CT than crystals the most convenient types of scintillators. Scintillators
in any other diagnostic X-ray application. Shorter scan times in for multislice CT scanner are built in two-dimensional (2D)
current CT scanners require high-power X-ray tubes and utili- arrays, with a pixel size around 1 mm. The arrays packaging also
zation of oil-cooled rotating anodes for efficient thermal dissi- incorporates a reflective material. The reason for the reflective
pation. Typical maximum tube power is on the order of 100 kW material is both to mechanically support the pixelated scintillator
or more. The advent of helical CT with nonstop scanner rota- array and to efficiently transport the scintillation light to the pho-
tion has introduced additional requirements for X-ray tubes. todetector, with minimal cross talk [3].
Several technical advances in component design have been
made to achieve these power levels and manage the issues of
target temperature and heat dissipation. As scan times have 8.2.6 Photovoltaic Detector Array (PDA)
diminished, the anode heat capacities have expanded by as
much as a factor of 5, avoiding the need for cooling delays dur- To date the light exiting from scintillators is collected by a
ing most clinical procedures. Tubes with capacities of 5–8 mil- photovoltaic detector array (PDA), which converts it into an
lion heat units are available. Furthermore, improvement in the electric current. In traditional detector design, the small ana-
heat dissipation rate (kilo–heat units per minute) has increased logic electrical current from the photodiode was amplified and
the heat storage capacity of modern X-ray tubes. The huge heat changed into a digital signal on an external board, requiring
capacities are achieved with thick graphite backing of target analogic connections between the detector and the electronic
disks, anode diameters of 200 mm or more, improved high- circuit components on the external board. These types of
temperature rotor bearings, and metal housings with ceramic detectors are referred to as conventional detectors. Current
insulators among other factors. The working life of current commercial CT systems are equipped with detectors featuring
fully integrated electronics. In these detectors, the photodiode, 8.2.8 Multi Detector CT
the preamplifiers, and the analog-to-digital converters are
integrated into the same silicon chip that is attached to the The term multislice CT (MSCT) scanner refers to a CT
scintillation ceramic, thus avoiding the need for analog con- system in which multiple-row detector arrays are assem-
nections. Such integration reduces the time during which the bled to simultaneously acquire data at different slice loca-
signal is in an analog format, so the amount of electronic noise tions [10]. Clinical applications benefit from multi-detector
that can be superimposed to the signal is reduced. These types row CT (MDCT) technology in several ways: (1) shorter
of detectors are referred to as integrated detectors [7]. scan time, (2) large longitudinal (z) scan, (3) continuous
Currently, semiconductor-based direct converting detec- data acquisition (enabling the images to be reconstruct at
tors are under development for CT imaging. With such as any z position), and (4) improved longitudinal resolution.
innovation, each detected X-ray photon generates a very The CT detectors must provide different slice widths to
short pulse enough to count each photon. In addition, the adjust the optimum scan speed, longitudinal resolution, and
area under the pulse, and thus the pulse height, is propor- image noise for each application. Earlier CT scanners were
tional to the photon energy. The goal is to design photon- equipped with a single detector. Different slice widths were
counting detectors with energy discrimination capabilities to acquired by means of pre-patient X-ray collimation. To
capture spectral X-ray information without the need to apply acquire more than one slice per rotation, CT scanners were
two different X-ray tube voltages [2]. then designed with a greater number of detectors. In 1998,
all major CT vendors introduced MSCT systems, which
typically offered simultaneous acquisition of four slices at a
8.2.7 Anti-Scatter Grids rotation time of down to 0.5 s [11]. Since then, the evolution
of CT acquisition has followed its own form of “Moore’s
Scatter is a significant source of image artifacts in CT. Scatter Law”; in particular, growth in the number of slices has been
management includes both scatter rejection and scatter exponential, doubling every 18 months. As a next step, the
correction via software. Scatter reduction becomes even
introduction of an 8-slice CT system in 2000 enabled
more important for large CT detectors with more than 64 shorter scan times but did not yet provide improved longitu-
rows, because image quality is influenced by the increasing dinal resolution (thinnest collimation 8 × 1.25 mm). The
quantity of scattered photons for large cone angles. latter was achieved with the development of 16-slice CT,
Furthermore, larger patients increase the scatter problem that which made it possible to routinely acquire substantial ana-
degrades image quality. tomic volumes with isotropic submillimeter spatial resolu-
Conventional CT scanners (with detector coverage not tion [12]. In 2004 the next generation of multislice CT
exceeding 40 mm along the patient axis) usually employ systems became commercially available, with 32, 40, and
one-dimensional (1D) anti-scatter grids (ASG) mounted in even 64 simultaneously acquired slices, with a strong
front of the detector. Scattered X-ray photons are suppressed impact on volume coverage speed. Similar to some estab-
by using lamellas made of X-ray absorbing material posi- lished 4-slice and 16-slice CT systems, the 64-slice scanner
tioned between the active detector cells [8]. Materials used to has an adaptive array detector design with different sizes of
make anti-scatter lamellas are of high Z numbers in order to the detector rows in the longitudinal direction. The inner-
allow effective absorption of scattered radiation. To ensure most elements can be used to collect thin slices or com-
adequate image quality in wide-cone CT, a two-dimensional bined in pairs to acquire thicker slices. Some vendors make
(2D) ASG is needed [9]. use of a periodic motion of the focal spot in longitudinal
direction (z-flying focal spot) to improve data sampling
along the z-axis and to double the number of simultane-
Key Learning Points ously acquired slices. To date the 320 detector row CT scan-
• Geometry of modern CT systems is based on a full ner represents the current pinnacle in the nominal number
X-ray fan beam that covers the complete field of view, of slices that can be scanned per gantry rotation. A detailed
coupled with a simultaneously rotating tube. story of the race for the increase of the axial coverage in
• Current CT technology is designed to increase the clinical CT scanners can be found in [13].
axial field of view, shorten the rotation time (to After 2007, the clinical market has reached a plateau in
increase time resolution and scan speed), and to terms of number of simultaneously acquired slices, consider-
reduce the radiation dose to the patients. ing some issues about patient dose, image quality, and cost.
• A combination of high-speed gantry rotation, tubes On the other hand, the advent of dual-energy CT (DECT)
with high heat dissipation rate, high performance seems now to have pulled manufacturers into a new battle-
scintillators, and new detectors with fully integrated field: the number of energy bin must grow instead of the
electronics, are the key components of modern mul- number of slices, even though true energy-resolved scanners
tirow CTs. based on photon-counting detectors are not yet mature for
clinical use.
8.2.9 A
djustable Parameters for CT Image
Acquisition tion, Tube current modulation, Tube voltage, Tube
voltage modification, Effective tube current-time
The common CT acquisition parameters are: product.
• The pitch is the table distance traveled in one rota-
• Tube current-time product (milliampere second, or mAs) is tion divided by total thickness of all simultaneously
the product of the X-ray tube current (in milliamperes) and acquired slices.
the CT scanner exposure time per rotation (in seconds).
• Tube current modulation (mA modulation) is an essential
tool to ensure proper patient exposure during CT exami- 8.2.11 Image Reconstruction
nations. It allows the tube current to be actively modu-
lated during the scan to reduce radiation to the patient. Image reconstruction in CT is a mathematical process that
• Tube voltage (kV): The X-ray tube potential indicates the generates tomographic images from X-ray projection data
peak energy of the X-ray photons (in kilovolts) in a spec- acquired at many different angles around the patient.
trum of X-ray energies. Analytical reconstruction algorithms, among these the
• Tube voltage modification (kV modification): Different Filtered Back Projection (FBP), assume that both the projec-
tube voltages can be selected for the CT examination tion data and the measurement process can be described by
depending on the size of the patient or the type of CT continuous functions. In a simplified model, a projection is
examination being performed. Lowering tube voltage obtained by drawing a set of narrow (pencil beam) X-rays
from 120 to 100 or 80 kV makes it possible to reduce the through the patient. Each single pixel in the projection is the
radiation dose provided to the patient and is advised for integral of the patient’s density along each pencil beam. A
small and average-sized patients. single projection of the patient is one-dimensional. CT
• Effective tube current-time product (effective mAs) is reconstruction of the patient requires several projections at
described as the ratio of tube current-time product to different angles between the pencil beam and the patient.
pitch. With this concept, the variation of pitch is exactly The X-ray source is rotated by an angle α and a new projec-
compensated by changing tube current or rotation time to tion is acquired, and the process keeps going until all the
maintain the same image quality [14]. necessary projections are acquired.
The collection of projections at several angles is called
sinogram. The inverse Radon transform is used in CT to
8.2.10 Pitch reconstruct a 2D image from the sinogram. The FBP is the
fastest tool to perform the inverse Radon transform. Each
Helical scanning with multislice CT scanners is indistin- projection is convoluted to a filter to compensate for the
guishable from that with single-slice CT scanners: table effect of the so-called low-pass blur that occurs because of
movement and tube-detector rotation occur simultaneously the different numbers of projections passing through the cen-
with continuous data acquisition. ter and the periphery, respectively, of an object.
For single-slice CT, detector pitch was defined as table In clinical practice, several filters (kernels) are available to
movement per rotation divided by slice thickness. For meet the requirements of spatial resolution and image noise.
instance, with a slice thickness of 10 mm and a table move- Increasing the strength of the low-pass blur increments the
ment of 15 mm per rotation, pitch would be 1.5. “sharpness” of the image, as well as expands image noise.
In multislice CT, beam pitch is defined as table distance Different kernels allow optimized depiction of high-contrast
traveled in one rotation divided by total thickness of all structures or soft tissue, such as lung tissue or bone. Image
simultaneously acquired slices. Pitches of greater than 1 sharpness and image noise are intrinsically coupled in image
mean that there are gaps between the X-ray beams from reconstruction via FBP: the sharper the image, the higher the
adjacent rotations. Pitches of less than 1 mean X-ray beam image noise. A major limiting feature of FBP is that it misses
overlap (and thus double irradiation of some tissue) and so the capability to account for image noise resulting from
are generally not clinically used [15]. Poisson statistical variations in photon number over the image
plane and system hardware details (focal spot size, active
detector area, and image voxel shape); in practice, the effort
Key Learning Points to reduce the radiation dose to patients translates into an
• The term multislice CT (MSCT) scanner refers to a increase in image noise. Delineation and low-contrast detect-
CT system in which multiple-row detector arrays are ability of a structure suffer from high image noise and arti-
assembled to simultaneously acquire data at differ- facts, so that to generate a diagnostic CT data set the radiation
ent slice locations. dose may not be lowered below certain minimal constraints.
• The CT acquisition parameters are: Tube current- Lowering image noise by choosing “smoother” kernels for
time product, CT scanner exposure time per rota- image reconstruction will result in impaired spatial resolution
when using a conventional FBP technique. [16]
The recent augmented computational capacities in normal involving ionizing radiation, CT will be responsible for the
workstations and the on-going attempts to reduce doses in major portion of the total collective dose delivered to the
CT have focused the attention of manufacturers on iterative public [18].
reconstruction (IR). All IR methods consist of three steps CT manufacturers designed the CT scanner to keep the
which are repeated iteratively. First, a forward projection of radiation dose to the patient as low as reasonably achievable
the object estimate creates synthetic raw data which, in a sec- (ALARA). X-ray source, detectors, and reconstruction algo-
ond step, are compared to the real acquired raw data in order rithm are involved in the process of dose reduction. Following
to calculate a correction term. In the last step, the correction CT evolution over the years, physicists developed new dose
term is back-projected onto the volumetric object estimate. indexes to evaluate and communicate the dose imparted in
The iteration process can start with an empty image estimate CT examinations. The AAPM report 96 and the AAPM
or using prior information, for example, a standard FBP report 204 are the most valuable references for detailed and
reconstruction or a volume of a similar object. In general, the standard definition of CT dose index currently in use. Report
process may converge faster toward a stable solution if the 96 also provides an overview on the methods for dose reduc-
initial images match the final images. The iterative process tion in CT. Hereinafter, the most commonly used CT dose
ends when either a maximum number of iterations are index parameters are discussed.
reached, or the update for the current image estimate is con- The primary quantity for assessing the dose in CT is the
sidered small enough, or when a predefined quality criterion computed tomography dose index (CTDI). CTDI repre-
in the image estimate is satisfied. More detailed information sents the average absorbed dose, along the z-axis, from a
about IR may be found in [17]. series of contiguous irradiations. It is measured from one
Iterative reconstruction algorithms have several advan- axial CT scan and is calculated by dividing the integrated
tages compared to analytical FBP-based methods: different absorbed dose by the nominal total beam collimation.
physical models might be integrated in the reconstruction CTDI100 is a linear measure of dose distribution over a
process, thus reducing image noise and artifacts depending 100 mm length pencil ionization chamber; since this param-
on the degree of modelling. They do not introduce new arti- eter underestimates dose for longer scan lengths and does not
facts related to approximations that are often exploited in take into account the topographical variations of a human
analytical reconstruction methods, and they are more body, it is not in clinical use. In fact, the CTDI100 across the
appropriate for dealing with missing data or irregular sam- field of view (FOV) is not constant. For instance, for a body
pling. The biggest advantage of IR methods is generally CT imaging, the CTDI100 at the surface is almost double or
considered the intrinsic potential for patient dose higher than the CTDI100 at the center of FOV. The weighted
reduction. CTDIw is defined as the sum of one-third the CTDI100 at the
All iterative reconstruction methods have the main disad- center and two-thirds the CTDI100 at the edge of a phantom.
vantage of requiring increased computational effort that CTDIw is closer to the human dose profile as compared with
becomes necessary due to multiple iterations instead of a the CTDI100.
single iteration as for analytical methods. To represent dose for a specific scan protocol, which
Also the image quality of reconstructed image depends on almost always involves more than one scan, it is essential to
the prior information used for the parameterization of the consider any overlaps or gaps between the X-ray beams
acquisition model, and the necessary number of iterations from consecutive rotations of the X-ray source. By dividing
might vary strongly depending on the measured object. the CTDIw by the pitch factor, one obtains the CTDIvol.
Different noise patterns and some artifacts that manifest While CTDIw is representative of the average absorbed
themselves in a way different from analytical reconstruction dose over the x and y directions at the center of scan,
methods had impact on the acceptance in clinical routine of CTDIvol is the average absorbed radiation dose over the x, y,
images reconstructed by IR methods. and z directions. This index was designed to offer a stan-
dardized method to compare radiation output levels
between different CT scanners using a reference phantom,
8.3 CT Dose Index usually a 16 cm and 32 cm diameter polymethyl methacry-
late (PMMA) cylindrical phantom often called “head” and
8.3.1 Dose in CT, CTDI, CTDIw, DLP, and SSDE “body” CTDI phantom, respectively. The SI units are mil-
ligray (mGy).
Interest in radiation exposure during CT examinations has To better represent the overall energy imparted by a given
increased during the last years. Scientific literature has scan protocol, the absorbed dose can be integrated along the
recently highlighted the importance of radiation dose from scan length to compute the dose length product (DLP).
X-ray CT examinations and the associated risk. The dose DLP is related to the total ionizing energy delivered to the
level imparted in CT is much greater than those from con- reference phantom attributable to the complete scan acquisi-
ventional radiology and the use of CT keeps growing, often tion and depends on all the adjustable acquisition parameters
by 10–15% per year. Thus, among medical procedures (kVp, mAs, pitch, slice thickness, and so on).
The CTDIvol or DLP, as visualized on consoles, do not Protons, neutrons, electrons, and whole nuclei show an
represent the actual dose administered to the patient; they analogue property known as spin, usually visualized as a
should be considered as an index of radiation output by the small sphere that rotates at a high speed around its axis even
system, which is useful only for comparison purposes. if such a picture is not strictly correct. Spinning masses gen-
Given the huge variability in patient size, CTDIvol is cor- erate an angular momentum. Since the single proton of
rected for the patient size, thus yielding the size-specific hydrogen carries a positive charge, its spin generates an elec-
dose estimate (SSDE); more information about SSDE may trical current similarly as a charge in a looped wire. This loop
be found in AAPM report 204. current provides the proton a torque, referred to as magnetic
It is important to notice that the potential biological momentum, when placed within a magnetic field [20].
effects from radiation depend not only on the radiation dose Nuclei useful for MRI must have both magnetic moment
to an organ or tissue, but also on the biological radiosensitiv- and angular momentum; typically this happens if nuclei have
ity of the tissue or organ irradiated. Effective dose, E, is the odd-numbered atomic mass: 1H, 13C, and 31P. Such nuclei are
dose descriptor that reflects this difference in biologic radio- said to have nuclear magnetic resonance (NMR) property
sensitivity. The units of effective dose are sieverts; generally and are generally referred to as spins. In the simplest classi-
millisieverts (mSv) are used in diagnostic radiology. cal model nuclei having NMR property may be imagined as
“Effective dose aims to provide a single number that is pro- small magnets or dipoles. It is well known that if a compass
portional to the radiobiological detriment from a particular, is placed close to a powerful magnet, the compass needle
often inhomogeneous, type of radiation exposure” [19]. aligns itself with the field in any orientation. If an external
magnetic field (B0) is applied, the hydrogen proton having a
spin quantum number ±½ tends to align with the field, but, as
known from quantum theory, only two states are allowed:
Key Learning Points with and against the direction of B0. In particular, the angle
• In CT images are reconstructed through a mathe- between the spin and the magnetic field B0 is 54° 44′ [21].
matical process that generates tomographic images Each state corresponds to an energy level: spins pointing
from X-ray projection data acquired at many differ- with B0 (also referred to as parallel or spin-up) are in a lower
ent angles around the patient. energy level, which is the preferred alignment; spins pointing
• The primary quantity for assessing the radiation against B0 (also referred to as antiparallel or spin-down) are in
dose in CT is the computed tomography dose index. a higher energy level. The energy-level difference, or energy
• The absorbed dose can be integrated along the scan gap, is given by the product of the proton gyromagnetic ratio,
length to compute the dose length product (DLP). the external magnetic field intensity |B0|, and the reduced
Planck constant ħ. Once known the energy-level difference,
which depends on the magnetic field strength, the distribution
at room temperature of spin-up and spin-down may be calcu-
8.4 Physical Principles of MRI lated by means of Boltzmann equation. For commonly used
magnets, operating at 1.5 T (Tesla), the population ratio is
8.4.1 The Source of the MR Signal approximately 1,000,000–1,000,009 [22]. Thus, at 1.5 T, only
nine hydrogen protons out of 1,000,000 are oriented with B0.
During the first half of the twentieth century, Rabi, Bloch, and That is why NMR is a fairly insensitive modality. Considering
Purcell discovered the physical principles of magnetic reso- the huge number of hydrogen nuclei (about 1027) available in
nance imaging (MRI). MRI is now an essential imaging tool in the body, the net magnetization is still measurable. It is pro-
modern medicine to image anatomy and other functions within portional to the field strength: a large field produces a greater
the body. MRI is capable of producing images of a patient magnetization and better signal-to-noise ratio [23]. This may
exploiting the abundance of hydrogen atoms within the tissues, explain why image quality increases at higher field strength.
mainly from water and fat. In contrast to X-ray computed Averaging the orientations of these few spins makes the
tomography, MRI is based on radio-frequency electromagnetic net magnetization vector M to appear; this vector can have
fields (RF-EMF). Typically, RF-EMF do not have enough any orientation with respect the magnetic field. The net mag-
energy to ionize the matter, thus avoiding the risk of X-irradiation. netization vector has a wobbling or precessional motion
Quantum mechanics describes what happens to tissues when around B0 at the Larmor frequency.
they are placed in a magnetic field. A simple classical model The central equation of MRI ω = γ|B0| is known as Larmor
may be sufficient to explain all the theory behind MRI. According equation, where γ is the gyromagnetic ratio and |B0| is the
to classical mechanics, an object rotating around its axis carries intensity of external field. The frequency ω is also known as
a quantity known as angular momentum that can be thought as resonant frequency or Larmor frequency.
a rotational inertia, reflecting the object’s size, shape, mass, and Transitions between the spin-up and the spin-down state
rotational velocity. It is typically represented as a vector point- can be induced by the absorption or the emission of electro-
ing along the axis of rotation. magnetic radiation of angular frequency exactly equating
the Larmor frequency. When the magnetic field is a few to as free induction decay (FID). The time necessary to
Tesla, the Larmor frequency may range between 1 and recover the thermal equilibrium is the relaxation time,
100 MHz. More specifically, at 1 T the Larmor frequency is which is constant for a given substance at a given magnetic
42.6 MHz [24]. field strength.
Two independent relaxation processes exist: longitudinal
relaxation and transverse relaxation. Both processes are
characterized by a particular time. The time taken by M to
Key Learning Points
recover 63% of its equilibrium value after a 90° rotation is
• MRI produces images of a patient exploiting the
called T1, or longitudinal relaxation time, and arises from the
abundance of hydrogen atoms within tissues,
interaction between the spins and the atomic neighborhood.
mainly from water and fat.
The second relaxation process that occurs simultaneously to
• Nuclei useful for MRI must have both magnetic
T1 relaxation is due to the spin-spin interaction and inhomo-
moment and angular momentum; typically this hap-
geneity of the main static magnetic field B0. It affects the
pens if nuclei have odd-numbered atomic mass,
phase coherence with which the spins precess around the
such as 1H, 13C, and 31P.
magnetization vector. Thus, as time passes, the signal
• The fundamental equation of MRI ω = γ|B0| is
recorded by the coil decreases because the spins begin to lose
known as Larmor equation, where γ is the gyromag-
their phase coherence (i.e., one spin becomes out of phase
netic ratio and |B0| is the intensity of the external
with another spin).
field.
Phase dispersion is mainly attributable to variation in the
• Frequency ω is also known as resonant frequency or
local precessional frequencies induced by random local
Larmor frequency.
magnetic inhomogeneities. Transverse relaxation is charac-
terized by T2 relaxation time. T2 relaxation time is the time at
which the signal decays to 37% of its original value. T2 time
8.4.2 Signal Formation is always shorter than T1 time [26]. In practice, the signal
decays with a faster time constant T2 because B0 is not
Seen from a frame of reference rotating at the common fre- uniform.
quency, the net magnetization vector M arising from the
hydrogen nuclei inside the magnetic field in its equilibrium
state is stationary, “so even though the individual spins are 8.4.3 S
ignal Localization and Image
precessing, there is no net emission of signal in equilibrium” Formation
[25]. Thus, to obtain useful information from the precessing
protons, we need to alter the direction of M. Radio-frequency Gradients are employed to localize the MR signal in a region
(RF) energy pulses of exactly the Larmor frequency are used of interest. Gradients are additional linear variations of the
to modify the M direction. [26] magnetic static field strength in a selected region. Gradients
RF energy pulses (B1) produce two main effects on the can be applied in any orthogonal direction using the three
protons inside a magnetic field: a certain portion of protons sets of gradient coils, Gx, Gy, and Gz, within the MR system.
are inverted into the excited state spin-down (oriented against In presence of gradients, the frequency of precession will be
B0), thus giving rise to a net magnetization vector oriented in different at different positions in the patient. Faster or slower
the plane transverse to B0. Phase coherence is established as precession is detected as higher or lower MR signal. Thus,
well; this means that nuclear spins are precessing synchro- the frequency measurements can be used to distinguish MR
nously. If the B1 field is continuously applied, the net magne- signals at different positions in space and enable image
tization M is pulled away from its initial alignment with B0. reconstruction in three dimensions [27]. Image reconstruc-
M can be rotated 10°, 90°, 180°, 270°, or any amount. After tion locates the signal within the patient by using the Fourier
every 360° rotation, M returns to its initial alignment with transformation, which is the mathematical operation neces-
B0. The rotating transverse magnetization can generate a sig- sary to break the signal down into its constituents. To select
nal in the receiver coil around the patient or close to the a slice, two steps are necessary: (1) a slice-select gradient is
patient’s surface. imposed along an axis orthogonal to the plane of the desired
Soon after the RF B1 is turned off, the net magnetization slice and (2) an RF pulse with a limited range of frequencies
M tends to realign itself with the external magnetic field B0 or bandwidth is applied to excite only those nuclei that match
to recover the thermal equilibrium. This means that the the frequencies of precession [28].
magnetization decays over time, which implies that the The frequency and phase encoding are used to detect sig-
magnitude of rotating transverse magnetization decreases, nals from each point (pixel) in the selected slice. In order to
thus the signal induced in the receiver coil around the encode the imaged object along the so-called phase-
patient will diminish over time. The phenomenon is referred encoding direction, a gradient along the direction of phase
encoding is applied, and thereafter the signal is sampled. To Image formation requires that similar measurements are
sample the whole object, it is necessary to apply this gradi- repeated several times, e.g., once per line in the image. The
ent several times, each time increasing the gradient by a repetition time, TR, is the length of time between two suc-
fixed amount. The phase-encoding gradient (GPE) is turned ceeding similar measurements or pulse sequence. TR con-
on for a limited time period. While it is applied, it modifies trols the amount of longitudinal magnetization that recovers
the spin resonance frequency, inducing dephasing. between each pulse. Indeed, each measurement makes use
Thereafter the GPE all the protons precess at the same fre- of the longitudinal magnetization present. Short repetition
quency but with different phases. The protons in the same time will result in magnetization not having fully relaxed
row, orthogonal to the GPE direction, have the same phase. back into alignment before the next pulse is made, hence
The frequency encoding is the final step used to differentiate every repeat will therefore only produce a small signal. A
pixels with the same phase encoding, i.e., a single row of the long repetition time allows the magnetization to relax back
object being imaged. A magnetic gradient during readout of into alignment with the main magnetic field. Short or long
the signal results in a specific shift of the precessing fre- TR is relative to the maximum T1 of tissues present in the
quency in the horizontal direction throughout the time it is sample. If TR is significantly longer than the maximum T1,
applied. Phase and frequency information are combined to then magnetization reaches complete equilibrium, and the
fill in a matrix or grid. This grid is called the K-space. T1 contrast disappears. Thus, using a long TR results in lim-
Finally, a fast Fourier transform transforms the raw data in a ited T1-weighting but a strong signal. Conversely, the choice
common image [26]. of applying short TR means that the signal is reduced for all
types of tissue, but the signal becomes more T1-weighted; in
other words, the images will lose intensity but will gain
greater signal variation between tissues with different T1.
Key Learning Points
Finally, both T1-contrast and T2-contrast may be mini-
• Radio-frequency (RF) energy pulses of exactly the mized, thus producing an image whose brightness is deter-
Larmor frequency are used to modify the net mag- mined by the number of protons present in a voxel, also
netization vector direction. known as proton density (PD). In such an image, variation in
• After the Radio-frequency B1 is turned off, the net the water content is the primary source of contrast. Short TR
magnetization vector tends to realign itself with the and TE are commonly chosen to acquire T1-weighted images,
external magnetic field B0. This event is referred to since T1 contrast is maximized and T2 contrast is minimized.
as free induction decay (FID). T2-weighted images are produced if long TR and long TE are
• The time necessary to recover the thermal equilib- chosen, since T1 contrast is minimized and T2 contrast maxi-
rium is the relaxation time, which is constant for a mized. Finally, PD-weighted images are obtained by apply-
given substance at a given magnetic field strength. ing long TR and short TE, since both T1- and T2-contrast are
• Gradients, i.e, additional linear variations of the minimized [23].
magnetic static field strength in a selected region,
are employed to localize the MR signal in a region
of interest. 8.4.5 Sequences
An MRI sequence is an ordered combination of RF and gra-

dient pulses designed to acquire the data to form the image.
8.4.4 Weightings and Image Contrast Several steps are required to acquire the data needed to cre-
ate an MR image. First an RF pulse excites tissue magnetiza-
Soft tissue contrast in MRI is principally due to the differ- tion in the presence of a slice-select gradient. Next, phase
ences among the tissues’ characteristic relaxation times. The encoding and frequency encoding/readout spatially localize
type of pulse sequence and its parameters determines whether the protons in the other two dimensions. Finally, after the
the image contrast is weighted by the T1 or T2 relaxation pro- data have been collected, the process is repeated for a series
cesses. To briefly summarize, two main controls determine of phase-encoding steps.
tissue contrast in MRI: TR and TE. The echo time, TE, is the Spin echo (SE) and gradient-recalled echo (GRE) are the
time between the application of the radio-frequency pulse two fundamental types of MR pulse sequences. All other MR
and the beginning of NMR signal sampling. Meanwhile, the sequences are variations of these, with different parameters
magnetization and the signal will diminish due to T2- added on. MR pulse sequences can be either two-dimensional
relaxation. The echo time controls the T2-weighting in the (2D), with one section acquired at a time, or three-
images. A long TE compared to T2 will thus result in consid- dimensional (3D), with a volume of multiple sections
erable T2-contrast, despite a little signal. obtained in a single acquisition.
In SE sequences, a 90° RF pulse flips the net magneti- reduced magnet length to improve patient acceptance. In
zation M in the transverse plane. As the spinning nuclei order to provide high-quality images, MRI magnets must
experience relaxation, the transverse magnetization is produce a magnetic field with very high temporal and spatial
gradually dephased. Thus, in order to rephase the spinning uniformity, on the order of several parts-per-million (ppm)
nuclei, a 180° RF pulse is applied at the time equal to one over the whole imaging volume. Typically, the 1.5 and 3 T
half the TE. Thus, at the total TE, the nuclei are spinning magnets commercially available can guarantee field unifor-
in phase and an echo is produced and read. In the GRE mity on the order of 10 ppm peak-to-peak in about 50 cm
sequence, an RF pulse is applied to partly flip the net mag- diameter volume. Magnetic shimming is necessary to
netization M into the transverse plane (variable flip angle). improve the magnet uniformity to the design value of 10 ppm
Gradients, instead of RF pulses, are exploited to dephase over the image volume. MRI systems are shimmed with two
(negative gradient) and rephase (positive gradient) trans- shimming methods: first, active shimming that makes use of
verse magnetization [29]. superconducting coils placed in the cryostat. Second, passive
shimming that adopts small pieces of iron located in the
magnet bore. The magnet must be shimmed whenever it is
moved to a new site because both shimming methods are
Key Learning Points site-dependent.
• Soft tissue contrast in MRI is mainly due to the dif-
ferences among the tissues’ characteristic relax- 8.4.6.2 Gradients
ation times. The second component of MRI systems is a set of additional
• The echo time, TE, is the time between the applica- coils of wire located inside the magnetic bore, designed to
tion of the radio-frequency pulse and the beginning create linear magnetic gradients in the three orthogonal
of NMR signal sampling; longer TE results in con- physical directions x, y, and z. Gradient pulses are used to
siderable T2-contrast. add a spatial variation to B0. Thus, spins at different physical
• The repetition time, TR, is the length of time locations will precess at different frequencies. As a result,
between two succeeding similar measurements or the resonant frequencies of the hydrogen nuclei are spatially
pulse sequence; a long TR results in limited T1- dependent within the gradient. At the middle of gradient,
weighting but in a strong signal. there is the null, where no change in precessional frequency
• An MRI sequence is an ordered combination of RF occurs. A linear increase (or decrease) in precessional fre-
and gradient pulses designed to acquire the data to quency exists with the variation of local magnetic field
form the image. strength away from the null.
Location of spins along the gradient is determined by
their frequency and phase. Often gradient pulses are trape-
zoidal in shape, since it takes time for the gradient field to
8.4.6 MRI Hardware ramp up to the desired amplitude and then to ramp down
again. Gradient amplitudes are commonly measured in mil-
8.4.6.1 Magnet liteslas per meter (mT/m); typical gradient amplitudes are in
The hardware of the MRI scanner consists of four main com- the range 30–50 mT/m. The time required for the waveform
ponents. The first component is the powerful magnet gener- to ramp up and down is known as rise time, which is typi-
ating the strong magnetic field B0 essential to induce the net cally in the range of 200–1000 μs.
magnetization M in the body. Clinical studies require high High-gradient amplitudes are useful to acquire thin slices
image quality. This means that high field strength is pre- or a small field of view (FOV), whereas short rise time per-
ferred. Superconducting magnets capable of producing at mits to select short echo time, echo-spacing, and repetition
least 1.5 Tesla are usually employed and, for some imaging time. Limitations on the maximum amplitude and rise time
studies, 3 Tesla instruments are preferred. must be obeyed to avoid possible peripheral nerve stimula-
To eliminate the electrical resistance, the solenoid coils of tion, pain, and other effects. Gradient switching is responsi-
niobium-titanium (NbTi) are enveloped by liquid helium ble for the typical knocking noise heard during MRI
(−269 °C). When NbTi becomes superconducting, the cur- examinations [31].
rent will circulate indefinitely without any external power
supply [30]. Iron-cored or permanent magnets have a more 8.4.6.3 Radio-Frequency System
patient friendly design, whereas superconducting MRI The third component of an MRI system is the apparatus that
system cryostats are usually shaped as a cylindrical tube that generates the radio frequency necessary to tip the net magne-
surrounds the patient. Although the cryostat bore is 100 cm tization M.
wide, additional coils must be accomodated inside the bore; The radio-frequency (RF) system is made up of a transmit-
hence the diameter is reduced to 60 cm. Recently manufac- ter, coil, and receiver. The transmitter is in charge of generat-
turers have introduced system with 70 cm apertures and ing suitably shaped pulses of current centered at the Larmor
frequency. When these pulses are applied to the coil, the alter- back to the host for image display, processing, archiving,
nating B1 field is produced. The coil will also detect the MR and networking.
signal from the patient. The transmitter must generate RF
pulses with appropriate center frequencies, bandwidths,
amplitudes, and phases in order to excite spins within the
Key Learning Points
desired slices or slabs. Setting the center frequency of the
• The main component of an MR system is the pow-
pulse is crucial to tell the scanner at exactly what frequency
erful magnet generating the strong magnetic field
the nuclei of interest are resonating in the magnet’s isocenter.
B0 necessary to induce the net magnetization M in
Setting the center frequency also allows identification
the body.
and tuning on a nucleus of interest. The bandwidth, or the
• Gradients are a set of additional coils of wire located
range of frequencies within the pulse, controls the thickness
inside the magnetic bore, designed to create linear
of the excited slice. The shape and duration of the RF pulse
magnetic gradients in the three orthogonal direc-
envelope determines the bandwidth. The amplitude of the
tions x, y, z.
RF pulse controls how much the magnetization is flipped by
• The radio-frequency (RF) system is made up of a
the pulse, whilst the phase controls along which axis the
transmitter, a coil, and a receiver; the transmitter
magnetization is flipped (in the rotating frame of reference).
generates suitably shaped pulses of current centered
The main transmitting coil is usually the body coil, which
at the Larmor frequency.
surrounds the entire patient. This is usually built into the
• MRI systems have a host computer employed to
scanner bore and is not generally visible. Since this coil is
adjust the scan parameters.
large it has a very uniform transmission field, but this also
means that it is not particularly sensitive if used as a receiver
coil” [32].
A receiver coil (or simply “coil”) has the function of max- 8.4.7 MR Safety
imizing signal detection, while minimizing the noise. The
tissues of the patients are the principal source of noise. A This section briefly describes the potential bio-effects and
method to reduce the noise and maximize the SNR is mini- risks of MRI environment for patients and personnel.
mizing the coil dimensions, i.e., the coil’s volume should be
filled as much as possible by the sample; hence, the large 8.4.7.1 Projectile Risk
number of anatomically optimized receiver coils available The static magnetic field B0 of an MRI scanner attracts fer-
(e.g., head, spine, shoulder, breast, knee). The signal from romagnetic objects and accelerates them toward the center of
each individual coil element is amplified and then digitized the bore of the MRI scanner. Potential dangerous projectiles
in preparation for reconstruction. that can be torqued or, in the worst case, accelerated by B0
Two types of receiver coil are currently exploited: volume are objects such as coins, hairpins, steel oxygen tanks, or
and surface. Volume coils completely surround the anatomy scissors. Every facility needs to warn personnel about the
of interest and are often combined transmit/receive coils. potential hazard and risks present in an MRI environment.
Surface coils are generally receiver only and, as the name Insufficient MRI safety training of ancillary medical person-
suggests, are good for detecting signal near the surface of the nel has led to fatal accidents [33].
patient. All patients undergoing MRI exams must be screened
for metal, medication patches, tattoos, body piercings,
8.4.6.4 Computer Systems and any electrically, magnetically, or mechanically acti-
Finally, the fourth component consists of a number of com- vated devices. Specifically trained personnel must per-
puters. The multitasking nature of MRI makes impractical to form the screening. The screening may be carried out by
control the many processes from the host computer, so each means of questionnaires, and it is fundamental to detect
subsystems will have dedicated microprocessors that down- any implanted or external metallic foreign bodies.
load commands from the host. Magnetic field strength compatibility of implanted devices
Typically, MRI systems have a host computer on which must be checked, possibly using device identification
the operator adjusts the scan parameters. These parame- card, medical record, manufacturers’ websites, and
ters are converted into commands and then are transferred MRIsaftey.com [34].
to the pulse programmer (PP) (another microprocessor
system) that controls the hardware. The PP manages the 8.4.7.2 Tissue Heating from RF
timing of RF, gradients, and data acquisition in such a The main biological effects related to the exposure to RF
way that they are all properly synchronized. As the data fields are due to the heat generated in the body during the
acquisition is completed, a separate computer system, interaction with the electric field component of RF.
known as the array processor, implements the image The specific absorption rate (SAR) is commonly used
reconstruction. The reconstructed images are then passed to describe the potential for heating of the patient’s tis-
sues; this parameter is a measure of the power absorbed in ciples, techniques, and anatomy. Radiographics. 2003;23(Suppl
1):S111–25.
tissue per unit mass. SAR is proportional to the product of 13. Panetta D. Advances in X-rays detectors for clinical and preclinical
the conductivity and the square of the electric field. The Computed Tomography. Nucl Instrum Methods Phys Res Sect A.
FDA and IEC limit SAR by anatomical site based on the 2016;809:2–12.
potential effects of heating. In most countries standard 14. Mayo-Smith WW. How I do it: managing radiation dose in

CT. Radiology. 2014;273:657–72.
MRI systems are limited to a maximum SAR of 4 W/kg 15. Goldman LW. Principles of CT: multislice CT. J Nucl Med Technol.
[35]. 2008;36:57–68.
16. Geyer LL. State of the art: iterative CT reconstruction techniques.
Radiology. 2015;276:339–57.
17. Beister M, Kolditz D, Kalender WA. Iterative reconstruction meth-
Key Learning Points ods in X-rays CT. Phys Med. 2012;28:94e108.
18. American Association of Physicists in Medicine. Report No. 096 –
• The static magnetic field B0 of an MRI scanner
the measurement, reporting, and management of radiation dose in
attracts ferromagnetic objects and accelerates them CT. College Park, MD: One Physics Ellipse; 2008. p. 20740–3846.
toward the center of the bore of the scanner. 19. Brenner DJ. Effective dose: a flawed concept that could and should
• Patients undergoing MRI scans must be screened be replaced. Br J Radiol. 2008;81:521–3.
20. Huettel SA, Song AW, McCarthy G. Functional magnetic reso-
for metal, medication patches, tattoos, body pierc-
nance imaging. Sunderland, MA: Sinauer Associates; 2008. p. 49.
ings, and any electrically, magnetically, or mechan- 21. Liang Z-P, Lauterbur PC. Principles of magnetic resonance imag-
ically activated devices. ing – a signal processing perspective. In: Huettel SA, Song AW,
• The main biological effects related to the exposure McCarthy G, editors. Functional magnetic resonance imaging.
Sunderland, MA: Sinauer Associates; 2008. p. 61.
to RF fields are due to the heat generated in the
22. Gibby WA. Basic principles of magnetic resonance imaging.

body during the interaction with the electric field Neurosurg Clin N Am. 2005;16:1–64.
component of RFs. 23. Hanson LG, Groth T (Trans.). Introduction to magnetic resonance
imaging techniques. Lyngby: Technical University of Denmark;
2009.
24. Taylor JR, Zafiratos CD, Dubson MA. Modern physics for scien-
References tist and engineers. Boston, MA: Addison-Wesley; 2003. p. 301–3..
ISBN-13: 978-0138057152; ISBN-10: 013805715X.
25.
Hanson LG. Is quantum mechanics necessary for under-
1. Hounsfield GN. Computed medical imaging. Med Phys.
standing magnetic resonance? Concepts Mag Reson Part A.
1980;7:283–90.
2008;32A(5):329–40.
2. Lell MM, et al. Evolution in computed tomography, the battle of
26. van Geuns RJM, et al. Basic principles of magnetic resonance
speed and dose. Invest Radiol. 2015;50(9):629–44.
imaging. Prog Cardiovasc Dis. 1999;42(2):149–56.
3. Shefer E, et al. State of the art of CT detectors and sources: a litera-
27. Grover VPB, Tognarelli JM, Crossey MME, Cox IJ, Taylor-

ture review. Curr Radiol Rep. 2013;1:76–91.
Robinson SD, McPhail MJW. Magnetic resonance imaging: prin-
4. Mahesh M. MDCT physics: the basics: technology, image qual-
ciples and techniques: lessons for clinicians. J Clin Exp Hepatol.
ity and radiation dose. Philadelphia, PA: Lippincott Williams &
2015;5(3):246–55.
Wilkins; 2012.
28. Hodgson RJ. The basic science of MRI. Orthopaed Trauma.

5. Mikla VI. Medical imaging technology. Amsterdam: Elsevier;
2011;25(2):119–30.
2013.
29. Bitar R, et al. MR pulse sequences: what every radiologist wants to
6. Lecoq P. Development of new scintillators for medical applica-
know but is afraid to ask. Radiographics. 2006;26:513–37.
tions. Nuclear Instruments and Methods in Physics Research.
30. Kagawa T, et al. Basic principles of magnetic resonance imag-
2016;809:130–9.
ing for beginner oral and maxillofacial radiologists. Oral Radiol.
7. Duan X. Electronic noise in CT detectors: impact on image noise
2017;33:92–100.
and artifacts. AJR Am J Roentegenol. 2013;201:W626–32.
31. Graves MJ, Zhu C. Basic principles of magnetic resonance imag-
8. Vogtmeier G, et al. Two-dimensional anti-scatter grids for com-
ing. In: Trivedi R, Saba L, Suri J, editors. 3D imaging technologies
puted tomography detectors. Proc. SPIE 6913, Medical Imaging
in atherosclerosis. Boston, MA: Springer; 2015.
2008: Physics of Medical Imaging, 691359, 2008.
32. McRobbie D, Moore E, Graves M, Prince M. Let’s talk techni-
9. Melnyk R, Boudry J, Liu X, Adamak M. Anti-scatter grid evalua-
cal: MR equipment. In: MRI from picture to proton. Cambridge:
tion for wide-cone CT. Proc. SPIE 9033, Medical Imaging 2014:
Cambridge University Press; 2006. p. 167–91.
Physics of Medical Imaging, 90332P, 2014.
33. Sammet S. Magnetic resonance safety. Abdom Radiol (New York).
10. Hu H. Multi-slice helical CT: scan and reconstruction. Med Phys.
2016;41(3):444–51.
1999;26(1):5.
34. Weidman EK, Dean KE, Rivera W, Loftus ML, Stokes TW, Min
11. Ulzheimer S, Flohr T. Multislice CT: current technology and future
RJ. MRI safety: a report of current practice and advancements
developments. In: Reiser M, Becker C, Nikolaou K, Glazer G, edi-
in patient preparation and screening. Clin Imaging. 2015;39(6):
tors. Multislice CT. Medical radiology. Berlin: Springer; 2009.
935–7.
12. Pannu HK, Flohr TG, Corl FM, Fishman EK. Current concepts in
35. van Osch MJP, Webb AG. Safety of ultra-high field MRI: what are
multi-detector row CT evaluation of the coronary arteries: prin-
the specific risks? Curr Radiol Rep. 2014;2:61.
PET/CT and PET/MR Tomographs:
Image Acquisition and Processing 9
Nicola Belcari, Ronald Boellaard, and Matteo Morrocchi
Contents
9.1 P hysical Principles of PET 200
9.1.1 Positron Emission 200
9.1.2 Annihilation of the Positron 200
9.2 adiation Detectors for PET and PET/MR
R 201
9.2.1 Scintillation Detectors 201
9.2.2 Photomultiplier Tube 201
9.2.3 The Block Detector 202
9.2.4 Solid-State Photodetectors 202
9.3 he PET System
T 203
9.3.1 Coincidence Detection 203
9.3.2 Data Acquisition System 203
9.3.3 Geometry of a PET System 204
9.3.4 From 2D to 3D PET 204
9.3.5 Spatial Resolution Issues: Physical Limitations and Technological Aspects 205
9.3.6 Noise in PET Events 206
9.4 I mage Reconstruction and Processing with PET/CT and PET/MR 208
9.4.1 Data Representation 208
9.4.2 Data Correction: Normalization, Random, Scatter, and Attenuation Corrections 208
9.4.3 Time-of-Flight PET 210
9.5 ET/CT and PET/MR Clinical Scanners: State of the Art
P 211
9.5.1 Hybrid Imaging 211
9.5.2 PET/CT Instrumentation 211
9.5.3 PET/MR Hybrid Systems 213
9.6 ET/CT and PET/MR Quantification
P 214
9.6.1 Technical Aspects Concerning in Particular Attenuation Correction 214
9.6.2 Factors Affecting PET/CT and PET/MR Quantification 214
9.6.3 Methods and Metrics for PET/CT and PET/MR Quantification 215
References 216
Learning Objectives • To acquire the knowledge of the working principles of the

• To understand the physical principles of positron emis- PET system and to understand its limitations in terms of
sion tomography. spatial resolution and noise.
• To build up the competence to describe the main compo- • To be able to understand the steps of normalization and
nents of radiation detectors used in PET. correction of PET data and to introduce the concept of
PET image reconstruction.
N. Belcari (*) · M. Morrocchi • To become acquainted with the current state of the art in
Department of Physics “E. Fermi”, University of Pisa and INFN, PET/CT and PET/MR imaging and to understand advan-
Pisa, Italy
e-mail: nicola.belcari@unipi.it tages and limitations.
• To understand the concept of quantitative imaging and
R. Boellaard
Department of Radiology and Nuclear Medicine, VU University describe the factors affecting quantification of PET/CT
Medical Center, Amsterdam, The Netherlands and PET/MR images.

200 N. Belcari et al.
9.1 Physical Principles of PET whereas the distance between the emission point and the
position where thermal equilibrium is reached is called posi-
Positron emission tomography or PET is a nuclear medicine tron range. The range of the positron depends on its energy
imaging technique able to measure, in vivo, the local concen- and the density and Z of the medium.
tration of a tracer labeled with a positron (β+)-emitting radio- When the positron-electron annihilation occurs, two high-
isotope (radiotracer or radiopharmaceutical). energy photons that are physically identical to the γ-rays are
Before acquiring a PET scan, the radiotracer is injected emitted by single-photon-emitting radionuclides. Since these
into the patient and spreads physiologically within the body. high-energy photons are not generated in the atomic nucleus
The radiotracer concentration, which provides an insight of but in the surrounding electron cloud outside the nucleus, they
the physiology and/or pathology of the patient, is propor- cannot formally denominated as γ-rays, although they share
tional to the radioisotope activity distribution ρ(x,y,z). with γ-rays all the properties of interaction with matter; for the
The emitted positron annihilates with an electron in tis- sake of simplicity, we will occasionally use in this chapter the
sue, thus producing two almost back-to-back 511-keV pho- term γ-rays also in the case of the high-energy photons upon
tons. These two photons are detected in time coincidence by which PET imaging is based. The high-energy photons gener-
using opposing pairs of detectors, typically arranged in a ated by the annihilation event are emitted at about 180° (back-
ring surrounding the patient. to-back), and their energy is always equal to 511 keV, because
of the Einstein’s mass-energy equivalence law:
9.1.1 Positron Emission

Eg = ( me c 2 + mb c 2 ) / 2 = ( 511keV + 511keV ) / 2 = 511keV

The positron-emitting radioisotopes are atoms whose nuclei
have an excess of protons with respect to neutrons and decay to where me and mβ are the electron mass and positron mass,
a stable configuration through β+ decay. The β+ decay of a respectively.
nucleus ZX with atomic number Z can be described as follows: As already mentioned, the positron annihilation occurs
when the positron has reached a small enough kinetic energy
X® Y * + b+ + n e (about 1/40 eV, thermal equilibrium). If we consider the
Z Z -1
annihilation in the center of mass of the electron-positron
The β+ particle is called positron and is the antiparticle of system, the two photons are exactly collinear, i.e., emitted at
the electron. This means that the electron and the positron 180°. Since in the laboratory reference frame the center of
have the same mass (melectron = mpositron = 9.109 × 10−31 kg) but mass is not at rest, the collinearity of the two photons will be
opposite charges (qelectron = −qpositron = −1.602 × 10−19 C). The lost due to the angle transformation from one reference
positron is then a positively charged particle. In some cases, frame to the other, the so-called Lorentz boost. This non-
the daughter nucleus Z−1Y can be in an excited state Z−1Y∗ with collinearity results in a Gaussian dispersion centered at 180°
a subsequent γ emission to reach the ground state. Because of a quantity Δθ. When a positron annihilates in water, the
of the presence of the neutrino (νe, a neutral massless parti- non-collinearity deviation Δθ = 0.5° FWHM.
cle), the β+ decay is a three-particle decay. The released The objective of a PET scan is to measure the activity
energy is almost shared between the lightest particles, the β+ distribution ρ(x,y,z) of a β+-emitting radioisotope. Thanks to
and the νe, because of the much larger mass of the recoil the nearly collinear emission of the γ-ray pair from the anni-
nucleus that causes its kinetic energy to be negligible. Hence, hilation of the positron with an electron, it is possible to
the positron exhibits a continuous energy spectrum ranging define the line L along which the annihilation occurred. L is
from zero to the maximum energy Emax released in the decay. usually called line of flight or LOF.
This value depends on the isotope and varies between few The activity distribution ρ(x,y,z) is measured in terms of pro-
hundreds of kiloelectronvolt to few megaelectronvolt. jections (Nγ–γ) along lines L using the line integral operator:
N g - g = k òr ( x,y,z ) dL
L
9.1.2 Annihilation of the Positron
This is an ideal model that assumes zero positron range,
The emitted positrons lose their energy mostly through mul- no deviation from collinearity, and an ideal behavior of the
tiple Coulomb interactions with electrons that are present in detector. In the practical situation, the lines L are defined by
the biological tissue. For the equivalence of the positron and all the possible lines of response connecting a detector ele-
electron masses, the positron can experience large angular ment i to a detector element j. Then, the line integral operator
deviations in each interaction. The positron travel is then can be written as:
characterized by a zigzag path. When the positron reaches
thermal equilibrium with the medium, it annihilates with an N ij = k ò r ( x,y,z ) dL
electron. The total path of the positron is called path length, LOR ij
9 PET/CT and PET/MR Tomographs: Image Acquisition and Processing 201
Thus, the problem of the PET measurement is reduced to and releases all or part of its energy inside it, the scintillator
the detection of pairs of 511-keV γ-rays. emits visible light isotropically. The quantity of emitted light
photons is usually proportional to the released energy. The
proportionality constant is called light yield and is measured
Key Learning Points in photons/MeV. The emission spectrum of the emitted light
• Radioisotopes used in positron emission tomogra- is characteristic for each scintillator but always in the visible
phy are subject to β+ decay. or near-visible range. The emitted low-energy photons can
• Emission of the positron and the subsequent annihi- be easily detected by standard photodetectors (such as a pho-
lation with an electron produces a pair of nearly tomultiplier tube) that convert the visible light signal into an
collinear photons of 511 keV each. electric current pulse, whose intensity remains proportional
• The number of detected γ-ray pairs can be corre- to the energy released by the original γ-ray.
lated to distribution of the activity inside the object Scintillators for PET, and nuclear medicine in general, are
of interest through a mathematical relationship. typically solid inorganic compounds with a high stopping
power for μ511 keV (linear attenuation coefficient in the range
of 0.3–1 cm−1 at 511 keV).
The choice between the different types of scintillators
9.2 adiation Detectors for PET
R stays in the optimal compromise among different features:
and PET/MR detection efficiency (expressed by the linear attenuation
coefficient at 511 keV and related to the crystal density and
The fundamental step in a PET measurement is to obtain the effective atomic number of the material), conversion effi-
spatial coordinates of the line of response where the count, ciency (expressed by the light yield), the output spectrum
corresponding to a positron emission and the following anni- (usually indicated by the peak wavelength), and the time
hilation, is detected. over which the light is emitted (the light is usually emitted
This can be achieved by measuring the coordinates P(x, y, with a fast light flash followed by an exponentially decaying
z) of the first interaction for both γ-rays in opposite detectors. intensity with a decay time characteristic for each scintilla-
The information that the PET detector must be able to pro- tor, from tens to hundreds of nanoseconds). For measuring
vide are the position of the first interaction of a 511-keV the original energy of the incident photon, it is essential that
γ-ray in the detector itself, the energy released in the interac- all the energy is released inside the scintillator, possibly with
tion or in the series of interactions, and the arrival time (at a single photoelectric interaction to preserve the information
least for making the coincidence measurement possible). on the point of the first interaction. For this reason, the rela-
Various detector technologies have been developed to pro- tive probability between photoelectric and total interaction in
vide all of this information. As of today, the most widely the scintillating material (photofraction) is also an essential
used solution is a scintillating material coupled to a feature when choosing a PET scintillator. Table 9.1 reports
photodetector. the most widely used scintillators for PET.
9.2.1 Scintillation Detectors 9.2.2 Photomultiplier Tube
Scintillation detectors consist of a dense crystalline material A photomultiplier tube (PMT) is a device capable of convert-
that acts as an interacting medium. When a photon interacts ing visible or near-visible light into an electric signal. A
Table 9.1 Physical properties of scintillating materials used in PET (adapted from [1, 2])
μ511keV
Material Density (g/cm3) Light yield (ph/MeV) Decay time (ns) (cm−1) Photofraction at 511 keV
Sodium iodide (NaI/Tl) 3.67 41,000 230 0.34 17%
Bismuth germanate (BGO) 7.13 8200 300 0.96 40%
Lutetium oxyorthosilicate (LSO/Ce) 7.40 30,000 40 0.87 32%
Lutetium yttrium oxyorthosilicate (LYSO/ 7.10 32,000 40 0.82 30%
Ce)
Gadolinium oxyorthosilicate (GSO/Ce) 6.71 8000 60 0.70 25%
Yttrium aluminum perovskite (YAP/Ce) 5.37 21,000 27 0.46 4.2%
Lutetium aluminum perovskite (LuAP/Ce) 8.3 12,000 18 0.95 30%
Barium fluoride (BaF2) 4.89 1400 (fast) 9500 0.6 (fast) 0.43
(slow) 630 (slow)
Lanthanum bromide (LaBr3/Ce) 5.08 63,000 16 0.47 15%
PMT is typically made of a vacuum glass envelope contain-

C
ing a series of electrodes called dynodes. The inner surface
of the glass entrance window is coated with a thin layer of a Y A
material called photocathode that releases electrons when hit
X
by a light photon (photoelectric emission). The photocathode D
is kept at a negative potential so that electrons are accelerated B
away from it. The probability of an electron to be emitted for
each light photon reaching it is called quantum efficiency (or
QE) and varies with the wavelength of the incoming light
photons and can be about 15–25% at peak. The electrons are Fig. 9.1 Diagrammatic representation of a block detector. A block of
then accelerated and multiplied by a series of dynodes that scintillator is subdivided by cuts of different depths in 4 × 8 rectangular
are held to a higher potential than the previous one by a volt- elements. The block is read out by a 2 × 2 matrix of photomultipliers
age divider resistor chain. The typical gain of a PMT is of the
order of 106. The whole multiplication step occurs in a very g ranularity of the pixel size. However, the capability to dis-
short time, making it possible for PMTs to provide sub-ns criminate neighbor clusters prevents to use very thin pixels.
electron time resolution. For whole-body PET scanners, a lateral side of 4–5 mm is
usually used. The main advantage of this type of detector is
that the cuts in the scintillator allow decoding of a large num-
9.2.3 The Block Detector ber of pixels by using only four photodetectors.
The standard configuration of a PET detector is represented

by the block detector, as developed by Casey and Nutt in 9.2.4 Solid-State Photodetectors
1986 [3]. The detector is composed of a scintillator having
lateral dimensions of few centimeters and approximately two Solid-state photodetectors take advantages of the properties
centimeters thickness. The scintillator is coupled to a 2 × 2 of a silicon p-n junction to detect optical photons generated
matrix of photomultiplier tubes (PMTs), as shown in Fig. 9.1. in the scintillation; depending on the arrangement of the lay-
To correlate the light collected by each PMT to the interac- ers that compose the junction, on its level of doping, and on
tion position of the impinging photons, a set of cuts of differ- the applied bias voltage, an avalanche can be triggered by the
ent depth are engraved in the scintillator to subdivide the incoming photon, thus generating an amplified signal with
detector in “pixels” of fixed dimension. These cuts are per- different levels of gain (avalanche photodiodes—APD), or a
formed regularly and orthogonally to the scintillator surface Geiger-mode avalanche (single-photon avalanche diode—
(to provide pixels of fixed dimension) and are covered with SPAD), capable to discriminate each single photon that is
reflective material. In addition, to increase the capability to detected by the device. A matrix of SPAD cells can then be
discriminate the lateral pixels of the block, the cuts are created to obtain a single-photon counter named silicon pho-
deeper closer to the edges of the detector. The X and Y coor- tomultiplier (SiPM). Several features of SiPMs make them
dinates of the photon interaction position can be determined particularly attractive for PET applications; in fact, solid-
by using the following formulas: state devices are insensitive to magnetic field (thus being
suitable for hybrid PET/MR systems); they are compact and
( S1 + S2 ) - ( S3 + S4 ) can be easily arranged in matrices (a single detector has a
X =
S1 + S 2 + S3 + S 4 size of few mm2), allowing the identification of thin scintilla-
tors adopted in preclinical PET systems. Furthermore, SiPMs
( S1 + S3 ) - ( S2 + S4 ) are fast and then suitable for the realization of a new genera-
Y= tion of PET systems with correction for the time of flight (or
S1 + S 2 + S3 + S 4
TOF—see further below).
where Si is the signal produced by the ith PMT and the
denominator (the sum of the signals of the four PMTs) is 9.2.4.1 Avalanche Photodiodes
related to the total energy released in the scintillator. The An avalanche photodiode (APD) is a solid-state photodetec-
reconstructed X–Y positions provide a map of clusters, called tor device. By applying a high reverse bias voltage, it is capa-
flood map, where each cluster corresponds to a pixel of the ble to accelerate enough the free electron produced in the
scintillator. Each region of the map is then assigned to a pixel active area of the device to generate other electron-hole (e/h)
of the block in the so-called pixel-identification process, and pairs by ionization while moving toward the cathode. If an
it is stored in a look-up table. Through this calibration, the e/h pair is generated by the interaction of an optical photon,
interaction position can be reconstructed with the same the charge triggers an avalanche that can multiply the
c ollected signal up to a factor of few hundreds. Nevertheless,

the gain of these devices is much smaller than that of com- • The standard detector configuration, called block
mon PMTs; moreover, it requires an external amplification detector, is composed of four photomultipliers cou-
to generate signals that can be easily processed, as well as an pled to a pixelated scintillator matrix.
effective cooling system to prevent signal drifts over time. • The development of new photodetectors based on
semiconductors enables to build PET systems com-
9.2.4.2 Silicon Photomultipliers patible with the MR environment.
In a SPAD the electric field in the junction is high enough to
trigger a self-sustaining avalanche in which both electrons
and holes produce new carriers (Geiger avalanche); the ava- 9.3 The PET System
lanche can be quenched by means of an external resistor con-
nected in series to the junction that, according to the current 9.3.1 Coincidence Detection
generated by the carriers, reduces the voltage on the device,
thus disabling the avalanche. Such device generates a signal As described in the previous sections, through the detection
of a fixed amplitude when a photon is detected and is not of the position of the interaction of the two annihilation
capable to discriminate more than a photon at a time; thus, photons, it is possible to define a line, called line of response
matrices of SPAD are necessary if several photons need to be or LOR. The position of the annihilation event is supposed to
counted. Such device is named SiPM (the same detector is occur along this line. The problem is then to recognize a pair
also referred as Geiger-mode avalanche photodiode, G-APD, of photons as being generated by the same annihilation pro-
or multi-pixel photon counter, MPPC). The SiPM is com- cess. This can be done by exploiting the information pro-
posed of thousands of cells, each cell is composed of a SPAD vided by the time of arrival of a γ-ray into the detector. The
and its quenching resistor, and all the cells are connected in two photons are simultaneously generated and then, neglect-
parallel so that the resulting signal is the sum of the signals ing the delay that may occur due to a difference in distance
generated by each triggered cell. Each SPAD has a size of from the annihilation point to the two detectors and consider-
tens of micrometers, and the surface of the full detector is of ing that they travel at the speed of light, they are simultane-
few mm2; such a high number of SPADs is necessary to ously detected. We call coincidence detection the process of
avoid saturation when SiPMs are coupled to bright scintilla- selecting the annihilation γ-ray pair using the arrival time
tors commonly used in PET. information. The event associated with the occurrence of a
Adoption of the Geiger mechanism ensures a gain on the coincidence detection of two photons is called coincidence
order of 106, comparable to PMTs, and the signal generated event, while in a single event, only one γ-ray is detected.
is very fast: a resolution of 80 ps has been reached with For this reason, the determination of the time when a pho-
single-photon signals, while about 20 ps can be reached irra- ton has struck a detector is a critical step in the PET detection
diating a SiPM with a higher number of photon by means of process. After having retrieved this information, the time of
a fast laser source. all detected events can be compared to determine which ones
The fabrication of sensors on a silicon substrate led also arrived closely enough in time to be identified as an annihila-
to the development of SiPMs composed of a digital part fully tion pair. The ability of a pair of detectors to determine the
integrated in the device equipped with an active quenching time difference in arrival of the annihilation photons is
circuit for each cell, an addition circuit to count the number known as the timing resolution and is typically in the range
of triggers in a defined time window, and time to digital con- from few hundred of picoseconds to few nanoseconds. The
verters to extract the timestamp of the event. These photode- maximum difference in time for a pair of detected photons to
tectors are usually called digital silicon photomultipliers or be identified as a coincidence event is called time window or
dSiPM. In such devices, no analog signal processing is τ. In order to avoid missing coincidence events, the time win-
required, thus improving the compactness of the system and dow is usually larger than two times the time resolution (e.g.,
minimizing the pickup noise in the propagation of the analog 2–3 ns using LSO).
signal.
9.3.2 Data Acquisition System
Key Learning Points In its simplest form, a PET data acquisition system has a
• Annihilation photons can be detected using a scin- two-branch structure: a timing circuitry provides the coinci-
tillator (to convert them into optical photons) and a dence information and enables to acquire data on the involved
photodetector (to generate an electrical signal). detectors, while a data collection circuitry converts position
and energy signals into digital values.
Amplitude
AMP
discriminator
1
DET.
Timing
discriminator
2
Inhibit
τ
Coincidence
Consensus
circuit Acquisition
τ
Inhibit
DET.
Timing
discriminator arc 2 arc 1
Amplitude
AMP discriminator
Fig. 9.2 Diagrammatic representation of the data acquisition electron- Fig. 9.3 Diagrammatic representation of a PET ring. Several LORs
ics of a simplified PET system comprising one detector pair only involving two single detectors are drawn. For example, detector 1 is in
coincidence with all detectors in the arc 1. The full FOV (displayed in
dark gray) is given by the intersections of all arcs
Figure 9.2 illustrates the PET data acquisition for a sim-
plified two-detector system. Along the timing branch, the
PMT signal passes through a fast amplifier (to preserve the line integrals defined by the possible lines of response is col-
time information) and enters into a discriminator that pro- lected (tomographic acquisition), thus sampling the object
duces a digital signal when a γ-ray with enough released along the spatial and angular coordinates.
energy is detected. The width of the digital signal generated Each detector can accept coincidence with any detector
is set to be equal to the time window τ. In order to improve belonging to an opposite arc of detectors (Fig. 9.3), thus
the time resolution, a constant fraction discriminator (CFD) defining a sort of wedge. The intersection of all the wedges
is used. The CFD is a device that produces a timestamp when is a circle centered on the scanner axis that represents the
the signal reaches a fraction of its maximum (instead of a field of view (FOV) of the PET system. Thus, a single ring
fixed amplitude value like in standard time discriminators), PET is able to provide images of slices of the object with an
thus avoiding time measurement systematic errors (called axial extension equal to the detector size along the ring axis.
time walk) when signals of different amplitudes are pro- In order to increase the FOV size along the axial direc-
cessed. The timing pulses are fed into a coincidence circuitry tions, modern PET systems feature multiple rings (multi-ring
that, through a series of logic ports (a simple AND port in geometry) with a typical axial extension of 15–20 cm.
case of a two-detector system), identifies pair of signals hav-
ing a certain overlap in time (time coincidence). In this way,
a coincidence is accepted when the time difference is 9.3.4 From 2D to 3D PET
between −τ and +τ. Hence, the effective time coincidence
window is actually 2τ. Multi-ring PET systems are classified into two categories:
The detection of a coincidence event enables the acquisi- 2D and 3D scanners. In 2D PET, coincidences among detec-
tion of energy and position signals of the involved blocks tors belonging to two different rings are not allowed. This
only, along the other branch. The extension of this concept to simplification makes the image reconstruction process eas-
a multiple detector system can be obtained by implementing ier. In addition, each ring is physically separated from the
a more complex coincidence circuitry. adjacent one with a septum made of a high-Z material, so as
to limit the number of single events reaching a single detec-
tor. The advent of more advanced reconstruction algorithms
9.3.3 Geometry of a PET System and more powerful hardware made 3D PET scanners, where
inter-ring coincidences are also recorded, possible. The
PET detectors are usually positioned around the object under major advantage of the 3D modality versus 2D PET is the
study in the so-called ring geometry. In this way, a full set of higher system sensitivity with a maximum increase of the
number of rings. As of today, all clinical PET systems fea- to the geometry of a PET system that does not sample all the
ture the 3D modality as the default setting. lines of flight in the same way.
Degradation of the spatial resolution is essentially due to
the uncertainty in the determination of the line of flight
9.3.5 Spatial Resolution Issues: Physical (LOF) that depends upon factors related either to the detector
Limitations and Technological Aspects geometry or to the physics of the detection process. When a
coincidence is detected, a pair of crystals defines the
The physics of the β+ decay and the consequent positron- LOR. Because of the finite size of the detectors, the LOR is
electron annihilation described in Sects. 9.1.1 and 9.1.2 rep- not actually a line but a region where the annihilation has a
resent the first limitations in spatial resolution of the PET certain non-zero probability to have occurred. This probabil-
technique. In fact, the average range (in water or other tissue- ity is described by a coincidence response function. For a
equivalent materials) of the positrons emitted from most pair of crystal elements, the FWHM of the coincidence
PET radionuclides is about 1–2 mm. Besides the positron response function in the midplane between detectors is equal
range, non-collinearity also affects the spatial resolution of a to half the size of the crystal (d) along the same direction and
PET system (Fig. 9.4). Assuming that the PET measurement worsens by moving closer to one of the two detectors. Thus,
is performed with a ring of detectors of diameter D, the effect the contribution (in terms of FWHM) to the spatial resolu-
on the PET spatial resolution can be expressed by the empiri- tion due to the finite crystal size is minimum at the center of
cal formula [4]: the FOV and equal to d/2. When crystals in a matrix are sepa-
rated by a reflective material of non-negligible thickness, the
FWHM = 0.0022 ´ D value d is actually the crystal pitch, i.e., the distance between
The best achievable spatial resolution is also limited by the centers of adjacent crystal elements. As the crystals
other factors related to the detection process and to the tech- defining the LOR are not aligned, the finite thickness of the
nology in use. In general, the spatial resolution of a PET sys- crystals comes into play, as typically no information on the
tem is not constant along the whole FOV. This fact is related depth of interaction (DOI) is provided. This fact introduces a
further contribution to the FWHM that is usually indicated
A with the letter p and is called parallax error (Fig. 9.5).
C
A C
C
I
DO
1 2
1 2
B D Parallax
B error
D
Fig. 9.5 Representation of the parallax error in the vicinity (case 1)
Fig. 9.4 Diagrammatic representation of the effects of positron range and far from (case 2) the center of the FOV (indicated by the cross in the
(case 1) and non-collinearity (case 2) on spatial resolution. In case 1, figure). In case 1 the two high-energy photons reach the detectors along
the positron moves away from the emission point before annihilating, a direction parallel to crystals long axis (0° interaction): parallax error
and the emitted high-energy photons are then detected by two detectors is minimum in this case, irrespective of the depth of the interactions.
A and B. The derived LOR passes from the annihilation point but not Thus, the LOR passes very close to the annihilation point. In case 2, the
from the point where the radioactive nucleus was actually present. In high-energy photons reach the detectors with an angle significantly
case 2, it is possible to appreciate how the detection of two high-energy larger than 0°: a larger parallax error occurs (the LOR passes at a certain
photons not emitted at exactly 180° generates a LOR (defined by detec- distance from the annihilation point), especially when the interaction
tors C and D) that does not cross the annihilation point occurs deep in the crystal
In PET systems with ring geometry, the parallax error has The detection efficiency D refers to the efficiency with which
a significant effect along the radial direction, resulting in a a detector converts emissions from the radiation source into use-
sort of radial elongation of the images. The radial resolution ful data and can be defined as D = R/A, where R is the counting
worsens by moving far from the center of the FOV. In fact, rate and A is the source activity. High detection efficiency is
the contribution to the FWHM of the PSF along the radial desirable to obtain an image with minimum statistical noise
direction can be approximated by the following equation: with a minimum amount of radioactivity (dose reduction).
Counting statistics is only one of the factors in PET deter-
r
p =a mining an increase of noise in the image (or, more correctly,
(r 2
+ R2 ) an increase of image roughness) that are related to both the
physics of the γ-ray interaction with matter and to the detec-
where r is the radial position where the PSF is derived, R is tor technology. A way to observe this effect is to evaluate the
the radius of the PET ring, and α is a term that depends on uniformity of a uniform radiation source in terms of standard
the material and thickness of the scintillating crystals. For deviation of the measured value.
example, α = 12.5 for a 30-mm-thick BGO crystal [5]. Besides the already discussed statistical limitations, uni-
In addition, some errors may also occur in associating the formity is affected by the presence of LORs not appropri-
interaction point to a certain crystal element. This process is ately recorded, i.e., not generated by a true count. A true
called crystal (or pixel) identification and can be performed count is a coincidence event with both annihilation photons
in different ways. The most common one is to calculate the being detected without experiencing any other interaction
centroid of the light spot emerging from the crystal and to along their path. Under these conditions, the LOR passes
compare the calculated position to a pre-calculated look-up through the annihilation point. Actually, not all the recorded
table where positions are associated to crystal elements. The LORs are generated by true counts. Figure 9.6 shows four
source of this error is twofold: there could be a possible error
in the crystal identification process itself, and there is the
possibility of multiple interactions in the scintillator to occur.
When the multiple interaction occurs in the crystal involving
more than one crystal element, the event is called an inter-
crystal scatter (ICS). The contributions of both crystal iden-
tification and ICS effects are included in the so-called coding
error term (b).
The best achievable spatial resolution in PET can be sum-
marized with the following formula that takes into account
both physical effects and the technological limitations:
( d / 2) + b 2 + ( 0.0022 D ) + r 2 + p 2
2
FWHM = 1.25
where 1.25 is a term related to the image reconstruction pro-

cess and is estimated assuming an analytical reconstruction
algorithm such as the filtered back-projection. The other fac-
tors are the detector size (d), the coding error (b), the non-
collinearity term (where D is the scanner diameter), the
positron range (r), and the parallax effect (b).
Modern clinical PET systems feature a spatial resolution
of about 4 mm FWHM. Dominating factors are non-
collinearity and pixel size, while the positron range factor r Fig. 9.6 Diagrammatic representation of true (T), scattered (S), and
may become important for high-energy positron emitters random (R) events. High-energy photons produced in point 1 reach
such as 68Ga. detectors A and B with no interactions. The LOR defined by detectors
A and B crosses the annihilation point. This type of event is called a true
count. Photons produced in point 2 reach detectors C and D, but one of
the two experiences a Compton scattering in the object before being
9.3.6 Noise in PET Events detected. The LOR defined in this way in incorrect. This type of event
is called a scattered count. A random count is recorded when two (and
only two) high-energy photons emitted by two independent annihilation
Noise characteristics have important implications for image events are simultaneously detected, e.g., by detectors E and F in the
quantification and detection performance in PET imaging, figure. Also in this case, the recorded LOR is not related to any
especially in high-resolution scanners. radioactive decay and contributes to image degradation
possible types of coincidence events that can be recorded square of the a ctivity, and the image quality is reduced as
during a PET scan. the activity increases. In fact, the random count rate can
The event originating in location 1 corresponds to a true be expressed as:
count (T) between detectors i and j, thus generating a count
Rij = Ci ´ C j ´ 2t
in the LORij. Instead, at least one of the two γ-rays emitted
from position 2 experiences a Compton scattering along its where Ci and Cj are the singles count rate on detector i and j
path, and the corresponding count in the LORkl considers an and τ is the coincidence time window. Hence, a narrow time
annihilation that does not occur along the same line of window helps reducing the effect of random counts.
response. Such type of event is called scattered count (S). A The noise equivalent count rate (NECR) is a figure of
third type of event is called a random count (R). It is gener- merit that quantifies the amount of background and statisti-
ated when two photons are emitted from two different points cal noise characteristic of a given PET scanner, thus evaluat-
(e.g., points 3 and 4 in the picture), and they are recorded, in ing the effect of the presence of scattered and random counts.
time coincidence, in detectors defining a geometrically The formulation of NECR is as follows:
acceptable LOR. All types of T, S, and R counts are generi-
cally called prompt (P) counts, but only true counts contain RT 2
RNEC =
useful information for the image reconstruction. An addi- RTOT
tional type of event is a multiple count, where more than two
γ-rays are recorded in time coincidence. Because of the where RTOT is the sum of true (RT), random (RR), and scatter
ambiguity on deciding which are the two photons actually (RS) count rates:
generated in a single annihilation, this type of event must be
RTOT = RT + RS + kRR
discarded. The relative weight of the effect of scattered
events on the final image is indicated by the scatter fraction where k is a factor that takes into account the method used
(SF), defined as: for estimating random counts, usually k > 1, while k = 1 for
noiseless random counts. Being NECR proportional to the
RS signal-to-noise (SNR) ratio in the final reconstructed images,
SF =
RT + RS it constitutes a good parameter to compare the performances
of different PET scanners and to estimate the optimal activity
The scatter fraction can be reduced by discarding some of to be injected to the patient.
the scatter events. A scatter event can be rejected using the Figure 9.7 shows examples of true, scatter, random, and
energy information, i.e., only photons that release 511 keV NECR curves as a function of the activity concentration.
in the crystal are accepted as a sign of no interaction of the
γ-ray before reaching the detector. The ability of the detector
to determine the energy of the photon is known as the energy
resolution. Typical values of energy resolution for LSO read- Key Learning Points
out by PMTs are 15–20% at 511 keV. Due to the finite value • Pairs of high-energy photons are identified as being
of the detector energy resolution, it is not possible to accept originated from the same annihilation event on the
events with an energy release of exactly 511 keV, but a wider basis of time coincidence detection.
energy window (EW), typically 450–650 keV in clinical sys- • A typical data acquisition system for PET features a
tems, should be set. time measurement branch and an energy measure-
However, when setting an energy window around the full ment branch.
energy peak, a fraction of scatter events is still included. • The typical detector arrangement in PET is a ring
Typical scatter fraction values in PET may range between geometry so as to collect tomographic information.
15% and 20% up to 50% in obese patients. • In addition to physical limitations due to the posi-
For this reason, a number of methods to correct the effect tron range and the non-collinearity of emitted
of the scattered events in the degradation of image quality γ-rays, spatial resolution is limited by other techno-
have been developed and are applied before or during the logical factors, such as pixel size, coding, and paral-
reconstruction process. They are usually classified with the lax error.
term “scatter correction” and used to restore the object con- • Image noise in PET is due to count statistics but
trast in the final image. also to the detection of mispositioned lines of
Random events also contribute to image degradation responses.
by generating a more uniform background than scatter • The major sources of error are scattered counts and
counts. While the scatter fraction is almost constant with random counts.
the activity, the random count rate increases with the
Fig. 9.7 Example of the plots 400

True
of RT, RS, RR, RTOT, and RNEC
for a clinical PET scanner Random
using the NEMA NU 2 (2001) Scatter
phantom [6]. (Data courtesy
of IFC-CNR Pisa (Italy). Data NECR (1R)
obtained with the GE 300
NECR (2R)
Discovery SRX PET/CT
system)
Count rate (kcps)

200
100
0
0 20 40 60 80
Activity concentration (kBq/cc)
9.4 I mage Reconstruction and Processing In case respiratory or cardiac gating signals are included
with PET/CT and PET/MR during histogramming, multiple sinograms are generated—
one for each gate or bin. The sinograms can be saved sepa-
9.4.1 Data Representation rately or as a single file containing all sinograms for all
(specified) time intervals and/or respiratory or cardiac gates.
On most current PET systems, coincidence data are ini- In order to save space and processing time, multiple lines of
tially stored in a list-mode file. The list-mode file is a list responses can be summed into a single sinogram bin. The
of all detected coincidences stored sequentially in the number of and which lines of responses are combined into a
order of detection. For each coincidence event, the infor- single bin is typically described by “span” and “mashing.” A
mation on the detector pair that measured the coincidence, detailed description of data representation, histogramming,
the type of detection (prompt or random), and, if applica- and rebinning can be found in [7].
ble, the difference in arrival time of the coincidence pho-
tons (time of flight) is stored. In addition, at pre-set times,
typically every millisecond, a so-called time tag is inserted 9.4.2 D
ata Correction: Normalization,
in the list-mode file. Random, Scatter, and Attenuation
Other additional information may be inserted during the Corrections
PET study as well. For example, when acquisitions are per-
formed using respiratory or cardiac gating, information on Before being ready for the image reconstruction process,
patient motion will be continuously inserted in the list-mode PET data need to be corrected. In fact, the line integral model
file. By adding time tags and motion information into the assumes that the efficiency of the various LORs follows an
list-mode file, the data can be sorted afterward in time frames ideal behavior, while nonuniformities of individual detectors
and/or respiratory or cardiac gates. Sorting of the list-mode may introduce inter-LOR efficiency variations versus the
data is often referred to as histogramming. During histo- theoretical value. In addition, scatter and random events and
gramming the sequential list of coincidences can be sorted the attenuation of γ-rays may corrupt the number of counts in
into integrated time intervals (frames) and saved in one or each LOR, thus causing a wrong evaluation of the activity
more sinograms. The sinogram consists of a large number of along the LOR itself.
bins. These bins indicate the total number of coincidences
measured for each line of response during a time interval. 9.4.2.1 Normalization and Scatter Correction
Bins are sorted in the sinogram as a function of transaxial Data normalization in PET is the process for compensating
position and angular orientation of the lines of response. the nonuniformities of LOR’s efficiencies by multiplying the
counts in each LOR by an appropriate factor. Normalization technique can be more accurate than the indirect method,
values are typically derived by scanning a geometrically uni- even if it is typically characterized by a noisier distribution
form object, such as a cylinder with a uniform activity con- of Rij values.
centration (or using rotating rod sources of 68Ge), mimicking
an annular source. The actual acquired counts per LOR are 9.4.2.3 Attenuation Correction
then compared with the theoretical values, thus obtaining the The linear attenuation coefficient μ in soft tissues for 511-
normalization factors. In order to reduce the effect of the sta- keV photons is about 0.096 cm−1, while it is about 0.17 cm−1
tistical noise that can be introduced in this process, the data in the bone. Hence, the probability for a γ-ray to interact
acquisition must be long enough to collect a large number of (through Compton scattering or photoelectric absorption) in
counts per LOR. In some cases, component-based variance the patient tissues along its path is relatively high. Events
reduction methods are implemented in order to further miti- that are recorded after a Compton interaction are corrected in
gate this effect [8]. the scatter correction process. In fact, scatter correction is
As already discussed, the use of an energy window can typically performed before attenuation correction. In the
reduce the number of scatter events that are included in the remaining cases, counts are simply lost. Taking into account
data sinogram. Nevertheless, the residual scatter events are the γ-ray attenuation along a LORij of this effect, the line
mixed with true counts, thus reducing image quality in the integral should be modified as follows:
reconstructed image, typically resulting in a loss of image
contrast and the accuracy of quantification. To minimize the N ij = k
LOR ij
ò r ( x ,y , z ) d L × P × P i j
effect of scatter counts in the reconstructed image, a proce-

dure called scatter correction is often performed. where Nij is the number of counts recorded along LORij and:
There are several methods implemented by scanner man-
ufacturers for scatter correction including analytical methods ò
- m ( x ) dx ò
- m ( x ) dx
Pi = e ; Pj = e
Li Lj
[9, 10], Monte Carlo simulation techniques [11], multiple

(e.g., dual [12] or triple [13]) energy window methods, and are the probabilities for the two γ-rays to reach detectors i
model-based algorithms [14, 15]. and j, respectively, and Li and Lj are the lines connecting the
annihilation point to detectors i and j (Fig. 9.8). The com-
9.4.2.2 Random Correction bined probability P for both γ-rays to reach the detectors is:
Random counts in the data sinogram can be reduced by
implementing a narrow time window. As for the scatter ò
- m ( x ) dx ò
- m ( x ) dx - ò m ( x ) dx
P = Pi × Pj = e ×e =e
Li Lj Li + L j
counts, residual random counts cannot be distinguished from

true counts. A possible way to correct data for random counts where, assuming L = Li + Lj, Nij can be then written as:
is to perform a statistical estimation of the random counts
with either a direct or an indirect procedure. The estimated ò
- m ( x ) dx
random counts (Rij) in each LOR can be either subtracted

N ij = k ò r ( x ,y , z ) d L × e
LOR ij
L
from prompt events online or stored as a separate sinogram

for later processing. i.e., the probability for a count to be detected depends only
An indirect way to perform the random estimation is to on the line integral of the attenuation coefficient along the
apply to the single count rate recorded in each detector the LOR itself and the quantity:
formula:
1 òm ( x )dx
RTOT = RT + RS + kRR = eL
P
Although this method may be precise, it requires an a
priori estimation of τ that may lead to estimation inaccuracy. is usually called (ACF). For a 40-cm-diameter patient, the
As an alternative, the so-called delayed window technique ACF is about 50.
[16] can be used to directly derive the random count rate. Attenuation correction is necessary not only to perform a
With the use of an additional coincidence processor, time correct estimation of the activity concentration but also to
coincidences between a logic pulse from one detector and a avoid image artifacts that may occur causing incorrect con-
delayed logic pulse from another one can be detected and centration ratios between different regions inside the patient
measured. If the delay is large enough, there is no chance body. Typically, its image borders and/or activity concentra-
that the coincidence comes from an annihilation pair. Hence, tion present in low attenuation tissues (such as lungs) is
the delayed coincidence count rate Rdelayed,ij between each exalted with respect to other regions.
pair of detectors can be considered a good estimation of the ACFs for each LOR can be estimated through a direct or
random count rate Rij in the same pair. The delayed window indirect measurement of the distribution of the attenuation
annihilation actually occurred. A possible way to estimate

the exact position of the annihilation point is to measure the
difference of arrival time of the two photons onto the oppos-
ing elements. This technique is called time-of-flight PET
(TOF-PET). In fact, the difference in arrival time is given by:
dA - dB 2 × DS
DT = TA - TB = =
c c
where the times of arrival of the two photons to detector A
and B are given by TA = dA/c and TB = dB/c, where dA and
dB are the distances from the annihilation point to A and B
detectors, respectively, and c is the speed of light. ΔS is the
displacement of the annihilation point from the center, which
can be then estimated as:
c × DT
DS =
2
The error associated with this measurement is related to
the precision σT in the estimation of ΔT. Thus, the uncer-
tainty on ΔS is:
c ×s T
sS =
2
Fig. 9.8 Diagrammatic representation of the paths of two annihilation
high-energy photons (γ1 and γ2 in this example) crossing the patient’s In order to limit the uncertainty in the position estimation,
body. The two high-energy photons follow the paths L1 and L2, where the PET system must feature a very high time resolution. For
L1 + L2 = L, i.e., the intersection of the LOR with the object. The attenu- example, considering the time resolution of a standard clini-
ation of counts in each LOR is not related to the position of the emis-
sion point P(x,z) but only to the linear attenuation coefficients measured cal PET (about 1–2 ns FWHM), this would provide a spatial
along the LOR resolution of 15–30 cm FWHM in the estimation of ΔS, not
providing as such any useful information. A PET system is
coefficients μ(x, y, z) at 511 keV. A way to directly measure considered TOF-enabled when the time resolution is about
μ(x, y, z) is via the so-called transmission scan [17], i.e., per- 500 ps FWHM, i.e., the precision in the determination of the
forming an acquisition by rotating a positron-emitting source position of the annihilation point is ≈7.5 cm FWHM.
(typically 68Ge) around the object, thus creating a sort of Although this does not allow a sufficient precision to
transmission scan and blank scan sinograms. By comparing obtain a significant direct gain in spatial resolution of the
the counts recorded along a certain line in the presence of the images, it has been proved that this information improves the
object (transmission scan) to those obtained without the signal-to-noise ratio (SNR); in fact:
object (blank scan), attenuation data (i.e., the ACFs) along all
the LORs are measured. Although the low technological D
complexity and the accuracy of the results, there are several SNR TOF = × SNR no-TOF
c × DT
limitations: long scan times, noisy attenuation sinograms
due to low statistics and the needs for replacement of the This estimate is approximate, but it gives a clear indica-
transmission source every 12 or 18 months. tion of the importance of timing resolution [18]. More opti-
In the last years, with the advent of hybrid PET/CT systems, mistic estimations of the gain in SNR include the beneficial
the transmission scan technique is not necessary anymore, and role of TOF in rejection of random counts [19]. TOF-PET
the attenuation correction is mostly performed on the basis of image reconstruction is also typically faster and its conver-
the estimation of the linear attenuation coefficients from the CT gence is more stable.
image (CT-based attenuation correction, or CT-AC).
9.4.4 Image Reconstruction

9.4.3 Time-of-Flight PET
Image reconstruction is the process of generating a PET
When a standard PET data acquisition system detects a count image, representing the activity distribution in the patient,
in a certain LOR, it is not possible to determine where the from the collected projection data (either stored as list-mode
files or sinograms). There are two major types of image volumes. A further advantage of hybrid imaging is the pos-
reconstruction algorithms: (1) algorithms that are based on sibility to improve the image quality of each technique
(filtered) back-projection or (2) algorithms that apply an using the provided mutual information. When two tech-
iterative reconstruction process. The latter is the de facto niques are combined in a single scanner, we refer to it as a
standard in most clinical systems. During image reconstruc- hybrid system.
tion the corrections mentioned above (normalization, ran-
dom and scatter corrections, attenuation, dead time, and
decay) are applied in order to generate a quantitatively cor- 9.5.2 PET/CT Instrumentation
rect PET image. When available, time-of-flight information
is included, as it allows for improved image quality. A In a hybrid PET/CT system, a PET scanner and an X-ray
detailed review on various image reconstruction methods computed tomography (CT) are combined into a single
and their properties can be found in [20]. More recently, iter- hybrid system. Computed tomography or CT is a medical
ative reconstruction methods that include a model for the imaging technique able to measure the distribution of linear
PET system spatial resolution have been introduced. By attenuation coefficients (μ(x, y, z)) in sections of the subject
including the system spatial resolution, expressed as point of interest. From the clinical point of view, it provides ana-
spread function, PET images with improved spatial resolu- tomical information distinguishing the various tissues or
tion can be generated resulting in improved lesion cavities within the body and the localization of a contrast
detectability but at the cost of so-called Gibbs artifacts,
agent, if used.
which may cause quantitative biases [21]. CT provides information that are highly complementary
to the PET information, since the CT provides an anatomical
reference to the activity density map ρ(x, y, z) measured by
Key Learning Points PET. With CT, issues can be analyzed for possible disease-
• The sequential list of coincidences can be sorted related anatomical abnormalities, while PET is able to pro-
into integrated time intervals (frames) and saved in vide functional or molecular information only. A typical
one or more sinograms. application of PET/CT is in oncology, where CT can provide
• Data corrections are needed before projection data the size and location of a tumor, while PET can characterize
can be reconstructed into a PET image. the lesion from the metabolic point of view. Knowing the
• The principle of time-of-flight PET has a beneficial size and shape of the lesion improves the quantification of
effect on image quality with a significant gain in PET images, for example, by providing information for cor-
SNR. recting the measured activity in a certain volume with the
• Iterative algorithms are today the most widely used most appropriate recovery coefficient and thus correcting for
image reconstruction technique to obtain a quanti- the partial volume effect.
tatively accurate PET image, representing the dis- A further advantage of the PET/CT approach is the pos-
tribution of the radiotracer in the patient. sibility to obtain attenuation coefficients from CT to be used
for the attenuation correction of PET data, thus further
improving the quality of PET image quantification.
From the engineering point of view, the major difficulty
9.5 ET/CT and PET/MR Clinical Scanners:
P stays in the precise mechanical integration that is required by
State of the Art the combination of the two systems. Possible misalignments,
due to nonoptimal translational motion of the patient bed
9.5.1 Hybrid Imaging along the axis, may cause significant image mismatch with
direct consequences on the quality of image fusion. The PET
Hybrid imaging can be defined as the combination of dif- component of a typical PET/CT is actually a “normal” stand-
ferent imaging techniques. A major advantage of hybrid alone PET, while the CT component may differ from a stand-
imaging is the possibility to provide complementary infor- alone CT. In fact, the quality of CT images to complement
mation. In the case of PET/CT, molecular information the PET information can be lower than that of diagnostic CT
from PET can be combined with morphological informa- systems, where a higher spatial resolution is required.
tion from CT in a single imaging session and without mov- However, to fully exploit the potential of PET/CT, a high-
ing the patient from the imaging bed. Images can therefore quality multi-slice CT component is mounted in high-end
be fused with high accuracy and minimal distortions or commercial PET/CT.
motion blurring, thus facilitating image analysis and PET Because of the great advantages of PET/CT versus PET
quantification. For example, regions of interest can be alone, as of today, no stand-alone PET systems are available
drawn on the CT images improving the definition of target on the market anymore.
9.5.2.1 CT Technology in Brief in scan time for whole-body acquisition and a reduced dose
A comprehensive technological description of a CT system delivered to the patient with respect to sequential scans.
is beyond the scope of this chapter. However, some technical
features of a modern CT system are described here for better 9.5.2.2 CT-Based Attenuation Correction
illustrating the potential of PET/CT. of PET Data
Computed tomography is based on the transmission of an The use of the attenuation information obtained from a CT
X-ray beam through the patient’s body. In this way, the CT scan can overcome limitations of transmission scans. Lower
system collects line integrals of the distribution of the linear noise attenuation maps can be obtained, but the values of the
attenuation coefficients μ(x,y,z) at different angles. Then, attenuation coefficients can be less accurate. In fact, the
similar to other tomographic techniques such as PET, an attenuation coefficients μCT(x, y, z) are first obtained at the
algorithm reconstructs the image. In CT imaging, analytical CT energies and with a non-monochromatic X-ray beam and
reconstruction algorithms, such as the filtered back- are therefore numerically far from those at 511 keV. A con-
projection (FBP), are often used. version to the linear attenuation coefficients at 511 keV
The basic components of a CT system are an X-ray source μ511 keV (x, y, z), for each of the crossed tissue, should be then
and an X-ray detection system located at the opposite sides performed through a procedure called energy scaling.
of the patient. This entire system is rotated around the object CT images are typically displayed in Hounsfield units
so as to collect tomographic information and create a sino- defined as:
gram to be fed into the image reconstruction algorithm.
X-ray sources for CT are high-power X-ray tube featuring a m tissue - m water
HU = ´ 1000
rotating tungsten anode with an output energy spectrum in m water
the 80 and 140 kVp range. The output spectrum is then
continuous (i.e., non-monochromatic). X-ray detectors for The conversion of μCT to μ511 keV cannot be obtained
CT are based on ceramic scintillators coupled to solid-state through a simple multiplicative factor, because of the differ-
photodetectors. In order to collect enough line integrals, the ent fraction of Compton and photoelectric interactions in the
detector extends along the necessary arc so as to cover the various tissues (strong variations can be observed between
whole diameter of the field of view and is finely pixelated. In bones and soft tissues) and because of such a wide range of
fact, the size of each element of the matrix affects the spatial energies. There are different ways to perform energy scaling.
resolution that can be obtained in the reconstructed image. In its simplest form, energy scaling is obtained with a linear
The rotation is typically continuous with a speed of about scaling with a coefficient from −1000 HU up to a given HU,
one round per second or less. Modern CT detectors have thus including mainly soft tissues that are approximated as a
600–900 columns and 1–64 rows (the latter placed along the mixture of air and water and with a different coefficient (typ-
scanner axial direction). Each element is about ically lower) above that value of HU, where tissues can be
0.6 mm × 1 mm. The number of rows defines the number of approximated as a mixture of water and bones or, more accu-
sections (or slices) of the object that can be obtained in a rately, of air and bones. Conversion coefficients and thresh-
single system rotation (without any motion of the bed along old values are usually suggested by the scanner manufacturer
the axis). Scanners with more than one slice (multi-slice CT) and may vary due to the assumed approximations and for the
are classified according to the number of slices: the greater is technological differences of used CT scanners. The linear
the number of slices, the larger is the section of the patient scaling approximation can fail in the presence of high-Z
that can be scanned with a single rotation and the shorter is materials that cannot be approximated as a mixture of air,
the scan time (and the patient dose) required. A further water, and bones, such as metal implants or when CT con-
reduction of patient dose can be obtained at the expenses of trast agents are used.
the spatial resolution when detector pixels are grouped
together to mimic larger pixels. 9.5.2.3 State of the Art in Clinical PET/CT
In order to scan a longer section of the patient, a sequen- As of today, almost all PET/CT systems use lutetium-based
tial scan can be performed, i.e., images from standard circu- scintillators with only one still using BGO with the aim of
lar acquisitions that are taken at different axial positions of maximizing system sensitivity [22]. The pixel pitch of mod-
the patient are stitched together to create the whole image ern systems ranges from 6.3 mm down to 4.0 mm. Time-of-
volume. As of today, all CT systems offer the further possi- flight capability is featured or available or offered as an
bility to perform helical scanning. In this procedure, called optional by all PET systems, except the one using
spiral CT, the patient is translated during detector rotation. BGO. PMTs still represent the preferred choice for the read-
Line integral data obtained during the scan are then interpo- out of scintillating crystals in PET, but silicon photomultipli-
lated to create missing projections and generate a sinogram. ers have recently been introduced in clinical systems. For
The advantages of this approach are essentially a reduction example, the first commercial PET/CT clinical scanner based
on the digital SiPM technology has been released by Philips of research, thanks to the capability to provide co-registered
[23], thus making it possible to achieve a spatial resolution images and to monitor time-dependent metabolic processes.
of 4 mm FWHM and a coincidence resolving time of 345 ps The development of new bimodal tracers can be particularly
FWHM, which represents a significant improvement with useful in the study of neurological diseases and in preclinical
respect to commercial TOF-PET/CT systems based on PMT applications for pharmaceutical research.
technology, where coincidence time resolution is between The combination of PET and MRI in a single is techni-
500 and 700 ps. cally more challenging than PET and CT because of the
The sensitivity of a PET system is one of the most impor- presence of magnetic fields that impair proper functioning of
tant figures of merit. Sensitivity is dominated by detector PMTs and because of the fast PET electronics that may dis-
material and thickness but also by ring diameter and axial turb the detection of MR radiofrequencies.
field of view extension and recording efficiency. Regarding For this reason, a hybrid PET/MR scanner for clinical
the field of view diameter, one must also consider that only application has been initially proposed in a tandem configu-
PET with a larger bore can accommodate obese patients. As ration, i.e., with the two scanners mounted together back-to-
of today, all the commercially available PET/CT systems back allowing a sequential acquisition in which the patient
have, in general, comparable configuration and performance bed is automatically moved from the MR to the PET
parameters with a maximum sensitivity at the center of the scanner.
FOV in the range of 7.5–20 cps/kBq and a spatial resolution An example of tandem scanner is represented by Philips
of about 4–5 mm FWHM. It is also clear that time-of-flight PET/MR Ingenuity [24], which is composed of a 3 T magnet
imaging is generally superior to non-TOF-corrected imaging and of a modified version of the time-of-flight PET used in
because of the improvement in signal-to-noise ratio. the commercial PET/CT hybrid systems. The two systems
However, some differences among the available systems can are aligned at a distance of about 4.2 m, so that the same bed
be found in the gantry aperture (70–90 cm) and axial FOV can automatically slide from one scanner to the other. Even
(15–26 cm) extension, or in the CT performance, or in the if the PET ring is located away from the magnet, the mag-
availability of TOF capability. netic field is still high enough to interact with the PMTs; an
In the last 15 years, PET/CT has changed the role of additional shielding is then mounted around the PMT detec-
nuclear medicine with an impressive image quality improve- tors of the PET ring. The PET system can achieve a coinci-
ment with respect to stand-alone PET. On the other hand, dence time resolution of 525 ps and a sensitivity of 7.0 cps/
some drawbacks are still present. PET/CT still remains a kBq. The main disadvantage of this configuration is that the
sequential measurement, and involuntary subjects or physi- two images need to be acquired separately and then merged
ological movement is possible, thus leading to movement together.
artifacts. Furthermore, the additional radiation exposure due The availability of new-generation photodetectors based
to the CT cannot be neglected, and this can be critical in on semiconductor materials, such as avalanche photodiodes
pediatric investigations and in repeat follow-up studies. and silicon photomultipliers, boosted the development of
PET/MR scanners where the two systems are fully integrated
within the same gantry.
9.5.3 PET/MR Hybrid Systems The first fully integrated PET/MR system introduced in
the market was the Siemens mMR scanner [25]. In this
In magnetic resonance imaging (MRI), strong magnetic model, the PET ring is installed inside the 3 T-MR bore,
fields, radio waves, and field gradients are used to reveal between the gradient coil and the RF coil, the PET is com-
structure and functions of the organs in the body. This modal- posed of matrices of LSO crystals readout by APDs, and the
ity has a good sensitivity (10−3–10−5 mol/L) and an excellent scintillating light is shared on multiple photodetectors so that
spatial resolution (≈1 mm isotropic for clinical systems). The a matrix of 8 × 8 scintillators can be decoded by only 3 × 3
combination of morphological/functional MRI information APDs. The reduced volume available for the PET system
with the nanomolar functional information given by PET can leads to a PET diameter smaller than that of standard PET
be used to provide a whole spectrum of information that can scanners; the transaxial FOV is in fact of 58.8 cm. The adop-
be used to understand new aspects of the anatomy and physi- tion of APDs as photodetectors does not enable to achieve
ology of a disease well beyond the capability of each of the TOF capability (coincidence time resolution is about 2.9 ns),
two modalities alone. The main applications of hybrid PET/ but thanks to the high axial elongation of about 25 cm, it can
MR are cancer diagnosis, treatment, and follow-up, in par- feature a sensitivity of 13.3 cps/kBq.
ticular when the superior imaging capabilities of MR for soft A PET/MRI scanner employing SiPM technology is now
tissues over computed tomography (CT) are more relevant, produced by GE [26]. The use of SiPMs enables to fully inte-
such as in the head and neck, abdomen, and prostate. grate a TOF-graded scanner inside the MR bore reaching a
Furthermore, in neurology, hybrid PET/MR opens new fields resolution below the 400 ps (comparable with that of
h igh-level TOF-PET/CT scanners), without observing sig- For PET/MR systems, attenuation correction is a particu-
nificant degradations in the MR image quality. Thanks to the lar challenge. MR images do not represent attenuation coef-
adoption of the time-of-flight correction and to the high ficients as CT images do. Typically, MR data relate to proton
angular coverage, the system enables to reach a sensitivity of density; therefore, the conversion of MR data into 511-keV
22 cps/kBq at the center of the field of view. photon attenuation coefficients is not as straightforward as
with CT. On most PET/MR systems, a dedicated MR attenu-
ation correction protocol is provided [29]. It typically starts
with the collection of a fast, dedicated MR sequence fol-
Key Learning Points lowed by segmentation of the image into three or four tissue
• Hybrid systems allow the collection of complemen- classes. Often the bone is ignored and lungs are assumed to
tary morphological and functional/molecular be uniform. Moreover, the MR field of view is smaller than
information. that of the PET, and typically arms are truncated. Finally,
• CT can be used also to correct for the attenuation of patient’s fixation devices and MR coils are not seen by MR
the high-energy photons used for PET imaging, and thus are not visible or present in the MR image. These
thus increasing the PET capability to yield quantita- assumptions and limitations result in quantitative biases in
tive information on tracer uptake. the reconstructed PET images from hybrid PET/MR studies.
• PET/MRI enables to reduce the dose with respect to Despite these limitations of MR-based attenuation correc-
PET/CT and to take advantage of the whole spec- tion, the quantitative performance of PET/MR examinations
trum of information provided by an MRI exam. appears to be comparable to that of PET/CT studies, yet spe-
cific attention to MR attenuation correction artifacts is war-
ranted during image interpretation. Currently, several new
MR image processing methods, PET image reconstruction
9.6 PET/CT and PET/MR Quantification algorithms, and specific MR sequences have been developed
that can mitigate many of the MR attenuation correction
As described earlier in this chapter, PET yields images that shortcomings mentioned above. The use of these methods
represent the activity distribution of the radiotracer in the makes PET/MR studies equally quantitatively accurate as
patient. However, for quantitative interpretation of the those with PET/CT, although the commercial release of
images, it is required that all necessary corrections (e.g., nor- some of these methods is lagging behind.
malization, randoms and scatter, attenuation) are accurately
implemented and applied. Patient’s motion, the presence of
metal implants, and the limited spatial resolution of PET sys- 9.6.2 F
actors Affecting PET/CT and PET/MR
tems may still result in quantitative biases and/or image arti- Quantification
facts that impair both image quality and quantification. An
overview of image artifacts can be found in [27]. When all necessary corrections for generating quantitative
PET images have been applied during data acquisition and
image reconstruction, the PET images can in principle be
9.6.1 T
echnical Aspects Concerning analyzed quantitatively. Yet, several other factors impact the
in Particular Attenuation Correction quantitative accuracy and precision of PET. Such factors can
be classified as technical, physical, and biological in nature.
In PET/CT the attenuation correction is derived from the CT Technical factors relate to clock synchronization, calibration
data after conversion of the CT Hounsfield units into 51-keV of the PET/CT/MR system, correct information on the net
photon attenuation coefficients by bilinear scaling [28]. The injected activity, and accurate activity measurements.
presence of contrast agents and metal implants may cause Physical factors are those that impact on image quality, such
scaling errors and thus result in incorrect attenuation coeffi- as the settings and methods used during image reconstruc-
cients. Moreover, sometimes CT images are truncated, i.e., tion, methods used for image interpretation and analysis, etc.
part of the patient anatomy is not covered by the Biological factors are, for example, tracer uptake time and
CT. Obviously, in case of truncated CT data, the PET attenu- variation in tracer delivery and circulation (arterial input
ation correction will not be complete, and attenuation will be function). For many tracers, in particular for [18F]FDG oncol-
underestimated. Finally, patient motion may cause a spatial ogy studies, detailed recommendations and guidelines are
misalignment between the PET and the CT data, therefore in provided to assure optimal image quality and reduce vari-
under- and overcorrection of the PET attenuation. ability in (quantitative) results [30].
9.6.3 M
ethods and Metrics for PET/CT 9.6.3.2 Standardized Uptake Values
and PET/MR Quantification The most commonly used parameter to quantify radiotracer
uptake is the standardized uptake value (SUV). SUV is the
Two main approaches are being used for the quantification of uptake of the tracer in a defined region of interest at a speci-
radiotracer uptake. The first one is often referred to as full fied uptake time as measured by the PET system (CROI,t,
quantification by means of pharmacokinetic modeling, while expressed, e.g., in kBq/mL) normalized to injected activity
the second one is tracer uptake quantification by so-called (Ainj.) per patient weight (total body weight, or TBW), i.e.:
standardized uptake values.
CROI,t [ kBq / mL ]
SUVTBW =
9.6.3.1 Kinetic Modeling Ainj. [ MBq ] / TBW [ kg ]
Full quantification by kinetic modeling requires a dynamic
PET acquisition in order to measure the uptake of the tracer Normalization of tracer uptake by injected activity per,
continuously starting directly with the tracer administration e.g., body weight is based on the assumption that this nor-
and continuing over a scanning period sufficient to charac- malization factor is equally proportional to the (integral of
terize the tracer kinetics. For many tracers a dynamic study the) input function for all subjects. Sometimes weight is
requires a PET acquisition of 60–90 min. The time course of replaced by lean body mass (SUVLBM), body surface area
radiotracer uptake, or time activity curve (TAC), is fitted with (SUVBSA), or body mass index (SUVBMI). The optimal nor-
a kinetic model. malization factor depends on the distribution of the tracer
The kinetic model is a mathematical model describing the throughout the body, organs, and fat. In case tracer uptake in
fate of the tracer over time in the tissue(s) of interest. The fat is low, lean body mass may prove to be more optimal than
model requires an arterial input function, which describes body weight.
concentration of the radiotracer in plasma over time. The The main advantage of radiotracer uptake quantification
input function thus reflects the availability of the radiotracer by SUVs is that it does not require a dynamic imaging proce-
from plasma to tissue(s) as a function of time. The input dure, which limits the anatomical coverage often to a single
function can be measured invasively by inserting an arterial bed position (typically <20 cm). SUV-based quantification
cannula and withdrawing arterial blood samples over time. can be performed with static PET imaging where tracer
Additional corrections for plasma to blood ratio and tracer uptake is measured at a specific uptake time. Moreover, it
metabolism may require additional processing of arterial allows for whole-body imaging which is of utmost impor-
blood samples. Once the input function is measured and pro- tance in oncology imaging to assess the presence of metasta-
cessed, it is used in combination with the model to fit the sis and thus clinical disease staging. Yet, quantification with
observed PET time activity curves in the tissue(s) of interest. SUVs requires a careful validation to guarantee that the
Once an optimal fit has been achieved, the model provides assumptions made are not violated. The main assumptions of
detailed information on the transport of the tracer from SUV quantification are the following: (1) the normalization
plasma to tissue(s) and the various kinetic parameters that factor used, activity per weight or lean body mass, is a good
characterize this process. Kinetic modeling provides a way approximation for (the integral of) the input function under
of quantifying tracer uptake that includes as many as possi- all conditions, i.e., there is no or there is minimal intra- (lon-
ble factors and corrections. gitudinal) and intersubject (cross-sectional) variability in the
The procedures required for full quantitative kinetic mod- input function; (2) postinjection, there is a point in time
eling can sometimes be simplified deriving the activity in where radiotracer uptake is fairly constant and where changes
blood over time directly from the dynamic acquisition, i.e., in uptake within and between subjects are larger than those
by using the so-called image-derived input functions. For caused by unwanted variations in uptake time. The latter
brain studies, often a noninvasive reference tissue approach often occurs during busy schedules in the imaging center.
can be applied enabling to quantitatively measure radiotracer Alternative but similar quantitative measures are SUV
binding without the need for arterial sampling [31]. Yet, it is ratios (SUVR), thereby uptake of the tracer in the region of
clear that the use of (plasma input) kinetic models is inva- interest is normalized to uptake of the tracer in a reference
sive, dynamic scans are time consuming, and kinetic analysis region or organ. In brain studies, the reference region is ide-
require specific expertise not commonly available in most ally a brain region free of specific tracer binding. In oncol-
clinical imaging department. Therefore, simplified ogy studies the uptake in blood, derived by placing a region
approaches that can be applied using a static imaging proce- of interest in a blood pool structure, is sometimes used as
dure are most commonly used and are clinically more reference or normalization factor, thus providing tumor-to-
feasible. blood ratios (TBR).
The optimal imaging procedure and simplified quantita- 9. Bailey DL, Meikle SR. A convolution-subtraction scatter correc-
tion method for 3D PET. Phys Med Biol. 1994;39:411–24.
tive uptake metrics, SUV, SUVR, or TBR, need to be care- 10. Bentourkia M, Msaki P, Cadorette J, Lecomte R. Assessment of
fully explored. Assumptions regarding constancy of input scatter components in high-resolution PET: correction by nonsta-
function, and/or appropriateness of normalization factors to tionary convolution subtraction. J Nucl Med. 1995;36:121–30.
approximate these functions, need to be validated both before 11. Levin CS, Dahlbom M, Hoffman EJ. A Monte Carlo correction for
the effect of Compton scattering in 3-D PET brain imaging. IEEE
and during or after treatment and/or between subject groups Trans Nucl Sci. 1995;42:1181.
in order to avoid misinterpretation of SUVR variations or 12. Grootoonk S, Spinks T, Sashin D, Spyrou N, Jones T. Phys Med
(treatment response) changes. Biol. 1996;41:2757.
13. Shao L, Freifelder R, Karp JS. Triple-energy- window scatter cor-
rection technique in PET. IEEE Trans Nucl Sci. 1994;13:641–8.
14. Ollinger JM. Model-based scatter correction for fully 3D PET. Phys
Key Learning Points Med Biol. 1996;41:153–76.
• CT-based attenuation is relatively straightforward, 15. Watson CC. New, faster, image-based scatter correction for 3D
PET. IEEE Trans Nucl Sci. 2000;47:1587–94.
but energy scaling of the CT image is required 16. Knoll G. Radiation detection and measurement. Chichester, NY:
before it can be used to derive the attenuation cor- John Wiley & Sons; 2010.
rection for PET. 17. Bailey DL. Transmission scanning in emission tomography. Eur J
• In PET/MR systems, the attenuation correction Nucl Med. 1998;25:774–87.
18. Karp JS, Surti S, Daube-Witherspoon ME, Muehllehner G. Benefit
must be derived from an MR image, an approach of time-of-flight in PET: experimental and clinical results. J Nucl
that is more challenging than CT-based attenuation Med. 2008;49:462–70.
correction. 19. Surti S. Update on time-of-flight PET imaging. J Nucl Med.

• PET image quantification is affected by many fac- 2015;56:98–105.
20. Tong S, Alessio AM, Kinahan PE. Image reconstruction for PET/
tors, being of technical, related to imaging physics, CT scanners: past achievements and future challenges. Imaging
or biological in nature. Med. 2010;2:529–45.
• Full kinetic analysis provides the most accurate 21. Rahmim A, Qi J, Sossi V. Resolution modeling in PET imaging: the-
way of image quantification but requires complex ory, practice, benefits, and pitfalls. Med Phys. 2013;40(6):064301.
https://doi.org/10.1118/1.4800806.
imaging procedures and image analysis methods. 22. Morzenti S, De Ponti E, Guerra L, Zorz A, Landoni C, Crivellaro
• Simplified methods, such as SUV and SUVR, are C, et al. Performance evaluation of the discovery IQ - GE PET/
more easily applicable in the clinical setting but CT scanner according to NEMA NU2-2012 standard. J Nucl Med.
need to be validated carefully for new PET tracers 2015;56:1846.
23. Miller M, Zhang J, Binzel K, Griesmer J, Laurence T, Narayanan
to understand possible sources of error and M, et al. Characterization of the Vereos digital photon counting
limitations. PET system. J Nucl Med. 2015;56:S434.
24. Zaidi H, Ojha N, Morich M, Griesmer J, Hu Z, Maniawski P, et al.
Design and performance evaluation of a whole-body Ingenuity TF
PET-MRI system. Phys Med Biol. 2011;56:3091–06.
25. Delso G, Fürst S, Jakoby B, Ladebeck R, Ganter C, Nekolla SG,
References et al. Performance measurements of the Siemens mMR integrated
whole-body PET/MR scanner. J Nucl Med. 2011;52:1914–22.
1. van Eijk CW. Inorganic scintillators in medical imaging. Phys Med 26. Grant AM Deller TW, Khalighi MM, Maramraju SH, Delso G,
Biol. 2002;47:R85–R106. Levin CS. NEMA NU 2-2012 performance studies for the SiPM-
2. Lecomte R. Novel detector technology for clinical PET. Eur J Nucl based ToF-PET component of the GE SIGNA PET/MR system.
Med Mol Imaging. 2009;36:S69–85. Med Phys. 2016;43:2334.
3. Casey M, Nutt R. A multicrystal two dimensional BGO detector 27. IAEA. Human health series no. 27. PET/CT atlas on quality control
system for positron emission tomography. IEEE Trans Nucl Sci. and image artefacts. Vienna: International Atomic Energy Agency;
1986;33:460–3. 2014.
4. Derenzo SE, Moses WW, Huesman R, Budinger TF, et al. 28. Kinahan PE, Hasegawa BH, Beyer T. X-ray-based attenuation cor-
Quantification of brain function. Amsterdam: Elsevier Science rection for positron emission tomography/computed tomography
Publishers; 1993. p. 25. scanners. Semin Nucl Med. 2003;33:166–79.
5. Moses WW. Fundamental limits of spatial resolution in PET. Nucl 29. Beyer T, Lassen ML, Boellaard R, Delso G, Yaqub M, Sattler B,
Instrum Methods Phys Res A. 2011;648(Suppl 1):S236–40. et al. Investigating the state-of-the-art in whole-body MR-based
6. National Electrical Manufacturers Association. NEMA stan- attenuation correction: an intra-individual, inter-system, inventory
dards publication NU 2-2001: performance measurements of study on three clinical PET/MR systems. MAGMA. 2016;29:75–87.
positron emission tomographs. Rosslyn, VA: National Electrical 30. Boellaard R, Delgado-Bolton R, Oyen WJ, Giammarile F, Tatsch
Manufacturers Association; 2001. K, Eschner W, et al. FDG PET/CT: EANM procedure guidelines
7. Fahey FH. Data acquisition in PET imaging. J Nucl Med Technol. for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging.
2002;30:39–49. 2015;42:328–54.
8. Badawi RD, Marsden PK. Developments in component-based nor- 31. Lammertsma AA, Hume SP. Simplified reference tissue model for
malization for 3D PET. Phys Med Biol. 1999;44:571–94. PET receptor studies. Neuroimage. 1996;4(3 Pt 1):153–8.
Further Reading Vandenberghe S, Marsden PK. PET-MRI: a review of challenges and

solutions in the development of integrated multimodality imaging.
Phys Med Biol. 2015;60:115–54.
Boellaard R, Quick HH. Current image acquisition options in PET/
Zaidi H, Del Guerra A. An outlook on future design of hybrid PET/MRI
MR. Semin Nucl Med. 2015;45:192–200.
systems. Med Phys. 2011;38:5667–89.
Del Guerra A, Belcari N, Bisogni MG. Positron emission tomography:
its 65 years. La Rivista del Nuovo Cimento. 2016;39:155–223.
Rausch I, Quick HH, Cal-Gonzalez J, Sattler B, Boellaard R, Beyer
T. Technical and instrumentational foundations of PET/MRI. Eur J
Radiol. 2017;94:A3–A13.
Essentials of Quantitative
Imaging with PET 10
Adriaan A. Lammertsma
Contents
10.1 Characteristics of PET 220
10.2 Issues in Quantification 220
10.3 Data Acquisition Issues 220
10.3.1 Normalization 220
10.3.2 Injected Activity 220
10.3.3 Cross-Calibration 221
10.4 Data Processing Issues 221
10.4.1 Decay and Dead-Time 221
10.4.2 Random and Scattered Coincidences 221
10.4.4 Tissue Attenuation 221
10.4.5 Patient Motion 222
10.4.6 Volume of Interest Definition 222
10.4.7 Partial Volume Effects 222
10.5 Data Analysis Issues 223
10.5.1 Arterial Input Function 223
10.5.2 Tissue Uptake 224
10.6 Compartmental Models 224
10.6.1 Single-Tissue Compartmental Model 224
10.6.2 Two-Tissue Compartmental Model 225
10.6.3 Reference Tissue Models 227
10.7 Practical Modelling Issues 228
10.7.1 Weighting Factors 228
10.7.2 Model Selection Criteria 229
10.8 Parametric Imaging 230
10.8.1 Linearizations 230
10.8.2 Patlak and Logan Plots 230
10.8.3 Basis Function Methods 230
10.8.4 Validation 231
10.8.5 Semi-Quantification 231
10.9 Conclusions 231
References 232
Learning Objectives
• To understand the basic principle of PET quantification
A. A. Lammertsma (*) • To realize that uptake images represent summed images
Department of Radiology and Nuclear Medicine,
Amsterdam UMC, Vrije Universiteit Amsterdam,
of several biological processes
Amsterdam, The Netherlands • To get an understanding when and how PET data should
e-mail: aa.lammertsma@vumc.nl be quantified

220 A. A. Lammertsma
• To realize that quantification requires attention to many rial sampling may be essential. Finally, for many applica-
data acquisition, data processing, and data analysis issues tions, an intermediate semiquantitative method may be
• To understand the need for an accurate arterial input function adequate. Nevertheless, all semiquantitative methods should
• To know when a reference tissue input function is first be validated by comparing results with those derived
sufficient from a fully quantitative method.
• To understand the basic principles of tracer kinetic
modelling
• To understand the basic principles of parametric imaging Key Learning Point
• To understand the limitations of simplified semiquantita- • The level of quantification required depends on the
tive methods clinical or research question to be addressed.
10.1 Characteristics of PET 10.3 Data Acquisition Issues
The two most important characteristics of positron emission PET is based on the coincidence detection of two annihila-
tomography (PET) are its high sensitivity (pico- to nano- tion high-energy photons resulting from the interaction of an
molar range) and its ability to accurately measure the con- emitted positron with an electron in tissue. Therefore, the
centration of a positron-emitting radionuclide within the line of response (the line along which the annihilation took
human body. In fact, PET was developed in the 1970s as a place) can be identified, which is the basis for the quantitative
noninvasive in vivo method to measure regional physiology nature of PET. This is primarily due to the fact that the total
(originally blood flow, oxygen utilization, and glucose path length of the two annihilation photons through tissue is
metabolism) in humans [1, 2]. Its high sensitivity facilitated known. A separate acquisition of a CT scan provides infor-
the development of tracer studies of neuroreceptors in the mation about attenuation, to correct the detected photons for
1980s. The high sensitivity of PET made it possible to use attenuation within the tissues. Absolute measurements of
(tracer) amounts of labeled compounds that did not affect the radioactivity concentrations require attention to detail, and
receptors themselves (negligible receptor occupancy). At the the following data acquisition issues should be considered.
same time, proper quantification remained important for cor-
rect interpretation of the signals captured in the image. Both
sensitivity and quantitative accuracy are unique for PET and 10.3.1 Normalization
make it the method of choice for investigating molecular tar-
gets and interactions in vivo. Normalization of the scanner is needed to correct for differ-
ences in sensitivity between detector pairs (similar to perform-
ing a “flood” correction with a single-photon gamma camera).
Key Learning Points
In general, the procedure prescribed by the manufacturer suf-
• The two most important characteristics of PET are
fices. This procedure needs to be repeated occasionally depend-
high sensitivity and quantitative accuracy.
ing on the possible drift of the scanner. Again, the manufacturer
• PET is the method of choice for measuring molecu-
usually has a daily quality control procedure in place to detect
lar targets and interactions in vivo.
changes that would compromise scanner performance.
10.2 Issues in Quantification 10.3.2 Injected Activity
Clinically, PET studies are usually performed using simple The injected activity should be in a range that can be handled
qualitative imaging. However, in many circumstances, quan- by the scanner. It is clear that reconstructed images will be very
tifying the data provides additional, clinically relevant infor- noisy if the activity is too low, making it difficult to extract reli-
mation (e.g., measuring absolute myocardial blood flow at able (precise) radioactivity measurements at least at the voxel
rest and stress in a myocardial perfusion scan). The particu- level. On the other hand, if the activity is too high, the signal
lar type of study (and associated scanning protocol) will may be dominated by random coincidences (quadratic increase
depend on the actual research or clinical question that needs with activity) rather than true coincidences (linear increase
to be addressed. For example, for staging in oncology, it is with activity). As the true coincidences are obtained by sub-
important to identify and localize sites with increased [18F] tracting the random counts from the total counts [3], the true
FDG uptake, and, consequently, a simple static scanning signal will become increasingly noisy, as it is obtained by sub-
protocol with visual (qualitative) interpretation is sufficient. tracting two large numbers from each other. In addition, with
In other cases, e.g., for monitoring response to therapy, full very high activities, dead-time may become an issue, and the
quantification using a dynamic scanning protocol and arte- response of the scanner could even become nonlinear.
10 Essentials of Quantitative Imaging with PET 221
10.3.3 Cross-Calibration outside the subject [4]. Particularly vulnerable are areas with
low tracer uptake, where an error in the scatter contribution
Fully quantitative PET studies usually require measurement has more impact on the accuracy of estimated radioactivity
of the arterial input function. Measurement of such an input concentrations. This means that the accuracy of the scatter
function will be discussed later, but independent of those correction algorithm should really be checked if a region
measurements themselves, the scanner needs to be cross- with low tracer uptake is important for the final outcome
calibrated against a well-counter used for measuring manual measures (e.g., reference tissue models, target-to-reference
blood samples. This is done by scanning a homogeneous tissue ratios).
phantom and measuring an aliquot of the radioactivity solu-
tion in the well-counter.
For semiquantitative studies the PET signal is often normal- 10.4.3 Image Reconstruction
ized to injected activity. This means that a cross- calibration
between scanner and dosimeter is required. This cross-calibra- An essential step in the processing of acquired PET data is
tion needs to be carried out with care, as technically it is more their reconstruction into images displaying the distribution
demanding than the cross-calibration with a well-counter. of radioactivity within the field of view. Originally, a filtered
back projection (FBP) algorithm was used. This algorithm
provides an analytical solution to the reconstruction prob-
Key Learning Points lem. It is accurate but at the same time sensitive to noise
• For accurate results it is important to check perfor- (poor image quality for low count densities), sometimes even
mance of the scanner on a regular basis. showing streak artifacts. State-of-the-art scanners are no lon-
• The injected dose should be low enough such that ger equipped with the FBP algorithm, at least not as the stan-
the scanner response is linear and high enough to dard choice. Nowadays, an iterative reconstruction algorithm
avoid excessive noise. will be the standard option for clinical studies. Iterative
reconstructions provide significantly better image quality
(hence their popularity), but they may be biased especially in
10.4 Data Processing Issues high-contrast images where the optimal solution has not
been reached within the limited (preset) number of itera-
Before data can be analyzed to extract quantitative biological tions. It is important to check the validity of an iterative
information, the acquired data need to undergo several cor- reconstruction algorithm, especially when an image-derived
rection and processing steps. Each of these steps could affect arterial input function is used [5].
the accuracy of the final results. The most common issues are
listed below.
10.4.4 Tissue Attenuation
10.4.1 Decay and Dead-Time During reconstruction acquired data need to be corrected for
tissue attenuation. For (older) stand-alone PET scanners, this
Prior or during reconstruction, acquired PET data need to be could be done by acquiring a transmission scan using an
corrected for decay (if analysis is based on decay-corrected external positron-emitting source. In theory, this is the most
data), dead-time, random coincidences, and scattered coinci- accurate method to correct for tissue attenuation, as the
dences. These corrections routinely are incorporated in the external source allows for direct measurements of the attenu-
scanner software. Usually these corrections are taken for ation of 511 keV photons, the main limitation being the num-
granted, but for dynamic scanning protocols, it may be use- ber of counts that can be acquired within a reasonable
ful to check their accuracy, e.g., the accuracy of dead-time scanning time.
corrections for very short frames. In current state-of-the-art PET/CT scanners, correction
for tissue attenuation is based on a (fast) low-dose CT scan
[6]. Accuracy of this method has been well validated,
10.4.2 Random and Scattered Coincidences although misregistration of the emission and transmission
scans can occur if the CT scan is acquired during breath-hold
Nowadays the correction for random coincidences is usually and the emission scan during tidal respiration. In addition
performed by subtracting the random signal from the total artifacts may occur due to metal implants.
signal (see above). More challenging is the scatter correc- Attenuation correction for PET/MRI is an entirely differ-
tion, which accounts for about 30–50% of the total number ent issue. As it is not possible to image bone with standard
of acquired counts. This correction is based on both mea- MRI sequences, attenuation needs to be estimated in a differ-
surements and calculations, e.g., by modelling the scatter ent way. Two approaches have been followed. In the first
distribution within the subject based on counts measured method, attenuation coefficients are allocated to three or four
different tissue types, which are defined by segmentation of is to use a fixed-size VOI centered on the maximum value.
standard MRI images [7]. This method is actually imple- This reduces operator variability but may be prone to errors
mented on most commercial PET/MRI scanners. It is, how- when following effects of therapy in small lesions (that may
ever, well documented that this type of attenuation correction get smaller thereby imposing a variable partial volume effect).
(which basically ignores bone) can lead to substantial errors Another common option is to use a thresholding technique,
and therefore it is not suitable for quantitative PET studies where all voxels are incorporated that have a higher uptake
unless serial measurements are performed (e.g., measuring than a certain percentage of the maximum (e.g., 50% or 70%).
response to therapy), where the error will be similar for suc- To account for variable background levels, the threshold val-
cessive scans. A more promising technique is the use of ues can also be adjusted for the actual background level.
ultrashort MRI sequences, which may be able to image bone If the structures of interest are not directly related to
directly. This is a field that still is in development and the uptake in the PET scan, but to anatomical structures, a (low-
best method has yet to be identified. dose) CT scan or, especially for brain studies, an MRI scan
should be used. For brain studies it is common practice to
use one of the VOI templates that have been published [9]. In
10.4.5 Patient Motion that case, PET and MRI scans need to be co-registered, as
templates usually are based on MRI. Again, co-registration
An important source of error can be patient motion during accuracy for individual scans should be checked.
scanning, especially when long dynamic scans are acquired. Interestingly, often VOIs are defined for gray matter only,
It is, therefore, important to monitor each patient carefully following a gray/white matter segmentation of the MRI scan.
during scanning and to correct for motion immediately, when For multicenter studies and for studies comparing results
possible. This can, for example, be done by projecting laser with published data, it should be noted that this gray/white
beams on the skin and to mark the projected points at the matter segmentation may depend on the MRI sequence. In
start of the scan. In addition, it is important to assess whether addition, results from different scanners may not be the same
patient motion has occurred retrospectively, i.e., before ana- even if the same sequence is used. This leads to some inter-
lyzing a scan, especially when volumes of interest are small. study variability that should be taken into account when
For a dynamic scan, it is then possible to co-register all interpreting results.
dynamic frames to the same frame (if there is enough ana-
tomical detail in the individual frames; otherwise a series of
short frames need to be added together). It should be noted, 10.4.7 Partial Volume Effects
however, that if there has been motion during a scan, there
will also be a mismatch between the attenuation (transmis- Measured concentrations in small structures suffer from par-
sion or low-dose CT) scan and the frames that were acquired tial volume effects, i.e., the concentration in a structure
after the motion occurred. This means that co-registration is smaller than twice the spatial resolution of the scanner will be
not sufficient but that also the attenuation correction needs to underestimated. This is due to the fact that part of the activity
be adjusted [8]. This is also the case when there is patient within the structure will spread out of the boundaries of the
movement between attenuation and emission scans. In prin- structure due to the limited spatial resolution. Similarly,
ciple, it is also possible to monitor patient motion online activity from outside such a structure will spill-in.
using, for example, an external laser beam system, which Consequently, measured concentrations may be underesti-
provides a means to correct for motion during a frame rather mated or overestimated. Many different algorithms have been
than just between frames. Although systems have been avail- developed to correct for partial volume effects [10, 11], but
able for some years, they have not found wide application. there is no consensus on which is the most accurate one, and
results may differ. In the brain, for example, partial volume
corrections usually are based on gray/white matter segmenta-
10.4.6 Volume of Interest Definition tion of MRI scans. As mentioned above, this segmentation is
scanner-dependent, and therefore partial volume corrections
Whether data are analyzed on a region of interest basis or at a based on these segmentations will exaggerate interinstitu-
voxel level, at some stage volumes of interest (VOI) need to be tional differences. In addition, partial volume corrections
defined for comparing results between different patient may result in bias as they usually are based on regular (phan-
groups, for evaluating changes over time, or simply for report- tom) structures, while tissue structure are less regular. In gen-
ing results to the scientific community. If the structures of eral, it is good practice to analyze PET data with and without
interest can easily be identified on the PET scan, such as for partial volume corrections. This could, for example, be a way
tumors or metastatic lesions on an [18F]FDG scan, VOIs can to check whether observed differences between patient
be defined directly on the PET scan. The simplest way is to groups are due to partial volume effects or not.
manually define VOIs. However, results will be very operator In some cases, such as measuring myocardial blood flow
dependent and test-retest reliability may suffer. A better option using [15O]H2O, it is possible to include a partial volume
correction term in the operational equation [12]. In other result of an intervention (e.g., a drug). For fully quantitative
words, in fitting the data also a term accounting for partial studies, it is therefore necessary to measure the arterial input
volume is included. This leads to increased accuracy and function, i.e., the radioactivity concentration in arterial
there is no need to rely on anatomical imaging techniques. plasma as a function of time.
In general, a tracer is administered using a bolus intrave-
nous injection. Arterial concentrations first peak and then
Key Learning Points decline continuously during a scan, usually with very rapid
• Accuracy of various corrections should be checked changes early after injection. The arterial input function can
under conditions that resemble actual scans. be measured by taking multiple manual arterial blood sam-
• Attenuation correction on PET/MRI scanners ples during the scan, initially with fast sampling to catch the
should be validated for each application. peak of the curve and subsequently with lower frequency
• Several methods exist to correct for patient motion. when changes in the arterial concentration become slower.
• For larger motions, an adjustment of the attenuation Although quite labor intensive, fast manual sampling cer-
correction may also be required. tainly is a feasible option. The main drawback in terms of
• Reconstruction algorithms may have an effect on accuracy is the fact that timing errors are easily made, as it is
the final results, and therefore the same algorithm very difficult to keep track of the exact times arterial samples
should be used throughout a study. are actually taken, especially when unexpected hiccups
• Definition of volumes of interest should also remain occur, given that initially a frequency of once per 5–10 s is
consistent (and reproducible) within a study. required.
• Where appropriate, data should be corrected for par- A better solution is the use of an automatic online blood
tial volume effects, but even in those cases, data should sampler in which blood is withdrawn continuously from an
also be analyzed without partial volume correction. artery (usually the radial artery) using a dedicated pump,
with the radioactivity in the line monitored by an external
detector [13]. Using this setup a time resolution of 1 s can
easily be obtained. Apart from saving manpower (not only
10.5 Data Analysis Issues with respect to sampling itself but also when processing the
samples), the advantages of an online system are better time
After acquiring and processing PET data, taking into account resolution and no potential timing errors. A potential disad-
all issues mentioned above, the resulting images should be vantage of online blood withdrawal is the risk of obstruction
ready for analysis. At this stage, these images should provide due to blood clots. Care should be taken to keep the line open
an accurate estimation of the spatial distribution of radioac- throughout the study period by regular flushing, e.g., imme-
tivity within the field of view of the scanner, nothing more diately following manual (see below) samples. In addition,
and nothing less. However, the clinician will not really be because of the distance between cannula and detection site, a
interested in radioactivity measurements, but rather in the correction for delay (compared with arrival at the tissue)
underlying biology reflected by these radioactivity concen- should be made. It should be noted that a correction for delay
trations. In other words, a final step has to be taken, extract- should also be made in case of manual sampling as the transit
ing relevant biological information from these radioactivity time from the heart to the wrist will be longer than that from
measurements. To do so, two issues are relevant. the heart to, for example, the brain. For manual samples this
delay is ignored, as it is almost impossible to measure with
the time resolution of manual samples. Depending on the
10.5.1 Arterial Input Function application, it may also be necessary to correct for dispersion
in the tubing [14], but for many tracers this is not really nec-
Obviously, the actual uptake in tissue will depend on the essary as long as every effort is made to limit delay and dis-
injected activity. If in the same patient (at the same time) persion by using as few connectors as possible and keeping
twice as much would have been injected, the tissue concen- the line as short and thin (e.g., 1 mm) as possible.
tration would also have been twice as high. However, when a An alternative (less accurate) approach for measuring
tracer is administered intravenously, it will circulate through concentrations in blood is to include a large vascular struc-
the body. Exchange with tissue at the capillary level will ture (such as the aorta, carotid, or iliac artery) in the image
depend not only on the arterial concentration at that time but and determine the total blood activity from a volume of inter-
also on residual tissue activity from exchange at earlier est set within the structure.
times. Over time, tracer will be cleared from the circulation Most ligands are metabolized in the body (liver). This
by uptake in all tissues, by clearance from the body (e.g., means that for a radiotracer at least one of the metabolic
through the kidneys), and by metabolism (of the tracer itself). products (metabolites) will be radioactive. These radioactive
This means that the arterial input can vary from subject to metabolites will also circulate through the body and could
subject, between patient groups, and within one subject as a compromise the signal obtained with the parent radiotracer.
For brain studies it is therefore good practice to use only (tissue) as a function of time. Each time-activity curve is a
tracers for which the radioactive metabolites do not cross the direct result of (the history of) the arterial input function. A
blood-brain barrier. As the brain signal is then due to the par- tracer kinetic model is needed to extract relevant quantitative
ent tracer alone, it is logical that the arterial plasma curve biological, physiological, or molecular information from the
needs to be corrected for radioactive metabolites to extract measured tissue tracer concentrations.
the true arterial input function Such a model is a mathematical description of the fate of
A metabolite-corrected arterial plasma input function can the tracer in the tissue under study. In practice all tracer
be derived from an online measured arterial whole blood kinetic models are compartmental models in which the uptake
curve by using information from a limited number of manual of a tracer is distributed over a (limited) number of discrete
samples withdrawn during the scan at times where changes compartments [16]. For most applications single-tissue or
in the curve are not too rapid (i.e., after at least 5 min). Both two-tissue compartmental models suffice. It should be noted
whole blood and plasma radioactivity concentrations of that the various compartments do not necessarily reflect phys-
these samples need to be measured together with metabolite ical compartments. For example, nonspecific and specific
fractions. The whole blood concentrations can be used to bindings are different compartments in a compartmental
calibrate the curve obtained from the online samples. The model although they are spread out over the same physical
plasma-to-whole blood radioactivity ratios of these samples space. The main purpose of the compartments is that they
are fitted to a mathematical function, usually an exponential. provide a convenient way to describe tracer kinetics.
The whole blood curve should then be multiplied with this
mathematical function to generate a (total) plasma curve as
function of time. The measured plasma metabolite fractions 10.6 Compartmental Models
are then used to derive the parent fraction for each sample,
which is the fraction of total radioactivity in plasma that is 10.6.1 Single-Tissue Compartmental Model
due to the originally injected tracer. This fraction over time
should also be fitted to a mathematical function, often a so- As indicated by the name, in a single-tissue compartmental
called Hill function [15]. Finally, the total plasma curve model, tissue is represented by just a single tissue. This is a
derived above should be multiplied with this mathematical valid model for a perfusion tracer in which the tracer enters
function, resulting in the final metabolite-corrected arterial and clears the tissue in a flow-dependent manner without
plasma input function. having any significant molecular interactions (such as bind-
ing to a target) within that tissue. It can also be used for mea-
suring uptake in tissue, including binding, if interaction
10.5.2 Tissue Uptake within the tissue cannot be clearly separated.
The basic differential equation of a single-tissue compart-
The acquired PET data results in images displaying the distri- mental model, illustrated in Fig. 10.1, is given by:
bution of radioactivity concentrations (uptake) in organs.
Ideally, uptake is associated with a single biological process, dCT ( t ) / dt = K1 × CP ( t ) - k2 × CT ( t ) (10.1)

but in reality it is not. In fact, it is the result of various processes,
some of which may interact with each other. At any given time where CT and CP are concentrations as function of time t for
after injection, total uptake is the sum of intravascular activity, tissue and metabolite-corrected arterial plasma, respectively,
free tracer in tissue (not bound to anything), nonspecific and K1 and k2 represent rate constants for influx into and
binding (i.e., bound to tissue proteins), and specific binding efflux out of the tissue, respectively. In other words, the rate
(bound to the target of interest), all of which change over time.
The problem with a single measurement is that one can only
Plasma Tissue
guess how large the contribution of each signal is. For example,
a highly vascularized tumor will have a larger contribution K1
from intravascular activity than a more necrotic tumor, but
separation of vascular and tumor contributions cannot be CP CT
derived from a single image. In addition, over time, the specific
k2
signal may increase at the cost of the free signal, and again metabolites
kinetics of that process cannot be derived from a single
measurement. Therefore, dynamic scanning is required for
VB PET
fully quantitative PET studies.
By projecting VOIs onto a dynamic scan, tissue time-
Fig. 10.1 Schematic diagram of the single-tissue compartmental
(radio)activity curves can be extracted reflecting uptake, model. CP and CT represent arterial plasma and tissue concentrations, K1
retention, and clearance in those VOIs. Even such a time- and k2 represent influx and efflux rate constants, and VB represents
activity curve, however, only represents total uptake in a VOI blood volume within the PET region
of change of the tissue concentration is the difference the tissue time-activity curves with both vascular curves as
between influx and efflux. input functions until the fitted parameters K1, k2, and VB
Note that if radiolabeled metabolites would also enter the converge).
tissue, another input function (i.e., representing radiolabeled The fitted rate constants K1 and k2 also allow for calcula-
metabolites in plasma) and additional rate constants describ- tion of an important characteristic of tissue, the volume of
ing rates of exchange of those metabolites would be needed, distribution or partition coefficient, which is the ratio of tis-
seriously complicating the overall equation. Therefore, here sue and plasma concentrations at equilibrium. The volume of
it is assumed that radiolabeled metabolites do not enter the distribution VT is defined as:
tissue, but in practice this needs to be validated for each
VT = CT / CP (10.5)
tracer.
Equation (10.1) describes tissue uptake and clearance. In where CT and CP now represent equilibrium (independent
contrast to in vitro measurements where tissues can be of time t) tissue and arterial plasma concentrations, respec-
washed, in vivo PET measurements do not allow for separat- tively. It essentially indicates what tissue concentration
ing tissue from blood. Measured VOI concentrations are would be achieved in case of a constant arterial plasma
based on activity in both tissue and blood (vessels). To concentration. As mentioned above, VT can also be calcu-
account for this, Eq. (10.1) has to be extended as follows: lated from K1 and k2, i.e., from a study in which no equilib-
rium is reached. This can be derived by imposing
CPET ( t ) = (1 - VB ) × CT ( t ) + VB × CA ( t ) (10.2) equilibrium conditions onto Eq. (10.1). In that case, there

will be no change in any concentration, and the left hand of
where CPET is the concentration as measured by the PET Eq. (10.1) will be 0:
scanner, CA the concentration in arterial whole blood (this
time not corrected for metabolites, as the metabolites will dCT ( t ) / dt = K1 × CP ( t ) - k2 × CT ( t ) = 0 (10.6)

also circulate in the blood), and VB the fractional blood
volume. It then follows that:
The solution of Eq. (10.1) is given by:
VT = K1 / k2 (10.7)

CT ( t ) = K1 × CP ( t ) Ä exp ( -k2 × t ) (10.3)

where ⊗ represents a convolution operation. In other words,
CT at a given time is not directly dependent on CP at that 10.6.2 Two-Tissue Compartmental Model
time, but on the history of CP. By substituting Eq. (10.3) into
Eq. (10.2), the following solution for the measured PET con- If the tracer undergoes (kinetically measurable) interactions
centration is obtained: within tissue, more compartments need to be added to prop-
erly describe tracer kinetics. The most used model is a two-
CPET ( t ) = (1 - VB ) × K1 × CP ( t ) Ä exp ( -k2 × t ) + VB × CA ( t )
tissue compartmental model, which is shown in Fig. 10.2. This
(10.4)
is the typical model for, among others, receptor studies. A
in which the measured PET concentration is a function of good ligand will bind only to a single type of receptor, but the
two (measured) input functions, i.e., the metabolite-corrected PET signal is still the sum of free ligand, nonspecifically
arterial plasma curve and the non-corrected arterial whole bound ligand (i.e., binding to various proteins), and ligand
blood curve. In addition, the equation contains the parame- specifically bound to the receptor. In theory this even requires
ters K1, k2, and VB, characteristic of the tissue under study, a three-tissue compartmental model, but in practice the free
that can be obtained by nonlinear regression analysis (fitting and nonspecific compartments are lumped together into a
Fig. 10.2 Schematic diagram Plasma Tissue

of the two-tissue
compartmental model. CP,
CND, and CS represent arterial
plasma, and non-displaceable K1 k3
and specific tissue
concentrations, respectively. CP CND CS
K1 to k4 are rate constants for
the transport between k2 k4
metabolites
compartments, and VB is the
blood volume within the PET
region VB
PET
single non-displaceable compartment, as it is assumed that related to pharmacological parameters in the following
kinetics between free and nonspecific compartments are very way [19]:
fast.
k3 = f ND × kon × Bavail (10.12)
For a two-tissue compartmental model, the measured PET
concentration is still given by Eq. (10.2). The difference with k4 = koff (10.13)
the single-tissue compartmental model is the expression of CT,
which for the two-tissue compartmental model is given by: where fND represents the free fraction in the non-displaceable
compartment, Bavail represents the number of available recep-
CT ( t ) = CND ( t ) + CS ( t ) (10.8) tors, and kon and koff represent the rate constants for associa-

tion and dissociation of the ligand-receptor complex,
where CND(t) and CS(t) represent concentrations in non- respectively. It is tempting to use the pharmacological
displaceable and specific compartments, respectively, both parameter Bmax (maximum number of receptors) in Eq.
as function of time. The kinetics of these two compartments (10.12), but this is not correct as some receptors might be
is given by the following differential equations: occupied by endogenous ligands, hence the use of Bavail. In
pharmacology, the equilibrium dissociation constant Kd is

defined by:
dCND ( t ) / dt = K1 × CP ( t ) - k2 × CND ( t ) - k3 × CND ( t ) + k4 × CS ( t )
(10.9) K d = koff / kon (10.14)
Combining Eq. (10.11) with Eq. (10.14) results in:
dCS ( t ) / dt = k3 × CND ( t ) - k4 × CS ( t ) (10.10)
BPND = f ND × Bavail / K d (10.15)

As for the single-tissue compartmental model, the solu- Assuming that fND is constant, Eq. (10.15) illustrates that
tions of both Eqs. (10.9) and (10.10) contain convolutions BPND is related both to the number of available receptors
similar to the one shown in Eq. (10.3). These solutions can (Bavail) and to the affinity of the ligand for the receptor (Kd).
be substituted in Eq. (10.8), which in turn can be substituted Reliability of estimated BPND values can be poor, espe-
in Eq. (10.2). The final nonlinear equation, now containing cially when kinetics between both tissue compartments are
two convolutions [17], can again be fitted using nonlinear fast, making it difficult to distinguish them from each other.
regression analysis, resulting in values for the four rate con- In that case, a better outcome measure will be VT, which
stants and VB. tends to have higher precision.
Fitting five parameters (K1, k2, k3, k4, VB) from a single VT is defined by Eq. (10.5) and, at equilibrium, both
time-activity curve is a difficult task, especially since nonlin- dCND(t)/dt and dCS(t)/dt are 0. From Eq. (10.10) it then fol-
ear regression is sensitive to noise. In addition, there can be lows that:
high correlation between parameters, especially between k2
and k3 (both representing efflux from the first compartment), CS = ( k3 / k4 ) × CND (10.16)
which means that the precision of fitted rate constants, also
called micro-parameters, can be low. The rapidly increasing where CS and CND are now independent of time. From Eqs.
mathematical complexity with additional compartments (10.5), (10.8), and (10.16), it follows that:
(each compartment adds another convolution) means that
attempts of using a three-tissue compartmental model have V = CT / CP = ( CND + CS ) / CP = (1 + k3 / k4 ) × CND / CP
been rare and, in general, unsuccessful. T
(10.17)
Even if the precision of parameters derived from the two-
tissue compartmental model is low, it is still possible to Next, from:
obtain valuable tissue information by using combinations of
parameters, so-called macro-parameters. The most interest- dCND ( t ) / dt + dCS ( t ) / dt = K1 × CP - k2 × CND = 0 (10.18)
ing macro-parameter for receptor studies is the non-
displaceable binding potential BPND (more precisely it is the it follows that:
binding potential relative to the non-displaceable
CND = ( K1 / k2 ) × CP (10.19)
compartment), which, for tracer experiments, is given by
[18]:
Finally, substituting Eq. (10.19) into Eq. (10.17) results in:
BPND = k3 / k4 (10.11)

V = ( K1 / k2 ) × (1 + k3 / k4 ) = ( K1 / k2 ) × (1 + BPND ) (10.20)
BPND is of special interest because of its direct relation- T
ship with parameters known in pharmacology. k3 and k4 are
which shows the relationship between VT and BPND. It can be Plasma Tissue
seen that VT does not only reflect specific binding (BPND) but
that it also contains a non-displaceable component (K1/k2),
K1 k3
which is primarily due to nonspecific binding. Target
ND S Tissue
In the brain BPND can still be calculated from VT if another CT(t)
region exists with a similar level of nonspecific binding, but k2 k4
without specific binding, i.e., a region that is devoid of the
receptor being targeted. The volume of distribution VT′ of
such a reference region is given by: K1′
Reference
¢ ¢ ¢
V = K / k (10.21)
T 1 2 ND Tissue
CR(t)
where K1′ and k2′ are the rate constants for the reference tissue. k2′
If the blood-brain barrier is symmetric (i.e., increased
transport into the brain is matched by a similar increase in rate
of transport out of the brain) and the level of nonspecific bind- Fig. 10.3 Schematic diagram of the full reference tissue model. ND
and S represent non-displaceable and specific compartments, respec-
ing is constant across the brain, the following equality holds:
tively. In the solution of this model, the target tissue concentration CT is
expressed in terms of the reference tissue concentration CR with four fit
K1 / k2 = K1¢ / k2¢ (10.22) parameters (R1 = K1/K1′, k2, k3, BPND)

Combining Eq. (10.20) with Eq. (10.21) results in [20]:

where CR(t) represents the tissue concentration as function of
BP = (VT - K1 / k2 ) / ( K1 / k2 ) = (VT - V ) / V = VT / V - 1 time in the reference tissue.
¢ ¢ ¢
ND T T T
By combining Eq. (10.24) with Eqs. (10.8)–(10.10), it is
(10.23)
possible to express CT(t) as function of CR(t) rather than
In general, this estimation of BPND is more stable than CP(t), resulting in an operational equation that enables fitting
direct fitting of BPND, but it can only be performed if a refer- for R1 (=K1/K1′), k2, k3, and BPND [20]. As target tissue time-
ence tissue devoid of receptors does exist. In addition, the activity curves are now expressed relative to reference tissue
calculation is only valid if the blood-brain barrier is intact time-activity curves, it is no longer possible to fit for K1, but
(assumption of constant K1/k2 over the brain). rather for its value (R1) relative to K1′.
The main limitation of the full reference tissue model is
correlation between parameters, especially between k2 and
10.6.3 Reference Tissue Models k3, even more so than for the two-tissue plasma input com-
partmental model. As a result, convergence tends to be slow,
The last paragraph of the previous section describes a means and care should be taken to avoid solutions that correspond
to calculate BPND indirectly from VT values derived from a to local minima. This typically involves multiple fits using
two-tissue compartmental model, provided a reference tissue different starting values in order to find the global minimum.
exists that does not express the receptor being studied. In Due to these limitations, use of the full reference tissue
general, this method works very well, but it should be real- model has been limited.
ized that an arterial plasma input function is needed to fit for To improve fitting stability and speed, the simplified
VT. Apart from being invasive, measuring a metabolite- reference tissue model (SRTM) was developed (Fig. 10.4).
corrected arterial plasma input function is labor intensive. This model assumes that the exchange between non-dis-
If a reference tissue exists, a better alternative to derive placeable and specific compartments is fast enough, so
BPND is the use of a so-called reference tissue model. that the net effect of BPND can be represented by an appar-
Figure 10.3 shows a schematic diagram of the full reference ent change in clearance from the tissue (i.e., k2). The
tissue model, which is directly related to the two-tissue resulting differential equation for the target tissue is
plasma input model. Kinetics of the target tissue are still given by [21]:
described by Eqs. (10.9) and (10.10). For the reference tis-
sue, however, the specific compartment does not exist, and dCT ( t ) / dt = K1 × CP ( t ) - k2 × CT ( t ) / (1 + BPND ) (10.25)

kinetics are described by a single-tissue compartmental
model, analogous to Eq. (10.1): while kinetics of the reference tissue are still given by Eq. (10.24).
The net result of this simplification is that only three
dCR ( t ) / dt = K1¢ × Cp ( t ) - k2¢ × CR ( t ) (10.24) parameters (R1, k2, BPND) need to be fitted rather than the

four needed for the full reference tissue model. More results in a much simpler operational equation, containing
importantly, this reduction in the number of parameters also only one rather than two convolutions [21]:
CT ( t ) = R1 × CR ( t ) + k2 × {1 - R1 / (1 + BPND )} × CR ( t ) Ä exp {-k2 × t / (1 + BPND )} (10.26)
Plasma Tissue
• Fitting tissue data requires two input functions, a
metabolite-corrected arterial plasma input function
and a non-corrected arterial whole blood input
K1 function.
Target
Tissue • A tracer kinetic (compartmental) model is needed
ND+S CT(t) to extract specific biological information from tis-
sue time-activity curves.
k2/(1+BPND) • The number of compartments depends on the
kinetic behavior of the tracer, but usually two-tissue
compartments are sufficient.
• Binding potential relative to the non-displaceable
compartment BPND is the best parameter to charac-
terize reversible binding to a target.
K1′ • If BPND cannot be determined reliably, volume of
Reference distribution VT can be used, but it should be realized
Tissue
ND that VT contains a nonspecific component.
CR(t)
• A reference tissue model can be used if (1) a region
k2′ exists that is devoid of the binding site under study
and (2) the level of non-displaceable uptake in tar-
get and reference tissues is the same.
Fig. 10.4 Schematic diagram of the simplified reference tissue model.

In this model, the non-displaceable binding potential BPND results in a
reduced apparent k2 for the target tissue, after non-displaceable (ND) 10.7 Practical Modelling Issues
and specific (S) compartments are lumped together. In the solution of
this model, the target tissue concentration CT is expressed in terms of
the reference tissue concentration CR with three fit parameters 10.7.1 Weighting Factors
(R1 = K1/K1′, k2, BPND)
Quantification of (dynamic) PET studies requires fitting of
This three-parameter model is much faster than the origi- tissue data using plasma or reference tissue input functions,
nal four-parameter model, and, in general, results are very resulting in estimates of various parameters (rate constants,
stable. More recently it has been shown that SRTM might fractional blood volume). In fitting a tissue time-activity
also be valid even if kinetics of specific binding are not as curve, it should be realized that the precision of successive
fast as expected in the underlying assumptions. Nevertheless, data points is not the same. Data are expressed as concentra-
use of SRTM and also of FRTM should always be validated tions, but typically early frames are short (e.g., 5–10 s),
against the optimal plasma input model. In particular, in both while late frames can be as long as 10 min. In other words
FRTM and SRTM (and any method that relies on a reference the number of acquired counts is much higher in the later
tissue), it is assumed that K1/k2 is constant over the brain, an frames, and, for similar concentrations, these later frames
assumption that is likely to be violated when the blood-brain will have higher precision. On the other hand, especially for
barrier is damaged [22]. short physical half-lives, the tracer will decay, and if data are
decay corrected, later frames will have lower precision. To
account for both effects during fitting, each point of a time-
Key Learning Points activity curve should be weighted according to its precision.
• An arterial input function should be acquired using Several weighting schemes have been proposed, and it
an automatic online blood sampler. seems that all of these schemes are better than not weighting
• Additional manual samples are needed to derive a at all (i.e., giving the same weight to all points), basically
metabolite-corrected arterial plasma input function. because all take into account differences in precision
between data points [23]. The most logical scheme is based
on total (true) counts per frame, and this is the scheme and, therefore, it is essential to fit data to the model that gives
presented here. the best description of the kinetics of the tracer. This is not
Suppose A total counts (not corrected for decay) have necessarily the model with the highest number of parame-
been acquired in a frame of duration L. Then the total count ters, as at some stage adding more parameters will not change
rate R for that frame equals A/L. Total counts will be distrib- the quality of the fit significantly. The main problem of too
uted according to Poisson statistics, so: many parameters is that these parameters cannot be fitted
with any degree of precision, i.e., the compartment they are
Var ( A ) = A (10.27) associated with is not identifiable. Therefore, it is important

to identify the optimal model balancing the number of addi-
SD ( A ) = A (10.28) tional parameters against the improvement in sum of squares

associated with the fit.
COV ( A ) = A / A = 1/ A (10.29) There are several model selection criteria that can be used

to select the optimal model for a certain tracer (and applica-
where Var represents variance, SD standard deviation, and tion). The most widely used one is the Akaike information
COV coefficient of variation. Assuming no error in frame criterion (AIC), which is given by [24]:
duration L, the following equations can be derived:
AIC = N × ln ( SS) + 2 P (10.37)
COV ( R ) = 1/ A (10.30)

where N is the number frames, P the number of parameters,
SD ( R ) = R / A (10.31) and SS the residual sum of squares.

The model that produces fits with the lowest AIC value is
Var ( R ) = R 2 / A (10.32) then considered to be the preferred (best) model. Equation

(10.37) demonstrates that additional parameters are penal-
Weight ( R ) = A / R 2 = L / R = L2 / A (10.33) ized, as they will increase the AIC value. In other words,

additional parameters are only considered if they result in a
It should be noted that these weights apply to non-decay- sufficient decrease in the sum of squares of the fit.
corrected data. It is common, however, to work with time- It is good practice to use a few additional model selection
activity curves that are corrected for decay, so that these criteria to check consistency in model selection, i.e., to make
curves only reflect biological processes. In case of decay cor- sure that indeed the optimal model is selected for further
rection, R is not equal to A/L, but needs to be modified to analysis of the data. Other selection criteria are, for example,
account for physical decay: the Schwarz criterion and the F-test. The Schwarz criterion
is given by [25]:
R = f × A / L (10.34)

where f is the decay correction factor for the frame. This fac- SC = N × ln ( SS) + P × ln ( N ) (10.38)
tor is given by:
Again, the model that produces the lowest SC value is con-
f = l × A / {exp ( -l × Ts ) - exp ( -l × Te )} (10.35) sidered to be the optimal model. The F-test is given by [26]:

where Ts and Te are start and end times of the frame (note that F = {( SS1 - SS2 ) / ( P2 - P1 )} / {SS2 / ( N - P2 )} (10.39)

L = Te − Ts) and λ is the decay constant of the radionuclide
used. The corresponding weighting factors for decay- where subscripts 1 and 2 represent fits with the lowest and
corrected data are then given by: highest number of parameters, respectively. Unfortunately,
evaluating results is more cumbersome than for Akaike and
Weight ( R ) = L / ( f × A )
2 2
(10.36) Schwarz tests, as an F-statistic table is required to assess
significance.
10.7.2 Model Selection Criteria Key Learning Points

• When fitting tissue time-activity curves, individual
In Sect. 10.6 a number of compartmental models have been data points should be weighted to account for dif-
described. When analyzing patient data, however, it is impor- ferences in precision between data points.
tant to stick to one model, as each model will have some bias, • Identification of the optimal model should be based
but this bias will be different for different models. on an objective model selection criterion.
Nevertheless, bias should be limited as much as possible,
10.8 Parametric Imaging Ki reflects the net uptake in the irreversible compartment.
K1 is the rate of transfer from plasma to the non-displaceable
10.8.1 Linearizations compartment (or in the case of [18F]FDG the free compart-
ment). Once a molecule is in this first compartment, the
The optimal tracer kinetic model is defined using nonlinear chance of moving on to the second compartment is
regression analysis of time-activity curves derived from pre- k3/(k2 + k3), and so the rate of net uptake in the second
defined volumes of interest. As mentioned earlier, such a (irreversible) compartment is this ratio times K1. Note that
time-activity curve can contain contributions from several for [18F]FDG, Ki is directly proportional to glucose
sources (compartments), e.g., free, nonspecifically, and spe- metabolism, explaining why the method became well known.
cifically bound tracer. If only part of the VOI is abnormal, the Using the Patlak plot on a voxel-by-voxel level, it is possible
resulting time-activity curve will be a mixture of normal and to generate images of glucose metabolism rather than [18F]
abnormal tissue. Apart from problems associated with tissue FDG uptake images.
heterogeneity, this “dilution” with normal tissue may mean Many years later a comparable linearization was devel-
that the disease-related abnormality will not be detected. If oped for reversible tracers. In this Logan plot [28], ∫CT/CT is
the abnormality is related to one process only, for example, plotted against ∫CP/CT, which, after an equilibration time and
specific imaging, while other processes (in this case free and ignoring intravascular activity, results in a straight line with
nonspecific binding) are unchanged, it is questionable a slope that represents VT:
whether an abnormality in the original radioactivity images
will be picked up. Clearly, it is preferable to perform tracer
ò CT / CT = VT × ò CP / CT + K (10.42)
kinetic modelling at the voxel level in order to fully utilize
the resolution of the scanner. Unfortunately, the one disad- where K is a constant that depends on k2, k3, and k4.
vantage of nonlinear regression is its sensitivity to noise. As Later the method was adapted to allow for a reference tis-
a result, fitting at the voxel level would result in parametric sue input [29]. With respect to reference tissue models, more
images (i.e., images that directly represent the distribution of methods have been proposed, in particular various versions
a certain parameter) with very high noise levels, and, conse- using multi-linear regression analysis [30, 31].
quently, it will still not be possible to detect subtle regional All parametric methods mentioned above are based on a
changes in such a parametric image. In addition, because of linearization of the operational equation with subsequent lin-
the high noise levels, fitting will be slow, as it will take time ear regression analysis of the data (fitting a straight line) to
to obtain convergence (for each voxel). obtain desired outcome measures. The major advantage of
Nevertheless various parametric imaging methods have this approach is that linear regression does not lead to noise
been developed successfully. To avoid noise amplification, amplification, resulting in high-quality images of the param-
they are all based on some form of linearization of the origi- eter of interest. A disadvantage is that bias may be intro-
nal nonlinear operational equations. duced, as fitted parameters are not independent of each other.
For example, CP occurs in both the left- and right-hand sides
of Eq. (10.40). In addition, because of interdependency of X
10.8.2 Patlak and Logan Plots and Y parameters, it is more difficult to calculate uncertain-
ties in the estimated parameters. Finally, only macro-
The first (and best known) linearized method is the Patlak parameters (e.g., Ki or VT) can be obtained.
plot [27], which originally was developed for analysis of
[18F]FDG data but which can be used for any tracer with irre-
versible uptake (for a two-tissue compartmental model, it 10.8.3 Basis Function Methods
means that k4 = 0), providing the net rate of influx Ki from
plasma into the irreversible compartment. It can be shown Another, more general, approach is the basis function
that, after an equilibration period, the plot of CT/CP against method, which in theory can be used for all compartmental
∫CP/CP becomes a straight line [27]: models. The method has indeed been used for plasma input
models [32], but is best known for its implementation of
CT / CP = K i × ò CP / CP + Vi (10.40) SRTM, known as RPM [33]. Taking SRTM as an example, it
can be seen that its operational equation, i.e., Eq. (10.26),
where Vi is the initial volume of distribution (i.e., volume of contains both a linear and a convolution term. In the basis
distribution of free tracer) and, for a two-tissue compartmen- function method, prior to fitting, first the convolution term is
tal model, Ki is given by: calculated for a series (usually 50–100) of so-called basis
functions that cover the entire range of physiological values
K i = K1 × k3 / ( k2 + k3 ) (10.41)
for the parameters involved in the convolution. Consequently,
for each basis function, Eq. (10.26) becomes linear, and this (drug) studies, the intervention itself may affect body uptake
linear expression can be fitted by linear regression. Next, lin- and clearance, so the arterial input function can be signifi-
ear regression is performed for all basis functions. Finally, cantly different (especially in shape) before and after the
parameter values are obtained from the fit with the basis intervention, even in the same subject [35].
function that results in the lowest sum of squares. In SUVr studies, SUV of the target tissue is normalized to
RPM provides parameter estimates of R1, k2, and BPND for SUV of a reference region, and differences in arterial plasma
each voxel. From Eq. (10.22) it follows that: input function will have a smaller effect on the results, as they
affect both target and reference tissues equally. However, SUVr
R1 = K1 / K1¢ = k2 / k2¢ (10.43) applications assume that there is some kind of equilibrium

between both tissues. This may not be the case if flow is affected
This can be rearranged to: in one or both of the regions, as equilibrium may be slower than
assumed when validating the method [36]. In addition, the rate
k2¢ = k2 / R1 (10.44) of clearance from the tissues may affect the bias [37].
Semiquantitative methods are attractive, as they are based
It will be clear that each set of R1 and k2 values will provide on simple scanning protocols (static scans). However, just as
another k2′ value. However, k2′ should be a constant as it rep- for the parametric imaging methods listed above, they should
resents k2 of the reference tissue. This provides a means to be validated for each new application and even for the same
reduce the number of fit parameters from 3 to 2 [34], thereby application in a new patient group. In addition, simulations
further improving stability. After running RPM with three should be performed to characterize their sensitivity to pos-
parameters, k2′ is fixed to the median value (not mean, as this sible violations of the underlying simulations. Finally,
would be sensitive to outliers), and subsequently a second run because of intersubject variability in kinetic parameters that
is performed with only two parameters (R1 and BPND). cannot be measured using SUV or SUVr themselves (which
are based on total uptake), test-retest variability can be sub-
stantially higher than for fully quantitative methods. This
10.8.4 Validation means that in some applications (many) more patients need
to be included in a trial [38]. Especially in drug trials, where
As mentioned above, all parametric methods are based on changes in delivery and/or flow may occur, semiquantitative
some form of linearization. This also means that each para- methods should only be used with the utmost care.
metric method essentially is an approximation of the full
compartmental model and that it may show nonlinear sensi-
tivity to noise. Interestingly, it seems that no single paramet- Key Learning Points
ric method is ideal for all tracers. In other words, for every • Parametric images are based on linearizations of
new tracer (and application), results of all methods need to underlying compartmental models.
be compared at a regional level with those of the optimal • Parametric imaging methods are fast (linear
compartmental model to identify the best method for para- regression) and do not amplify noise, but they
metric imaging of that tracer. should be validated as bias may occur.
• The best parametric imaging method is
tracer-dependent.
10.8.5 Semi-Quantification • Although semiquantitative methods such as SUV
and SUVr are attractive, given that they are based on
Many studies that claim to be quantitative use in fact semi- simple scanning protocols, they should be validated
quantitative methods, such as tumor-to-blood ratios, stan- even more vigorously, as substantial bias may occur.
dardized uptake values (SUV), or, in the brain, SUV ratios • SUV and SUVr should be avoided in longitudinal
(SUVr). studies where changes in tracer delivery can be
SUV represents uptake in a volume of interest (or voxel) expected.
normalized to injected activity and body weight (alterna-
tively, lean body mass and body mass index have also been
used). This approach basically assumes that delivery to a tis-
sue (arterial input function) is directly related to the injected 10.9 Conclusions
activity, i.e., that not only the height but also the shape of the
input function is predictable. For some tracers this may be Although PET is an intrinsically quantitative imaging
valid, but for others there may be significant intersubject modality, true quantification requires attention to detail in
variation in plasma clearance. In addition, in intervention data acquisition, data processing, and data analysis. For
fully quantitative studies, dynamic scanning is required coincidence measurements and coincidence efficiency calibration.
J Comput Assist Tomogr. 1986;10:845–50.
together with compartmental analysis of the data, in first 4. Watson CC. New, faster, image-based scatter correction for 3D
instance using an arterial plasma input function. Only after PET. IEEE Trans Nucl Sci. 2000;47:1587–94.
proper validation, i.e., comparison with full quantitative 5. Boellaard R, Van Lingen A, Lammertsma AA. Experimental and
analyses and appropriate simulations, can other methods clinical evaluation of iterative reconstruction (OSEM) in dynamic
PET: quantitative characteristics and effects on kinetic modeling. J
such as reference tissue models and parametric imaging Nucl Med. 2001;42:808–17.
methods be used. 6. Kinahan PE, Hasegawa BH, Beyer T. X-ray-based attenuation cor-
Simplified scanning protocols and data analysis methods rection for positron emission tomography/computed tomography
should only be used when tracer kinetics are fully under- scanners. Semin Nucl Med. 2003;33:166–79.
7. Beyer T, Lassen ML, Boellaard R, Delso G, Yaqub M, Sattler B,
stood. This applies to new tracers but also to new applica- et al. Investigating the state-of-the-art in whole-body MR-based
tions (different patient population or different intervention) attenuation correction: an intra-individual, inter-system, inventory
using an existing tracer. Otherwise it is possible that the sim- study on three clinical PET/MR systems. MAGMA. 2016;29:75–87.
plification is invalid for addressing the specific research or 8. Van den Heuvel OA, Boellaard R, Veltman DJ, Mesina C,
Lammertsma AA. Attenuation correction of PET activation
clinical question. When fully quantitative studies have been studies in the presence of task-related motion. Neuroimage.
performed, it is always possible to retrospectively apply a 2003;19:1501–9.
(validated) simplified analysis to the data, so that existing 9. Svarer C, Madsen K, Hasselbalch SG, Pinborh LH, Haugbøl S,
data are not lost. In contrast, when using a simplified method Frøkjær VG, et al. MR-based automatic delineation of volumes
of interest in human brain PET images using probability maps.
that later appears to be invalid implies that all data acquired Neuroimage. 2005;24:969–79.
with that method will be lost. Moreover, even progress in the 10. Hoetjes NJ, van Velden FHP, Hoekstra OS, Hoekstra CJ, Krak NC,
entire field may be delayed, as others will base their research Lammertsma AA, Boellaard R. Partial volume correction strate-
on incorrect results from those studies. gies for quantitative FDG PET in oncology. Eur J Nucl Med Mol
Imaging. 2010;37:1679–87.
Clearly, the level of quantification depends on the clinical 11. Greve DN, Salat DH, Bowen SL, Izquierdo-Garcia D, Schultz
or research question at hand. As mentioned above, for stag- AP, Catana C, et al. Different partial volume correction methods
ing in oncology even a qualitative approach will often suffice lead to different conclusions: an 18F-FDG-PET study of aging.
(finding metastases). Simplified quantitative and semiquanti- Neuroimage. 2016;132:334–43.
12. Iida H, Kanno I, Takahashi A, Miura S, Murakami M, Takahashi
tative methods can be useful, provided they have been vali- K, et al. Measurement of absolute myocardial blood flow with
dated against fully quantitative methods and are still valid for H215O and dynamic positron-emission tomography. Strategy for
the clinical or research question at hand. In practice, the quantification in relation to the partial-volume effect. Circulation.
challenge is to find the optimal balance between accuracy 1988;78:104–15.
13. Boellaard R, Van Lingen A, Van Balen SCM, Hoving BG,

and precision. Lammertsma AA. Characteristics of a new fully programmable
blood sampling device for monitoring blood radioactivity during
PET. Eur J Nucl Med. 2001;28:81–9.
14. Lammertsma AA, Frackowiak RSJ, Hoffman JM, Huang SC,

Key Learning Points Weinberg IN, Dahlbom M, et al. The C15O2 build-up technique to
• For a new tracer and even for an existing tracer used measure regional cerebral blood flow and volume of distribution of
in a new patient group or with a different interven- water. J Cereb Blood Flow Metab. 1989;9:461–70.
tion, simplified methods should only be used after 15. Gunn RN, Sargent PA, Bench CJ, Rabiner EA, Osman S, Pike
VW, et al. Tracer kinetic modelling of the 5-HT1A receptor ligand
rigorous validation against fully quantitative [carbonyl-11C]WAY-100635 for PET. Neuroimage. 1998;8:426–40.
models. 16. Gunn RN, Gunn SR, Cunningham VJ. Positron emission tomog-
• The level of simplification depends on the clinical raphy compartmental models. J Cereb Blood Flow Metab.
or research question to be addressed, aiming for an 2001;21:635–52.
17. Huang SC, Phelps ME, Hoffman EJ, Sideris K, Selin CJ, Kuhl
optimum trade-off between accuracy and precision. DE. Noninvasive determination of local cerebral metabolic rate of
glucose in man. Am J Physiol. 1980;238:E69–82.
18. Mintun MA, Raichle ME, Kilbourn MR, Wooten GF, Welch MJ. A
quantitative model for the in vivo assessment of drug binding sites
with positron emission tomography. Ann Neurol. 1984;15:217–27.
References 19. Innis RB, Cunningham VJ, Delforge J, Fujita M, Gjedde A, Gunn
RN, et al. Consensus nomenclature for in vivo imaging of reversibly
1. Phelps ME, Hoffman EJ, Mullani NA, Ter-Pogossian MM. binding radioligands. J Cereb Blood Flow Metab. 2007;27:1533–9.
Application of annihilation coincidence detection to transaxial 20. Lammertsma AA, Bench CJ, Hume SP, Osman S, Gunn K, Brooks
reconstruction tomography. J Nucl Med. 1975;16:210–24. DJ, et al. Comparison of methods for analysis of clinical [11C]raclo-
2. Ter-Pogossian MM, Phelps ME, Hoffman EJ, Mullani NA. A posi- pride studies. J Cereb Blood Flow Metab. 1996;16:42–52.
tron emission transaxial tomograph for nuclear medicine imaging 21. Lammertsma AA, Hume SP. Simplified reference tissue model for
(PETT). Radiology. 1975;114:89–98. PET receptor studies. Neuroimage. 1996;4:153–8.
3. Casey ME, Hoffman EJ. Quantitation in positron emission com- 22. Folkersma H, Boellaard R, Vandertop WP, Kloet RW, Lubberink M,
puted tomography: 7. A technique to reduce noise in accidental Lammertsma AA, et al. Reference tissue tissue models and blood
brain barrier disruption: lessons from (R)-[11C]PK11195 in trau- tion to [11C]DASB positron emission tomography studies of the
matic brain injury. J Nucl Med. 2009;50:1975–9. serotonin transporter in human brain. J Cereb Blood Flow Metab.
23.
Yaqub M, Boellaard R, Kropholler MA, Lammertsma 2003;23:1096–112.
AA. Optimization algorithms and weighting factors for analysis of 32. Boellaard R, Knaapen P, Rijbroek A, Luurtsema GJJ, Lammertsma
dynamic PET studies. Phys Med Biol. 2006;51:4217–32. AA. Evaluation of basis function and linear least squares methods
24. Akaike H. A new look at the statistical model identification. IEEE for generating parametric blood flow images using 15O-water and
Trans Automat Contr. 1974;19:716–23. positron emission tomography. Mol Imaging Biol. 2005;7:273–85.
25. Schwarz G. Estimating the dimension of a model. Ann Stat.
33. Gunn RN, Lammertsma AA, Hume SP, Cunningham VJ. Parametric
1978;6:461–4. imaging of ligand-receptor binding in PET using a simplified refer-
26. Cunningham VJ. Non-linear regression techniques in data analysis. ence region model. Neuroimage. 1997;6:279–87.
Med Inform. 1985;10:137–42. 34. Wu Y, Carson RE. Noise reduction in the simplified reference tissue
27. Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation model for neuroreceptor functional imaging. J Cereb Blood Flow
of blood-to-brain transfer constants from multiple-time uptake Metab. 2002;22:1440–52.
data. J Cereb Blood Flow Metab. 1983;3:1–7. 35. Cheebsumon P, Velasquez LM, Hoekstra CJ, Hayes W, Kloet RW,
28. Logan J, Fowler JS, Volkow ND, Wolf AP, Dewey SL, Schlyer Hoetjes NJ, et al. Measuring response to therapy using FDG PET:
DJ, et al. Graphical analysis of reversible radio-ligand binding semi-quantitative and full kinetic analysis. Eur J Nucl Med Mol
from time-activity measurements applied to [N-11C-methyl]-(–)- Imaging. 2011;38:832–42.
cocaine PET studies in human subjects. J Cereb Blood Flow Metab. 36. Van Berckel BNM, Ossenkoppele R, Tolboom N, Yaqub M, Foster-
1990;10:740–7. Dingley JC, Windhorst AD, et al. Longitudinal amyloid imag-
29. Logan J, Fowler JS, Volkow ND, Wang GJ, Ding YS, Alexoff ing using 11C-PiB: methodologic considerations. J Nucl Med.
DL. Distribution volume ratios without blood sampling from 2013;54:1570–6.
graphical analysis of PET data. J Cereb Blood Flow Metab. 37. Carson RE, Channing MA, Blasberg RG, Dunn BB, Cohen RM,
1996;16:834–40. Rice KC, et al. Comparison of bolus and infusion methods for
30. Ichise M, Toyama H, Innis RB, Carson RE. Strategies to improve receptor quantitation: application to [18F]cyclofoxy and positron
neuroreceptor parameter estimation by linear regression analysis. J emission tomography. J Cereb Blood Flow Metab. 1993;13:24–42.
Cereb Blood Flow Metab. 2002;22:1271–81. 38. Lammertsma AA. Forward to the past: the case for quantitative
31. Ichise M, Liow JS, Lu JQ, Takano A, Model K, Toyama H, et al. PET imaging. J Nucl Med. 2017;58:1019–24.
Linearized reference tissue parametric imaging methods: applica-
Principles of Radiation Biology
and Dosimetry for Nuclear Medicine 11
Procedures
Massimo Salvatori, Amedeo Capotosti,
and Luca Indovina
Contents
11.2 Diagnostic Nuclear Medicine 236
11.2.1 Radiation Carcinogenesis 238
11.2.2 Risk–Benefit Ratio 239
11.3 Nuclear Medicine Therapy 240
11.3.1 Overview on Dosimetric Methods 242
11.3.2 Nuclear Medicine Treatments 244
References........ 255
Learning Objectives
• Understand the conventional radiobiological models and 11.1 Introductory Background
the cell death mechanism(s).
• Understand dosimetry to both the target organ and other Medical imaging and therapy using ionizing radiation,
sites of radioactivity localization. including diagnostic nuclear medicine and therapy, must
• Understand the most important purpose of nuclear medi- consider the benefits and the possible risks of the exposure.
cine therapy, that is, to maximize the radiation dose Both diagnostic studies and therapy should be performed
absorbed by the tumor while minimizing radiobiological with the lowest radiation doses [1–3].
damage to normal tissues. The purpose of this chapter is to provide information
• Understand the need for dosimetry to individually option the computational methods in radiation dosimetry,
mize the therapeutic activity to be administered. relationship between radiation dose and biological effects,
risk model for low-dose radiation exposure, justification
and risk–benefit ratio of nuclear medicine procedures
[4–6].
M. Salvatori (*)
Nuclear Medicine Institute, Fondazione Policlinico Universitario The differences in administered activities, radiation doses
A. Gemelli, Rome, Italy and possible risks (i.e., deterministic versus stochastics
Nuclear Medicine Institute, Catholic University of the Sacred effects) of diagnostic and therapeutic studies are discussed
Heart, Rome, Italy in the diagnostic and therapy sections [4].
e-mail: massimo.salvatori@unicatt.it
A. Capotosti
Department of Physics, Catholic University of the Sacred Heart,
Rome, Italy
L. Indovina
Department of Medical Physics, Fondazione Policlinico
Universitario A. Gemelli, Rome, Italy

236 M. Salvatori et al.
11.2 Diagnostic Nuclear Medicine lar patient may deviate significantly from these average val-
ues [15–17].
Nuclear medicine is mostly a diagnostic specialty. Whereas For diagnostic nuclear medicine procedures, the radiation
the broad aim in diagnostic radiology is imaging anatomy, effects of practical concern are stochastic effects, in particu-
nuclear medicine is more closely linked to the investigation lar carcinogenesis [18–22].
of metabolism and pathophysiological processes [7]. The ED is used in radiation protection to compare the sto-
The National Council on Radiation Protection and chastic risk of a nonuniform exposure to ionizing radiation
Measurements (NCRP) Report 160 (“Ionizing Radiation with the risk associated with a uniform whole-body expo-
Exposure of the Population of the United States”), released sure. It is a weighted sum of the doses to the individual tis-
in March 2009, revealed that between 1972 and 2006, the sues/organs of the body, where each tissue weighting factor
annual number of nuclear medicine procedures increased reflects the relative susceptibility of that tissue to stochastic
from about 3.3 million to 18.1 million, an approximately damage [15].
448% increase [8]. EDs for nuclear medicine procedures are usually of the
Report 160 also reported that the myocardial perfusion order of 10 mSv for typical administered activities, meaning
scan has been the most commonly performed nuclear medi- that the overall risk of carcinogenesis is roughly equivalent
cine procedure. Mettler et al. [2] estimated that cardiac stud- to that of a uniform whole-body dose of 1 cGy. Anatomic
ies in the USA in 2006 accounted for almost 60% of the total models and relevant dosimetric quantities such as absorbed
number of nuclear medicine examinations, with an estimated fractions are now available [16], and the radiation burden
number of 10.3 million of examinations and more than 85% of radiopharmaceutical doses have thus been estimated for
of the collective effective dose from nuclear medicine newborns, 1-year-olds, 5-year-olds, 10-year-olds, and
procedures. 15-year-olds as well as adults [16, 23].
The number of cardiac procedures increased from 1% However, since there is a significant difference in doses
of the total in 1973 to 57% in 2005, whereas the percent- among organs for a particular radiopharmaceutical and ana-
ages of brain, gastrointestinal tract, and lung studies mark- tomic model [17], one should use organ-, age-, and gender-
edly decreased [2]. A distant second-place procedure is the specific doses and risk factors, rather than the ED and the
bone scan, but cardiac scans are performed at a rate approx- overall risk factor, to estimate the cancer risk associated with
imately three times greater than bone scans [8]. Report 160 a particular procedure [24–26].
analyzed also the age stratification of nuclear medicine Currently, diagnostic nuclear medicine is predominantly
examinations in 2003, showing that most patients were in performed with hybrid PET/CT or SPECT/CT imaging
40–74 age range [8]. machines. These devices improve diagnostic accuracy by
The radiation doses to patients resulting from administrations allowing better localization and definition of scintigraphic
of radiopharmaceuticals are determined by a variety of physical findings [27]. However, the combined acquisition of func-
and biological factors which include the amount and form of the tional and anatomic images can increase radiation exposure
radioactive material administered, the route of administration, to patients, particularly when using high mAs diagnostic
the biokinetics and physiological fate of the radiopharmaceuti- CT procedures [27, 28].
cal, and the decay scheme of the radionuclide [4]. When CT is performed as part of a SPECT/CT or a PET/CT
Absorbed doses to the various organs and tissues are gen- study, the axial field of view may be much larger than that of
erally estimated using the dosimetric formalism developed conventional CT studies, routinely extending from the base of
by the Medical Internal Radiation Dose (MIRD) Committee the skull to the mid-thigh and thus potentially yielding an ED
of the United States Society of Nuclear Medicine and considerably greater than that of conventional, limited field-of-
Molecular Imaging [9–14]. view CT studies [29]. The dose contribution of these CT scans
The radiation doses associated with administration of can account for a considerable percentage of the dose from the
diagnostic radiopharmaceuticals vary in effective dose (ED) whole examination, ranging between about 20–65% for para-
from 0.3 to 20 mSv, values well below the threshold doses thyroid scintigraphy with 99mTc-sestamibi, about 60–100% for
associated with deterministic effects [15, 16]. Table 11.1 bone scintigraphy with 99mTc-MDP, about 30–60% for soma-
reports a compilation of dose coefficients, and ED for com- tostatin receptor scintigraphy with 111In-pentetreotide, and
mon diagnostic nuclear medicine procedures normalized about 10–25% for scintigraphy with 67Ga-citrate.
to the 70-kg standard man anatomic model [15–17]. As a general rule, low-dose acquisition is suggested when
Average normal tissue doses received by the “standard” recent diagnostic CT data are available and when follow-up
patient are found in package inserts for approved radiophar- studies are to be performed (monitoring of response to treat-
maceuticals and in reports issued by authoritative bodies ment). Moreover, low-dose CT is indicated for further ana-
such as the International Commission on Radiological Units tomic localization or characterization of focal abnormalities
(ICRU), even though the actual doses received by a particu- observed on planar or SPECT images. SPECT-guided diag-
11 Principles of Radiation Biology and Dosimetry for Nuclear Medicine Procedures 237
Table 11.1 Radiaton doses for diagnostic nuclear medicine procedures must be scaled down according to the actually administered activities,
for the 70-kg standard man anatomic model. Since the administered that in several instances are much lower that the activities listed here
activity values listed here are at the upper limit of the recommended (e.g., with modern PET scanners the activity of [18F]FDG administered
activities for each radiopharmaceutical, the radiation dosimetry values for oncology normally do not exceed 370 MBq)
Procedure Radiopharmaceutical Administered activity (MBq) Effective dose cSv/MBq cSv total
Brain 99m
Tc-HMPAO-exametazime 740 0.00093 0.69
Brain 99m
Tc-ECD-neurolite 740 0.00077 0.57
Brain [l8F]FDG 740 0.0019 1.41
Thyroid scan l23
I Na 25 0.0075 0.19
Thyroid scan 99m
Tc-pertechnetate 370 0.0013 0.48
Parathyroid scan 99m
Tc-sestamibi 740 0.0009 0.67
Cardiac stress–rest test 201
Tl-chloride 185 0.022 4.1
Cardiac rest–stress test 99m
Tc-sestamibi 1-day protocol 1100 0.00085 0.94
Cardiac rest–stress test 99m
Tc-sestamibi 2-day protocol 1500 0.00085 1.3
Cardiac rest–stress test Tc-tetrofosmin 1500 0.00076 1.1
Cardiac ventriculography 99m
To-labeled red blood cells 1110 0.0007 0.78
Cardiac [18F]FDG 740 0.0019 1.4
Lung perfusion 99m
Tc-MAA 185 0.0011 0.20
Lung ventilation 133
Xe 740 0.000074 0.050
Lung ventilation 99m
Tc-DTPA 1300 0.00049 0.020
Liver–spleen 99m
Tc-sulfur colloid 222 0.00094 0.21
Biliary tract 99m
Tc-disofenin 185 0.0017 0.31
Gastrointestinal bleeding 99m
Tc-labeled red blood cells 1110 0.0007 0.78
Gastrointestinal emptying 99m
Tc-labeled solids 14.8 0.0024 0.04
Renal 99m
Tc-DTPA 370 0.00049 0.18
Renal 99m
Tc-MAG3 370 0.0007 0.26
Renal 99m
Tc-DMSA 370 0.00088 0.33
Renal 99m
Tc-glucoheptonate 370 0.00054 0.20
Bone 99m
Tc-MDP 1110 0.00057 0.63
Infection/tumor 67
Ga-citrate 150 0.0100 1.5
SST-receptor scintigraphy 111
In-Pentetreotide 222 0.0054 1.2
Infection scintigraphy 99m
Tc-HMPAO-leukocytes 740 0.0011 0.81
Infection scintigraphy 111
In-oxime-leukocytes 18.5 0.0360 0.67
Tumor [l8F]FDG 740 0.0019 1.4
Reproduced from: Salvatori M, Cremonesi M, Indovina L, Chianelli M, Pacilio M, Chiesa C, Zanzonico P. Radiobiology and radiation dosimetry
in nuclear medicine. In: Strauss HW, Mariani G, Volterrani D, Larson M, eds. Nuclear Oncology – From Pathophysiology to Clinical Applications.
New York: Springer; 2017:305–49
nostic CT is suggested when a recent diagnostic CT is not In CT, the dose dependence on acquisition parameters (such
available and when detailed anatomical information is man- as tube potential, tube load, rotation time, beam width, pitch,
datory to address the clinical needs [27, 28]. and reconstructed image thickness) should be considered when
Hulme et al. [30] demonstrated that attenuation correction addressing the need for CT dose optimization [27, 28].
is achievable with ultra-low-dose CT, despite the high CT Accordingly, when the CT component of a PET/CT
image noise, without compromising the quality of the study is used for attenuation correction and anatomic local-
SPECT image. While Buck et al. [31] indicate a dose range ization rather than for radiologic diagnosis, the CT scan
from 2 to 4 mSv for SPECT-guided low-dose CT, Roach parameters should be adjusted to appropriately reduce the
et al. [32] showed that CT scans used for anatomic localiza- CT dose.
tion in the chest or abdomen result in a dose of up to X-ray beams are characterized by the current [expressed
1–2 mSv. Sawyer et al. [33] reported typical values of around in units of milliampere (mA)] and the peak (p) voltage
1.1 mSv for the chest, around 1.3 mSv for abdomen–pelvis, [expressed in or kilovolts (kV) and represented as “kVp”].
and around 0.2 mSv for the head. Montes et al. [34] reported Spiral CT scans are also characterized by the pitch and the
similar values (~1.2 mSv) for chest and abdomen–pelvis and distance of travel of the imaging table per rotation of the
0.6 mSv for the head–neck region. Miller [35] showed even X-ray tube and detector assemblies. The X-ray intensity and
lower radiation exposure for the CT scan of the SPECT/CT therefore the radiation dose are proportional to the mA and
examination: 0.47 mSv for an abdominal nondiagnostic approximately proportional to the square of the kVp value.
localization and attenuation correction scan. The dose is inversely related to the pitch [17].
The two main dose descriptors in CT are the CTDIvol and 11.2.1 Radiation Carcinogenesis
the dose-length product (DLP). These quantities are not only
used to compare different protocols of acquisition, but they Various radio-epidemiological studies have demonstrated
also form the basis for estimating organ and effective doses increasing rates of radiation-induced cancers in humans. The
[16, 17]. largest and most important of these studies is the Life Span
A diagnostic-quality CT (e.g., with “standard” kVp of Study (LSS) performed by observing the survivors of the
140, mAs of 190, and pitch of 1.25) delivers an ED of 22 mSv atomic bombings in Hiroshima and Nagasaki [5]. Follow-up
[29, 36–38]. However, for attenuation correction of an [18] of the atomic bomb survivors has provided detailed knowl-
[18F]FDG PET scan in an adult, a quantitatively accurate but edge of the relationships between radiation risk and absorbed
nondiagnostic CT scan (e.g., with a kVp of 120, mAs of 60, dose, the age at exposure, the age at diagnosis, and other
and pitch of 1.5) delivers an ED of only 6 mSv [29, 36–38]. parameters. The data are largely supported by a multitude of
For pediatric patients a tube current as low as 60 mA can be smaller studies, mostly on groups of individuals exposed for
used (with a kVp of 120 and a pitch of 1.5), and the resulting medical reasons, both for diagnostic and for therapeutic pur-
ED will be only 1 mSv [29, 36–38]. poses [6].
Thus, if the purpose of the CT component of a SPECT/CT There is considerable controversy regarding the risk of
or PET/CT study is attenuation correction and anatomic reg- low levels of radiation (<100 mSv), typical of diagnostic
istration and not radiologic diagnosis, appropriately judi- radiation exposures, since radiation risks evaluated at low-
cious selection of the CT parameters can reduce the total ED dose levels are not based on experimental and epidemiologi-
by over 50% without compromising the diagnostic informa- cal evidence [39, 40].
tion content and quantitative accuracy of the study. Given this lack of evidence, estimates on risk derived
from high doses have been extrapolated down to low-dose
levels by various scientific bodies, including the International
Commission on Radiological Protection (ICRP), the United
Nations Scientific Committee on the Effects of Atomic
Key Learning Points Radiation (UNSCEAR), and the Biological Effects on
• The annual number of nuclear medicine procedures Ionizing Radiation Committee (BEIR).
is increasing throughout the world, with the myo- Estimates of risk per unit of dose have been derived using
cardial perfusion scan and the bone scan estimated the linear no-threshold (LNT) hypothesis, which is based on
as the most commonly performed nuclear medicine the assumptions that any radiation dose—no matter how
procedure. small—may cause an increase in risk and the probability of
• The radiation doses to patients resulting from this increase is proportional to the dose absorbed in the
administration of radiopharmaceuticals are deter- tissue.
mined by the amount and form of the radioactive Although the risk evaluated at low-dose levels is hypo-
material administered, the route of administration, thetical, it is prudent to assume that it exists and that the LNT
the biokinetics and physiological fate of the radio- model represents an upper bound for it. Current radiation
pharmaceutical, and the decay scheme of the protection standards as well as risk assessments are based on
radionuclide. the LNT hypothesis [41].
• The radiation doses associated with diagnostic Choice of the mathematical model depicting the rela-
administration of radiopharmaceuticals vary in tionship between the excess cancer risk and radiation dose
effective dose (ED) between 0.3 and 20 mSv, values for extrapolation of the A-bomb dose-response data from
well below the threshold doses associated with the “high dose” to “low dose” has long been controversial, and
deterministic effects. several mathematical extrapolations have been proposed
• Currently, diagnostic nuclear medicine is predomi- (Fig. 11.1).
nantly performed with hybrid PET/TC or SPECT/ The supralinear model implies that the cancer risk per cSv
CT imaging machines which improve diagnostic is actually higher at low doses than at high doses; there are no
accuracy but can increase radiation exposure to data or biophysical evidence to support such a model. The
patients, particularly when using a hybrid system sublinear or linear-quadratic model, which for many years
with diagnostic CT capabilities. has been the generally accepted dose-response model,
• However, judicious selection of the CT parameters implies that the risk per cSv is lower at lower doses than at
can reduce the total ED by over 50% without com- the higher doses. However the LNT model is now the prevail-
promising the diagnostic information content and ing model for solid tumors, while the linear-quadratic model
quantitative accuracy of the study. is the model of choice for leukemias [19, 20]. The LNT
model implies that any excess radiation dose above back-
Moreover, although stratified by age and gender, the risk

Current Historical
lower limit lower limit factors within each stratum represent population-averaged
of A-bomb of A-bomb values and thus do not account for individual differences in
survivor data: survivor data:
~5 cSv ~100 cSv radiation sensitivity, such as preexisting condition and
genetic susceptibility to radiogenic damage [26]. Brenner
Excess Cancer Incidence
and colleagues [3, 22] point out that the risk factors they
Data
employed to estimate the numbers of radiogenic cancers
associated with CT scanning are based on A-bomb incidence
data at doses as low as 50 mSv—comparable to cumulative
Supra-Linear
Model Current doses actually received by patients undergoing multiple CT
prevailing model procedures [4]—and not model-based extrapolation from
BEIR V/VII
Linear higher dose data.
No-threshold Model
Supra-Linear Previous
(Linear-Quadratic) prevailing model Key Learning Points
Model Pre-BEIR V
• The most important epidemiological study on
0 radiation-induced cancers in humans, the so-called
Radiation Dose
Life Span Study (LSS), showed increasing rates of
Fig. 11.1 Schematic representation of the basic choices for mathemat- radiation-induced tumors in humans.
ical extrapolation model: the supralinear model, the sublinear model
• Dose unit risk estimates were derived using the so-
(i.e., linear quadratic), and the linear no-threshold model (reproduced
with permission from: Salvatori M, Cremonesi M, Indovina L, Chianelli called linear no-threshold (LNT) hypothesis, which
M, Pacilio M, Chiesa C, Zanzonico P. Radiobiology and radiation is based on the hypothesis that any dose of radiation
dosimetry in nuclear medicine. In: Strauss HW, Mariani G, Volterrani can cause an increased risk of developing tumors.
D, Larson M, eds. Nuclear Oncology – From Pathophysiology to
• However, further mathematical models have been
Clinical Applications. New York: Springer; 2017:305–49)
developed including the supralinear model (that
implies that the cancer risk for cSv is actually
ground, no matter how small, carries with it a finite excess higher at low doses than at high doses) and the sub-
risk of cancer. linear or linear-quadratic model (that implies that
In reality, this issue remains controversial. The A-bomb the risk for cSv is lower at lower doses compared to
survivor data for solid tumors [18–21] are consistent with the higher doses).
LNT as well as with the linear-quadratic model, and on the • The data available on survivors of the nuclear bomb
basis of statistical considerations, one cannot distinguish and nuclear fallout for solid tumors are consistent
between the two models. The LNT model is more conserva- with the LNT and the linear-quadratic model, and
tive (i.e., predicts a higher risk per cSv at low doses) than the on the basis of statistical considerations, it is not
sublinear model and is therefore considered “safer” and thus possible to distinguish between the two models.
more appropriate by some authors for radiation protection
purposes.
The BEIR VII age at exposure- and gender-specific risk
factors for solid tumor and leukemia incidence and mortality
show marked differences in radiogenic cancer risk depend- 11.2.2 Risk–Benefit Ratio
ing on age at exposure and gender. There are large differ-
ences depending on the cancer site, with female breast, lung, Predicting the risk of cancer associated with low radiation
colon, and urinary bladder being the most susceptible to doses must also consider the benefit obtained by the medical
radiation-induced cancer [20]. procedure. Unfortunately, the scientific literature only con-
The risk factors incorporate a so-called dose-rate effective siders the radiogenic risks, without considering the clinical
factor of 1.5, meaning that high-dose radiation rate [HDR, benefits.
>~100 mSv/min] has a 50% higher risk factor than low-dose The quantitation of risk without comparable consider-
radiation rate [LDR, <~10 mSv/min] [20]. It is possible, ation of benefit fosters the mistaken perception that the med-
however, that the dose-rate effectiveness factor may actually ical benefit of a diagnostic procedure does not far outweigh
be higher than 1.5 and that the risk factors may therefore radiogenic risk. Risk–benefit analyses should thus quantita-
overestimate the cancer risk associated with the LDRs typi- tively incorporate both the medical benefit and the radio-
cal of diagnostic radiopharmaceuticals. genic risk of diagnostic procedures.
Pat Zanzonico and Michael M. Stabin analyzed and

quantitatively compared the benefits of some diagnostic Key Learning Points
nuclear medicine procedures with the LNT model-derived • Discussion about the risk of cancer associated with
estimates of radiogenic cancer mortality [42]. Risk and ben- low radiation doses must always consider the ben-
efit were evaluated by using the excess risk (ER, number of efit obtained through the medical procedure.
excess fatal cancers higher than the naturally occurring rate) • Radiation risks must be always weighed against the
and the number of lives lost by not performing the proce- clinical benefit.
dures or by performing an alternative invasive procedure, • Many studies analyzed and quantitatively compared
respectively. the benefits of some diagnostic nuclear medicine
van Tinteren et al. [43] compared the management of procedures with the LNT model-derived estimates
suspected non-small cell lung cancer (NSCLC) with and of radiogenic cancer mortality.
without preoperative PET/CT with [18F]FDG. Without [18F] • Risk and benefit were analyzed by using the excess
FDG-PET/CT, 81% of patients underwent thoracotomy; risk and the number of lives lost by not performing
41% of those thoracotomies were futile because of the pro- the procedures or by performing an alternative inva-
gression of disease disclosed at surgery, and there was a sive procedure.
6.5% surgery-related mortality. The preoperative use of • The use of radiation in medicine saves many lives
[18F]FDG-PET/CT reduced the proportion of patients with each year, while the theoretical risks predicted by
NSCLC undergoing thoracotomy to 65%, with only 21% mathematical models (especially LNT) are gener-
being futile and 20% of patients in whom a noncurative ally smaller than orders of magnitude.
operation was avoided. Zanzonico et al. [42] extrapolated
these data to the US lung cancer population and found a net
benefit of 2219 lives saved per year. Considering for a 370-
MBq administered activity of [18F]FDG an ED of 14 mSv, 11.3 Nuclear Medicine Therapy
and therefore a number of radiogenic cancer-related deaths
of 122, the result is a net benefit of 2097 lives saved per The main features of nuclear medicine therapy (NMT) are
year. the locoregional or systemic administration of a radiophar-
Mansel et al. [44] examined the use of sentinel lymph maceutical directed to a specific target and the rate at which
node biopsy versus standard axillary lymph node dissection the radiation dose is delivered to that target [5].
in operable breast cancer. The authors found a net 63% Radiopharmaceuticals have been used as therapeutic
reduction in the incidence of lymphedema and arm sensory agents for over 60 years for the treatment of both malignant
loss, without any significant increase in cancer risk in the and benign diseases but with patterns of practice that vary
515-subject cohort studied. significantly from country to country [7].
Myocardial perfusion imaging is useful for predict- The role of NMT is expanding with the development of novel
ing perioperative cardiac events (myocardial infarction due pharmaceuticals, the emergence of new indications for treat-
to occult coronary artery disease) associated with vascular ment, and improvements in results [7]. To date, NMT is under-
surgery (most commonly, carotid endarterectomy). Goyer utilized, and the numbers of such treatments will likely increase,
et al. [45] estimated that 109 patients per year would avoid, particularly for oncological indications. The total worldwide
perioperative cardiac deaths if they had preoperative myo- annual number of treatments for 1991–1996 was estimated to be
cardical perfusion imaging. Considering a potential risk of about 0.4 million, corresponding to an average frequency of
six excess cancer deaths per year using 99mTc-sestamibi (or 0.065 treatments per 1000 world population [7].
99m
Tc-tetrofosmin) for rest-stress myocardial perfusion According to the principle of very low-dose rate (LDR),
scintigraphy (ED = 12 mSv), the clinical benefit is in NMT the radiation dose is delivered to the tumor cells by
obvious. continuous, but declining, exposure that is a function of the
Although these analyses suffer from several limita- initial uptake and the effective half-life of the radionuclide
tions and in some cases quantitative analysis of the risk– employed for therapy. The average dose rate for NMT is
benefit ratio indicates that a diagnostic imaging procedure typically of the order of 2–8 Gy/day, and the maximum
is not justified, the conclusion is that “overall, the use of absorbed dose may be up to 50 Gy delivered over a period of
radiation in medicine saves hundreds to thousands of many days. This is in marked contrast with the constant high-
lives every year, whereas the theoretical risks predicted dose rate (HDR) of external beam radiation therapy (EBRT)
by the LNT model are typically orders of magnitude and brachytherapy where the doses are delivered typically at
smaller.” [42]. 1–5 Gy/min and 1–5 Gy/h, respectively [5].
To date, most radiobiological models used in NMT are The typical dose rates used for both clinical EBRT and
based on the extrapolation of data obtained following homo- radiobiological studies are of the order of minutes (~1–5 Gy/
geneous exposures to acute single or fractionated doses of min), exposure times not long enough for DNA repair to
EBRT and assume that EBRT and NMT are substantially occur during the irradiation process. As the dose rate
equivalent from the biological point of view. However, declines, the time taken to deliver a given dose increases, and
emerging evidence suggests that the mechanisms by which it is then possible for DNA repair to take place during irra-
cells respond to LDR exposures are fundamentally different diation and for radiosensitivity to gradually decline [5, 46].
from those occurring at HDR with either EBRT or brachy- Survival curves at decreasing LDR become straighter and ulti-
therapy [6, 46]. mately extrapolate the initial slope (α) of the HDR curve, with a
The combination of prolonged response, limited toxicity, sparing effect that would be attributed to the repair of sublethal
and the possibility of multiple treatments suggests that the lesions occurring during the prolonged exposure (Fig. 11.2).
mechanism of action of LDR therapies is different from that At very LDRs (≤2 cGy/min), cell proliferation can occur
observed with HDR exposures. Accordingly, treatment plans, during the irradiation, thus leading to repopulation of the
clinical trials, and outcome assessment should be designed to pool of clonogenic cells. Clearly, based on this assumption,
consider this difference. Current radiobiological research NMT should be noneffective, although this expectation is
points to a way in which a radiopharmaceutical for therapy can not consistent with many clinical observations. Such phe-
be more effectively administered to patients by consider- nomenon, in which decreasing the dose rate results in
ing the biology of the mechanism of action of therapy as well increased cell killing, has been defined “inverse dose-rate
as the physical characteristics of the radionuclide [5]. effect.” The effect has been attributed both to the lower-dose
A number of basic principles, emerged from early studies rate (permitting the cells to progress through the cell cycle
describing the effects of acute single or fractionated doses of into more radiosensitive phases) and to some hypersensitiv-
EBRT on the clonogenic survival of cells, generally held that
a cell must be “hit” by a radiation track in order to be killed, 1
that genomic DNA was the main “target” for killing, and that
double-strand breaks (DSBs)/clustered lesions/multiply
damaged sites were the principal DNA lesions leading to cell
death [46].
Early models focused largely on the mechanical or meta- 0.1
bolic challenges that DSBs might pose to a cell when it
attempts to replicate its DNA (in S-phase) or divide (in mito-
A
Surviving fraction
sis). The linear-quadratic (LQ) model of fitting and interpret-

ing clonogenic cell survival curves became a cornerstone of
experimental radiation oncology and indeed is still widely
0.01
used for predicting the clinical impact of changes in dose 1.6
fractionation or dose rate on biological effect [5].
This model assumes two components of cell killing, one B 7.6
proportional to the dose D (αD), while the other is a qua-
dratic term (βD2) in which two sublesions are presumed to
interact to produce a lethal event. The relationship between 0.001 150cGy/min
cell survival (surviving fraction, SF) and dose is then
described by: −ln [SF] = αD + βD2. The 2-hit/quadratic (β)
term represents the fraction of cell killing caused by damage
that can be spared either by dose fractionation or by decreas- 2 4 6 8 10 12 14
ing the dose rate. In contrast, the 1-hit/linear α-type lethal Radiation dose (Gy)
events are independent of time and thus of fractionation or
dose rate [5, 46]. Fig. 11.2 Conventional dose-rate effect for cell killing as the dose rate
is lowered from 150 to 1.6 cGy/min. The dashed curves represent the
Application of the LQ model to HDR exposures indicates best fit to the data set obtained using the lethal-potentially lethal (LPL)
that cell killing generally decreases as the dose is fraction- model assuming either full repair (a) or no repair (b), respectively
ated, and this sparing would reflect the repair that takes place (reproduced with permission from: Salvatori M, Cremonesi M, Indovina
between fractions of the individual “sublethal” lesions that, in L, Chianelli M, Pacilio M, Chiesa C, Zanzonico P. Radiobiology and
radiation dosimetry in nuclear medicine. In: Strauss HW, Mariani G,
combination, would otherwise have contributed to β-type cell Volterrani D, Larson M, eds. Nuclear Oncology – From Pathophysiology
killing. Survival curves for large numbers of small fractions to Clinical Applications. New York: Springer; 2017:305–49)
ultimately approximate the initial slope (α) of the HDR curve.
ity of cells to low-dose/LDR exposures, as it will be dis-

cussed below [5, 46]. function of the initial uptake and the variable half-
Most cellular responses to ionizing radiation-induced life of the radionuclide used for the therapy. This is
DNA damage are genetically regulated and involve special- in marked contrast with the constant high-dose rate
ized DNA damage recognition factors that trigger a cascade of external beam radiation therapy and
of signaling events that alter the expression and/or activity of brachytherapy.
specific genes/proteins involved in cell cycle arrest, DNA • The possibility to obtain prolonged response, lim-
repair, accelerated senescence, and apoptosis at the cellular ited toxicity, and the ability to treat on multiple
level and in tissue repair at the tissue level [5]. occasions suggest that the mechanism of action of
Necrosis, a form of cell death that has been variously “low-dose rate” therapies is different from the
described as generalized, nonspecific, accidental, or passive “high-dose rate” exposures.
(i.e., it is not a genetically regulated process), generally • The linear-quadratic model is one of the first mod-
occurs after delivering high doses to cells that enter mitosis els of experimental radiation oncology and indeed
carrying high levels of unrepaired DNA damage [5]. is still widely used for predicting the clinical impact
In contrast, apoptosis is an energy-dependent, genetically of changes in dose fractionation or dose rate on bio-
controlled “suicide” process that involves the activation of logical effect.
degradative proteolytic enzymes called caspases and tends to • Most of the cellular responses to ionizing radiation-
occur at low doses of radiation. Apoptotic postirradiation induced DNA damages are genetically regulated
cell death is mediated by two pathways, both resulting in and involve specialized DNA damage recognition
activation of the cascade of caspases that function as cell factors that alter the expression and/or activity of
executioners. First, signals from the nucleus or cell mem- specific genes/proteins.
brane can activate the “intrinsic” pathway of apoptosis with • There are other modes of cell death that are the
a series of sequential events that cause final activation of cas- “accelerated” or “premature” senescence or
pases 3, 6, 7, and 8. Second, the “extrinsic” pathway of apop- “STASIS” (stress or aberrant signaling-induced
tosis is mediated by death receptors that are activated by senescence), that is, a genetically programmed
ligands, such as the Fas ligand, tumor necrosis factor-α response to DNA damage, and the mitotic catastro-
(TNF-α), and TNF-related apoptosis-inducing ligand phe, defined as the failure of a cell to properly
(TRAIL). TRAIL induces apoptosis in response to ionizing undergo mitosis after DNA damage.
radiation through clustering of the DR4 and DR5 receptors
in the cell membrane [39].
Other modes of cell death are the “accelerated” or “pre-
mature” senescence or “STASIS” (stress or aberrant 11.3.1 Overview on Dosimetric Methods
signaling-induced senescence), a genetically programmed
response to DNA damage [47], and the mitotic catastrophe, The basic goal of nuclear medicine therapy is to ensure that
generally defined as the failure of a cell to properly undergo enough radiation-absorbed dose is delivered to the tumor
mitosis after DNA damage. It has recently been recognized while minimizing the risk of toxicity to the bone marrow and
that mitotic catastrophe is not a mode of cell death per se but to other normal tissues. Many physicians administer approx-
a death that generally occurs secondarily to mitotic catastro- imately the same activity to all patients, while ideally admin-
phe, apoptosis, or accelerated senescence. Death secondary istered activity should be adjusted using a patient-specific
to mitotic catastrophe has been suggested to account for a treatment planning strategy based on radiation-absorbed
significant proportion of the cytotoxicity observed after radi- dose. With this approach, activity administration is opti-
ation exposure. mized to maximize treatment efficacy not delivering a radia-
tion dose that could cause deleterious side effects to normal
organs.
Key Learning Points In NMT dosimetry, the absorbed dose is the energy depos-
• Nuclear medicine therapy is expanding with the ited (E) per unit mass of matter (m) (with units of J/kg, 1 J/
development of novel pharmaceuticals, the emer- kg = 1 Gy), where E is the number of radionuclide disinte-
gence of new indications for treatment, and grations in a particular volume multiplied by the energy
improvements in results. emitted per disintegration of the radionuclide and the frac-
• According to the “very low-dose rate” principle, the tion of emitted energy that is absorbed by a particular mass
radiation dose is delivered to the target tissues by (m).
continuous but decreasing exposure, which is a To improve the correlation between dose and the expo-
sure effect, other quantities must be defined, such as relative
biological effectiveness (RBE), radiation weighting factors, where D is the absorbed dose, μ is the exponential repair rate
and tissue weighting factors. This suggests that energy constant that quantifies the rate of sublethal damage repair,
absorbed per unit mass does not predict response at all lev- and λ is the effective clearance rate constant (given by the
els, and some other factors must be considered. sum of the physical decay and the biological clearance rate
Damage to cells is due primarily to indirect effects of constants). It means that the BED may be defined as the
radiation (formation of free radicals in water, that diffuse and product of the total physical dose D and a modifying factor
subsequently interact with cellular components, mostly named the relative effectiveness per unit dose (RE) that
DNA) and to some degree to direct effects (direct damage to quantifies dose-rate effects with respect to radiosensitivity
DNA from radiation interaction). Also, different tissues and and repair of radiation damage.
different individuals have different abilities to respond to and The BED equation is derived from the linear-quadratic
repair this damage. Thus, physical quantities such as the (LQ) model that describes the surviving fraction (Sf) of target
absorbed dose must be linked to radiobiological quantities to cells after a radiation dose (D):
completely understand and be able to predict effects in a
system.
(
Sf = exp -a D - b D 2 ) (11.4)
The internal dose can be calculated by the following sim-
ple equation from the MIRD Committee [9, 10]: where the linear component αD describes the DSBs induced
 by a single ionizing event and the quadratic component βD2
A ´ D ´ fT – S 
DT – S = = A ´ S = A0 ´ t ´ S (11.1) describes the same effect induced by two separate ionizing

MT events; α and β are the tissue-specific coefficients for radia-
 tion damage, α being proportional to dose (one single event
where D is the absorbed dose in a target organ (Gy), A is the is lethal) and β being proportional to squared dose (two sub-
cumulated activity in source region S, Δ is the energy emit- lethal events required for lethal damage). The α/β ratio is
ted by the radionuclide per disintegration, ϕT→S is the frac- also named “repair capacity” and quantifies the sensitivity of
tion of energy emitted—by the radionuclide in source region a given tissue to changes in fractionation. Typical values for
S that is absorbed in the target region, and MT is the mass of the α/β ratio are about 5–25 Gy for early-reacting normal tis-
region T. Furthermore,
 τ is the residence time, which is sim- sues and tumors and about 2–5 Gy for late-responding nor-
ply equal to A /A0, the cumulated activity divided by the mal tissues.
patient’s administered activity (A0). S is the absorbed dose Also with the biological effective dose, the focus in radio-
per unit cumulated activity, and it is given by nuclide dosimetry is to calculate the absolute amount of
energy delivered per mass unit of tissue, i.e., the absorbed
D ´ fT – S
S= (11.2) dose D. In order to plan and design an appropriate dosimetric
MT study, it is necessary to know approximately how the com-

pound will be taken up and cleared from the various organs
Equation (11.1) means that the absorbed dose depends on the and the whole body, by collecting samples at the appropriate
half-life of the radionuclide and its spatial and temporal dis- times. Most therapeutic agents have a relatively fast phase of
tribution in the target. The latter are typically obtained by organ uptake and initial clearance, followed by more general
images acquired at different times after administration of the systemic removal that lasts for many days. Therefore, a typi-
radiopharmaceutical and used to estimate the amount or con- cal sampling scheme is to collect several samples in the first
centration of radioactivity in a specific region. The level of hours after administration and then about once or twice a day
activity obtained at different times after injection, plotted for a few days to 2 weeks.
against time, gives a time-activity curve for a particular tar- To calculate the cumulated activity, the integral of the
get, such as an organ or tissue. The integral of this curve time-activity curve for each source organ may be obtained by
gives the total number of disintegrations or the cumulated (1) direct integration of data, for example, with the trapezoi-
activity ( A ) for the region. Therefore, the main input data dal method; (2) a fit of the data to yield a mathematical
needed for evaluation of radiation dose are the biokinetic expression of the uptake and retention in the target organ;
data that characterize the distribution and retention of the and (3) using compartmental models, if the pharmacokinet-
radiopharmaceutical throughout the biological system. ics inside the target tissue is known.
This absorbed dose could be used to estimate the biologic Specific quantification techniques have been summarized
effective dose (BED) [11]: for 2D and 3D imaging in the MIRD 16 and MIRD 17
publications [12, 13]. Using planar data (2D Imaging), the
é Dl ù most accepted technique is to obtain images from the poste-
BED = D ´ RE where RE = ê1 + ú (11.3)
êë (a / b ) ( m + l ) úû rior and anterior projections and then correct the projected

data in each region of interest (ROI) for attenuation and

scatter. quantities, such as relative biological efficacy, tis-
The most popular technique for attenuation correction sue oxygenation, radiation weighting factors, and
involves the use of a Cobalt-57 (or other appropriate radio- tissue weighting factors.
nuclide) projection source imaged with and without the • The absorbed dose depends on the half-life of the
patient in the view, the attenuation coefficient for the system radionuclide and its spatial and temporal distribu-
having been characterized in advance. tion in the target. The latter are typically obtained
For scatter correction, the two- or three-energy window by images acquired at different times after adminis-
method is widely accepted and applied when gamma camera tration of the radiopharmaceutical and used to esti-
software permits simultaneous acquisition in multiple energy mate the amount or concentration of radioactivity in
windows. a specific region.
One of the most known packages able to offer mean • Specific quantification techniques have been sum-
absorbed doses in the organs, based on uniform activity dis- marized for 2D and 3D imaging in the MIRD 16
tribution in organs/tissues, was the MIRDOSE3.1 package and MIRD 17 publications.
that implemented the use of whole-body MIRD stylized • One of the most known packages generating mean
mathematical phantoms representing adult males and absorbed doses in the organs, based on uniform
females, children, and pregnant women [14]. The activity distribution in organs/tissues, was the
MIRDOSE3.1 software could be used for calculating inter- MIRDOSE3.1 package that implemented the use of
nal dose for a large number of radiopharmaceuticals, the whole-body MIRD stylized mathematical phan-
rapid comparison of calculations for different cases, exami- toms representing adult men and women, children,
nation of dose contributions to different organs, and regional and pregnant women.
marrow dose calculations. • The essential requirements for 3D imaging-based
MIRDOSE3.1 has been updated to a new generation dosimetry are the availability of 3D anatomic imag-
code, Organ Level Internal Dose Assessment (OLINDA), ing studies, such as CT or MRI, at least one 3D imag-
employing the Java programming language and the Java ing study of the radioactivity distribution (e.g., PET
Development Kit environment [48]. The entire code was or SPECT), and software that implements a point
rewritten, but all of the basic functions of the MIRDOSE kernel or Monte Carlo calculation methodology to
code were retained, while others were extended. More indi- estimate the spatial distribution of absorbed dose.
vidual organ phantoms were included, the number of radio-
nuclides was significantly increased (including alpha
emitters), and the ability to perform minor patient-specific
adjustments to doses reported for the standard phantoms was 11.3.2 Nuclear Medicine Treatments
made possible.
Quantitative 3D imaging using single-photon emission Clinical applications of dosimetry are not widely adopted in
computed tomography (SPECT) methods is considerably NMT, mainly because data from prospective randomized
more complex. The essential requirements for 3D imaging- clinical trials that may prove the effectiveness of dosimetry
based dosimetry are the availability of 3D anatomic imaging in predicting clinical outcome after treatment are lacking. In
studies, such as CT or MRI, at least one 3D imaging study of order to prove that dosimetry-based NMT is of additional
the radioactivity distribution (e.g., PET or SPECT), and soft- benefit over administration of fixed empirical activities or
ware that implements a point kernel or Monte Carlo calcula- activities per body weight, prospective randomized phase III
tion methodology to estimate the spatial distribution of trials with appropriate endpoints should be undertaken [51].
absorbed dose. Two fully developed packages are 3D Internal So far, the lack of standardized methodology for calculat-
Dosimetry (3D-ID) and DOSIMG [49, 50]. ing the absorbed doses and the continuing use of approaches
based on administrations of fixed empirical activities irre-
spective of personalized radiation dose estimates have ham-
Key Learning Points pered efforts to compare clinical outcomes in different
• The fundamental goal of nuclear medicine therapy patient populations and has led to significant difficulty in
is to ensure that a sufficient dose of radiation is comparing results among different trials [4].
delivered to the tumor minimizing the risk of toxic-
ity to the healthy tissues surrounding the lesion. 11.3.2.1 Radioiodine Therapy of Differentiated
• To improve the correlation between the dose and Thyroid Carcinoma
the effect of exposure, it is necessary to define other Radioiodine therapy is largely used for ablation of thyroid
remnants and treatment of locoregional or distant metastases
in patients with differentiated thyroid carcinoma (DTC) [52]. administered activity is determined by the sum of contribu-
The treatment can be performed using empiric or fixed pre- tions of blood self-irradiation and of penetrating radiation
scribed activities of 131I-iodide (ranging from 1.1 to 5.5 GBq from the whole body.
for ablation to 5.5–11.0 GBq for treatment of metastases) or The mean blood absorbed dose per unit administered
dosimetrically determined prescribed activities [53, 54]. tracer activity can therefore be calculated as
The advantages of using the empiric prescribed activity
DBlood é Gy ù 0.0188
are convenience, a long history of use, an acceptable rate and ê GBq ú = 108 ´ t Blood [ h ] + wt kg ´ t TotalBody [ h ]
severity of complications, and the possibility to avoid the
A0 ë û (11.5) [ ]
“stunning” phenomenon due to the prior diagnostic
131
I-iodide activity. However, the persistence of disease in a
significant proportion of patients and the possibility for mul- The activity to be administered for a blood absorbed dose of
tiple small empiric activities to have less therapeutic benefit 2 Gy is:
than the same total activity given at one time have led to
2 ´ A0 [ Gy / GBq ]
attempts to improve such empiric approach to therapy by AAdm [ GBq ] = (11.6)
adopting a dosimetry-based approach [54]. Dblood

Two basic dosimetrically determined prescribed activity
approaches are available, each one addressing a different Under the conservative assumption that the activity concen-
aspect of this problem. A summary of the methodologies trations within the hematopoietic tissue and the blood are
used can be found in the review article by van Nostrand [54] identical [59], a red marrow-based approach for the determi-
and in the guidelines of the Dosimetry Committee of the nation of the maximum activity to be administered has been
European Association of Nuclear Medicine (EAMN) for pre- proposed [59]. Although this method seems to be accurate,
therapeutic dosimetry [55]. no systematic clinical validation of the red marrow absorbed
dose versus toxic effect has yet been undertaken.
Maximum Tolerated Activity (Benua–Leeper Method) The limited bone marrow dosimetry is easy to perform
This approach was developed at the Memorial Sloan both pre-therapeutically and peri-therapeutically, thus allow-
Kettering Cancer Center by Benua et al. and estimates the ing to increase the therapeutic activity for selected patients
131
I-iodide maximum tolerated activity (MTA) that would without risk of severe side effects. However, the original
deliver a radiation dose to the blood (considered as a surro- method might not to be applicable in the presence of extended
gate for the bone marrow) of 2 Gy [56–58]. metastatic bone involvement, as the blood-based absorbed
The method requires evaluation of the kinetics of activity dose calculation could underestimate the absorbed dose to
in the blood and in the whole body. The kinetic of activity in the red marrow. Similarly, it should be applied with caution
the blood is assessed using five blood samples obtained over also in the case of patients with diffuse lung micrometasta-
1 week and measuring the blood activity with a calibrated ses, because the critical organ could be the lung itself instead
well counter from aliquots of blood samples [54]. The time of the hematopoietic red marrow [60, 61]. The main limita-
dependent changes of activity in the whole body are obtained tions of this method can be summarized as follows: (1) the
by conjugate views of a whole-body scan (WBS) acquired method has not yet been validated with clinical data on out-
with a dual-head gamma camera equipped with high-energy come rates demonstrating a benefit of the strategy; (2) the
collimators [54]. The first WBS measurement obtained 2 h method does not provide dosimetry estimates on absorbed
after tracer administration is considered to represent 100% of dose to the thyroid remnant or metastasis; (3) the “stunning”
administered activity, while measurements are repeated at effect from prior administration of a diagnostic activity of
24, 48, 72, and 96 h (or later if uptake and/or renal clearance 131
I-iodide might alter the uptake and retention kinetics dur-
are impaired). ing subsequent therapy with radioiodine; and (4) the method
The curves, A(t), describing the activity in blood and in entails certain additional costs and practical inconvenience.
the total body as a function of time after the administration To calculate the whole-body activity as a function of time
are usually multi-exponential, and bi-exponential fitting is and circulating activity per milliliter of blood, the geometric
usually appropriate to determine the equation describing mean of corresponding net counts obtained by conjugate
time-activity curves in the blood and in the total body, views and all blood activities is normalized to the first data
respectively. point and to the administered activity, respectively [54, 55].
The residence times in the whole body and activity con-
centration in blood, t TotalBody [h] and t Blood [h], are cal- Lesion-Based Dosimetry (Maxon Method)
culated by integrating the corresponding retention function The second approach (Maxon method) aims at individualiz-
R(t) = A(t)/A0. According to the generally accepted MIRD ing radioiodine activity that delivers the recommended
formalism, the mean absorbed dose to the blood per unit absorbed doses to ablate thyroid remnant (traditionally
≥300 Gy) or to treat metastatic disease (traditionally nomenon (for 131I-iodide) [53]. For these reasons, PET/CT with
≥80 Gy) while minimizing the risk to the patient [57, 62]. In 124
I-iodide (124I-PET/CT) is becoming an emerging and attrac-
order to determine the activity required to deliver the tive methodology for lesion dosimetry in the management of
absorbed doses, it is necessary to measure the uptake, the patients with DTC [63], providing quantitative images with
clearance, and the concentration of 131I contained in the iden- high spatial resolution and sensitivity [63–65].
tifiable thyroid remnants and/or metastases. In general, a 124I-PET/CT dosimetry protocol involves esti-
One way to ascertain these parameters is through an anal- mating the absorbed dose per administered 131I activity for each
ysis of selected regions of interest (ROIs) on conjugate-view radioiodine-positive lesion, allowing the calculation and choice
gamma camera images or on SPECT images, obtained at of the actual recommended 131I-iodide activity. However, pre-
sequential time points after administration of a tracer activ- therapeutic 124I-PET/CT dosimetry requires knowledge of the
ity. Typically, these images should be acquired at 24, 48, 72, lesion volumes in order to adequately correct for partial volume
and 96 h after tracer administration, but later time sampling effects. If no morphological CT correlates can be assigned to
might be necessary if the uptake and clearance are impaired. 124
I-positive foci, the target volumes must then be assessed from
In addition, transmission images to correct for attenuation in PET images alone by applying various threshold methods.
the lesion area, scatter images, as well as calibration proce- The original 124I-PET dosimetry protocol entailed five
dures, are necessary. A curve-fitting procedure is then PET measurements at 4, 24, 48, 72, and 96 h after 124I-iodide
employed to determine the assumed single-exponential half- administration, although simplified models using four, three,
life value and to extrapolate the curve to zero time to deter- or two time-points (24 h/96 h) have been proposed [66, 67].
mine the initial activity in the lesion [54]. At the moment, considering the slow progression typical of
Pre-therapeutic dosimetric assessments of the activity DTC and the long follow-up period, it is difficult both to
required to achieve a certain prescribed absorbed dose to a evaluate the long-term benefits and to design classical
remnant or lesion are often based on adaptations of the blinded and randomized studies in DTC patients submitted
generic MIRD equation for absorbed dose, as previously to pre-therapy 124I-PET/CT dosimetry. However, pre-therapy
described. 124
I-PET/CT dosimetry seems to be very useful in selected
 patients with locoregional or distant metastases who may
D = A´ S (11.7) benefit from radioiodine therapy with an individually tai-
Where D denotes the mean absorbed dose to the remnant/ lored escalated activity or to abandon the therapy because of
lesion, A is the cumulative activity, and S is the “S factor” of insufficient obtainable dose in the lesions.
the MIRD scheme depending on the lesion/remnant mass.
For dosimetry of metastases, CT, MRI, or US can be used for
attenuation correction and to determine the mass, while no Key Learning Points
thoroughly validated method is yet available for ablative • Radioiodine treatment of locoregional or distant
treatments to exactly calculate the thyroid remnant volume metastases in patients with differentiated thyroid
after surgery. If the lesions are small, the “nodule module” of carcinoma (DTC) can be performed using empiric
the OLINDA/EXM software might be useful to generate a or dosimetrically determined activities.
spherical model of the remnant and/or tumor. Nevertheless, • Two basic dosimetrically determined prescribed
if the size of the lesion is smaller than 5.0 mm (assuming that activity approaches are available.
such small size can actually be accurately determined), then • The first method (the so-called Benua–Leeper
the tissue range of the beta particles can no longer be method) estimates the 131I-iodide maximum toler-
neglected in the dose calculation. The limitations of the ated activity (MTA) that would deliver a radiation
lesion-based method can be summarized as follows: (1) dose to the blood (considered as a surrogate for the
absorbed doses within each given lesion are generally not bone marrow) of 2 Gy.
uniform; (2) it is not always possible to estimate the lesion • The second approach (Maxon method) aims at indi-
mass; (3) statistical errors in the measurements may occur vidualizing radioiodine activity that delivers the
for lesions with low radioiodide uptake and low count rates; recommended absorbed doses to ablate thyroid
(4) the biological effectiveness of dosimetry-guide therapy remnant (traditionally ≥300 Gy) or to treat meta-
with 131I-iodide is not proven yet; and (5) the “stunning” static disease (traditionally ≥80 Gy) while mini-
effect from prior administration of a diagnostic activity of mizing the risk to the patient.
131
I-iodide might alter the uptake and retention kinetics dur- • PET/CT with 124I-iodide is emerging as an attractive
ing subsequent therapy with radioiodine. methodology for lesion dosimetry in the manage-
However, dosimetry performed with 123I-iodide or 131I-iodide ment of patients with DTC, providing quantitative
has limitations, such as cost and the logistics of supply (for images with high spatial resolution and sensitivity.
123
I-iodide) and poor imaging properties and “stunning” phe-
11.3.2.2 P eptide Receptor Radionuclide Table 11.2 Average absorbed dose estimates derived from the main
Therapy PRRT trials with 90Y-DOTATOC and 177Lu-DOTATATE
Patients with somatostatin receptor expressing tumors, Organ/tissue/region Absorbed doses per unit activity (Gy/GBq)
including neuroendocrine tumors, can be treated with Y-DOTATOC
90 177
Lu-DOTATE
peptide receptor radionuclide therapy (PRRT) [68, 69]. Red marrow 0.09 0.04
Kidneys 2.68 0.82
Due to radiation-induced renal damage and large inter-
Liver 0.69 0.17
patients’ variability in biodistribution and tumor uptake,
Spleen 4.87 1.33
such patients require accurate dosimetry-based treatment Bladder 1.75 0.31
[70–84]. Total body 0.18 0.05
90
Y-DOTATOC, 90Y-DOTATATE, and 177Lu-DOTATATE Average values calculated from data reported by: Salvatori M,
are the most widely employed radiopharmaceuticals for Cremonesi M, Indovina L, Chianelli M, Pacilio M, Chiesa C, Zanzonico
PRRT, showing similar biological half-lives for organs and P. Radiobiology and radiation dosimetry in nuclear medicine. In:
lesions, but variable uptake depending on the different Strauss HW, Mariani G, Volterrani D, Larson M, eds. Nuclear
Oncology – From Pathophysiology to Clinical Applications. New York:
expression of somatostatin receptors with respect to the pep- Springer; 2017:305–49
tide affinity.
90
Y- and 177Lu-peptides each have advantages and limita-
tions. The physical half-life of 90Y is compatible with the
peptide kinetics, and the high energy of the β− particles con- 80
fers high probability of killing all neoplastic cells through

the cross-fire effect. However, 90Y is not easily suitable for
imaging due to the lack of γ-emission [85–87]. 60
111
In-octreotide, 86Y-octreotide, and 68Ga-peptides have been cumulative A% eliminated
proposed as alternative options to 90Y imaging [88] with lim- A% Blood
A (%)
itations due to different biodistribution kinetics and physical 40

T1/2, complex image quantification, cost, and poor availabil-
ity [88].
Compared with 90Y, 177Lu induces less damage by 20
cross-fire effect but releases higher energy in smaller tis-
sues. The γ-rays emitted by 177Lu are suitable for imaging
and dosimetry, although for proper scintigraphic imag-
0
ing, diagnostic/dosimetric activities of at least of 370– 10 20 30 40 50
740 MBq should be administered. It should be noticed, hours
however, that most often dosimetry is evaluated after
administering a therapeutic dose of 177Lu-DOTATATE Fig. 11.3 Radiopeptides blood clearance (red curve) and urinary
radioactivity excretion (blue dotted curve) (reproduced with permission
(i.e., 3.7–7.4 GBq), since treatment is usually given in from: Salvatori M, Cremonesi M, Indovina L, Chianelli M, Pacilio M,
multiple cycles. So, the activity to be administered in Chiesa C, Zanzonico P. Radiobiology and radiation dosimetry in
cycles after the ‘therapeutic’ dosimetry can be adjusted nuclear medicine. In: Strauss HW, Mariani G, Volterrani D, Larson M,
based on the dosimetry results. eds. Nuclear Oncology – From Pathophysiology to Clinical Applications.
New York: Springer; 2017:305–49)
The absorbed doses per unit activity for 177Lu-peptides are
much lower than for 90Y-peptides [88]. The activities admin-
istered in 177Lu- or 90Y-PRRT should vary accordingly,
although the therapeutic protocols most commonly employed dose-limiting organs; and (4) no significant uptake in bone or
in clinical studies are not yet uniformly established with red marrow is typically reported.
respect to activities administered and number of cycles The infusion of protective agents (negatively charged
[89–91]. amino acids) allows 25–65% reduction of the kidney uptake,
Table 11.2 summarizes the main dosimetric results for the so reducing the renal radiation burden [92]. Despite such
most commonly used 90Y- and 177Lu-peptides (90Y-DOTATOC protection, the cumulative dose to the kidneys could still
and 177Lu-DOTATATE), which share some basic aspects: (1) induce nephropathy, especially with 90Y-peptides.
pharmacokinetics is characterized by very fast blood clear- An important issue in PRRT dosimetry is the possible
ance and urinary elimination, leading to low exposure of the variation in dose and biokinetics among cycles. It has been
whole body (Fig. 11.3); (2) the spleen, kidneys, and liver demonstrated that in most cases, the first two of the four
receive the highest absorbed doses; (3) the kidneys are the planned therapy cycles make the major contribution to the
tumor-absorbed dose, possibly due to a saturation of the pep- Although long-term data analyses of 90Y-trials and
tide receptors, while the different cycles contribute on aver- 177
Lu-trials have been published, the intrinsic potential toxic-
age equally, within 10%, to the absorbed dose to the kidneys ity of the two radiopeptides is still to be elucidated [88, 96,
[70]. 101, 102]. This is because the 177Lu-based and the 90Y-based
In an attempt to improve accuracy of renal dosimetry esti- clinical protocols are conceived in terms of fixed total activi-
mates, the standard methods that assume uniform activity ties that lead to lower kidney absorbed dose and BED values
distribution in kidneys have been shifted to the multi-region for 177Lu-schemes as compared with 90Y. A reliable compari-
MIRD model for a sub-organ kidney dosimetry and dose- son of 177Lu- versus 90Y-peptide therapy would require
rate effects [48, 82, 93, 94]. instead randomized trials conceived to release the same
The impact of the physical characteristics of the radionu- absorbed dose or BED using different radiopeptides. In prin-
clides with respect to inhomogeneous activity distribution is ciple, the physical characteristics of 177Lu, with lower tissue
a further issue concerning potential nephrotoxicity [91]. A penetration, make it more suitable for small tumors and
threshold biologically effective dose (BED) of 33 Gy for kid- micrometastases, while the cross-fire effect 90Y might have
ney radiation damage and a BED50 (defined as the BED value the advantage of a higher radiation burden to larger lesions
derived from the TD50 value derived from the EBRT) of [88]. This hypothesis has been supported by preclinical find-
44 Gy were found for 90Y-PRRT [82]. ings and is laying the basis for new clinical rationales pro-
Clinical experiences have shown that multiple-cycle proposing cocktails of 177Lu- and 90Y-peptides. New interesting
tocols lower nephrotoxicity, while therapy performed by approaches propose tandem treatments with 90Y- and
using few cycles increases the BED values and the side 177
Lu-peptides, with equal or different administered activities
effects to the kidneys [91, 95, 96]. This observation perfectly of the two radiopharmaceuticals [103, 104].
matches with the theoretical expectation from the BED con-
cept: for a given cumulative dose, the higher the number of
cycles the lower the total BED, thus the lower the damage
[91, 97]. Clinical results on 90Y-PRRT have provided evi- Key Learning Points
dence of the safety up to a BED of about 40 Gy to the kid- • Patients with somatostatin receptor-expressing
neys, cumulatively, and that risk factors (such as hypertension tumors, including neuroendocrine tumors, can be
and diabetes) lower the tolerability to a BED of about 28 Gy treated with peptide receptor radionuclide therapy
[82, 95]. (PRRT).
Treatment protocols based on multiple therapy cycles rep- • 90Y-DOTATOC, 90Y-DOTATATE, and 177Lu-
resent a useful modality to lower toxicity for a same cumula- DOTATATE are the most widely employed radio-
tive activity, and/or improve the therapeutic outcome for a pharmaceuticals for PRRT, exhibiting similar
same fixed BED to the kidneys, owing to the different radiobiological half-lives for organs and lesions but vari-
sensitivity of most tumors versus the kidney tissue. Cycles able uptake depending on the different expression
seem especially effective in 90Y-PRRT, where frequent occur- of somatostatin receptors relative to the peptide
rence of renal impairment has been clinically confirmed [91, affinity.
95, 98]. • 90Y- and 177Lu-peptides each have advantages and
177
Lu-peptides appear to be less nephrotoxic [69, 99, 100]. limitations, and the intrinsic potential toxicity of the
Possible reasons for these different outcomes are the differ- two radiopeptides is still to be elucidated.
ent particle ranges in tissues and peptide localization that • New interesting approaches propose tandem treat-
might deliver a higher dose to the tubuli (radioresistant cells) ments with 90Y- and 177Lu-peptides, with equal or
compared to the glomeruli (radiosensitive cells). The advan- different administered activities of the two
tage of cycling PRRT with 177Lu is not so prevalent as with radiopharmaceuticals.
90
Y. This is due to the observation that the cumulative
absorbed dose to the tumor after each therapy cycle gradu-
ally declines [70] and that the absorbed dose to the kidneys
with 177Lu-PRRT is much lower, leading to a lower risk of
nephrotoxicity. 11.3.2.3 Radioimmunotherapy of Lymphoma
Overall, in the balance of tumor response and kidney tox- Radioimmunotherapy (RIT) offers an important option for
icity, the use of too many cycles is not recommended for the treatment of follicular low-grade non-Hodgkin’s lym-
177
Lu-PRRT, while four cycles or possibly less seem to offer phoma (NHL) refractory or relapsed after treatment with the
a better chance to avoid repair of sublethal cell damage and current best practice [105]. The two radiolabeled antibodies
repopulation without dramatically alter the renal risks. commercially available for treatment (tositumomab and ibri-
tumomab) constitute the radiolabeled immunoglobulin com- The purpose of the scan is simply to evaluate the whole-
ponents of Bexxar® and Zevalin®, respectively [105, 106]. body total counts rather than to produce a diagnostic image,
Bexxar® is the 131I-labeled anti-CD20 antibody tositu- because the parameters involved in visualizing a tumor (dose
momab, while Zevalin® is the 90Y-labeled anti-CD20 anti- administered, tumor size, and location) are different from
body ibritumomab. Although both tositumomab and those involved in radiobiological response evaluations [105].
ibritumomab recognize the same CD20 antigen (one of many Whole-body counts are determined from the total counts
epitopes expressed on mature B cells), they have slightly dif- on the anterior and posterior WBS acquired 1 h postinjection
ferent binding characteristics [105]. (day 0), then 2 days (day 2), and 5 days (day 5) later, to deter-
For both radiopharmaceuticals, an unlabeled antibody is mine the radiopharmaceutical
 residence time (τ) and cumu-
infused before administration of the radiolabeled murine anti- lated activity ( A ) in the body.
CD20 antibody component. In the Bexxar® regimen, both the To calculate the desired therapeutic activity, the following
labeled and unlabeled antibody are tositumomab. In the Zevalin® equation can be used:
regimen, the unlabeled antibody is rituximab (the chimeric anti- 
body used as an immunotherapeutic and marketed as Rituxan®), A Desired TBD ( cGy )
Theraputic dose = (11.8)
while the labeled antibody is ibritumomab [105, 106]. t 75 cGy

In Europe only Zevalin® is commercially available, while in 
the USA and Canada both Zevalin® and Bexxar® have been The calculation simply divides a cumulated activity, A
approved for clinical use [105]. Nonetheless, Bexxar® has sub- (reported for several patient’s weight in Table 11.3), that
sequently been withdrawn from the market also in the USA and would result in the patient receiving a 75 cGy whole-body
Canada. No direct comparison in a randomized blinded trial radiation-absorbed dose, based on the patient-specific resi-
has been performed to assess whether the difference in physical dence time. This would result in the activity associated to a
characteristics of the two radionuclides, 131I and 90Y, have dif- 75 cGy whole-body absorbed dose that could be also res-
ferent therapeutic effects or toxicity. caled by an absorbed dose ratio if a value of total-body
In principle, the greater energy of the β− emitted by 90Y is absorbed (TBD) dose different from 75 cGy is desired.
thought to be advantageous in the therapy of larger tumor masses,
whereas the lower energy of the 131I particle emission would be
an advantage for treating patients with small tumor foci and mar- Table 11.3 Organ radiation absorbed doses (mGy/MBq) for
row involvement. In the non-myeloablative setting, the use of 90
Y-ibritumomab tiuxetan (Zevalin®), based on pre-therapeutic 111In
either agent is contraindicated for patients with lymphomatous imaging
infiltration exceeding 25% of the bone marrow [107]. Absorbed doses per unit activity
Organ/tissue/region (mGy/MBq)
For Bexxar® and Zevalin®, different protocols and dose
Median Range
determinations have been reported, also due to the different
Spleen 7.35 0.37–
approval requirements by regulatory agencies in the USA 29.7
and Europe. However, for both radiopharmaceuticals, the red Liver 4.32 0.85–
bone marrow is the dose-limiting organ due to its intrinsic 17.5
radiosensitivity, to the rapid equilibration of the radiolabeled Lungs 2.05 0.59–
antibodies within its extracellular fluid volume, and to the 4.86
Red marrow (blood-derived) 0.59 0.09–
long retention kinetics in the bone marrow [105]. 1.84
When using Bexxar®, dosimetry is estimated to determine Red marrow 0.97 n.a.
the whole-body radiation-absorbed dose, since hematologic (sacrum-derived)
toxicity as well as the dose-clinical response relationship Kidneys 0.22 0–0.95
depends on the whole-body radiation-absorbed dose rather Bone surfaces 0.53 0.09–
than on a MBq/kg dosing schedule [108]. 1.31
Bladder 0.89 0.38–
Optimal clinical benefit with acceptable hematologic tox-
2.32
icity has been observed at 65–75 cGy whole-body radiation- Other organs 0.41 0.06–
absorbed dose, considering 65 cGy for patients with platelet 0.62
counts between 100,000 and 150,000 and 75 cGy for patients Total body 0.54 0.27–
with platelet counts greater than 150,000 [105]. 0.78
The whole-body radiation-absorbed dose is determined by Average values calculated from data reported by: Salvatori M,
three WBS acquisitions after the administration of 185 MBq Cremonesi M, Indovina L, Chianelli M, Pacilio M, Chiesa C, Zanzonico
P. Radiobiology and radiation dosimetry in nuclear medicine. In:
of 131I-tositumomab (with prior infusion of 450 mg of unla- Strauss HW, Mariani G, Volterrani D, Larson M, eds. Nuclear
beled tositumomab), administered for determining the resi- Oncology – From Pathophysiology to Clinical Applications. New York:
dence time, which is necessary for dosimetric estimates. Springer; 2017:305–49
Seven to nine days after the initial “dosimetric” study, the

infusion sequence (unlabeled tositumomab followed by • For both radiopharmaceuticals, the red bone mar-
131
I-tositumomab) is repeated with the 131I-tositumomab ther- row is the dose-limiting organ due to its intrinsic
apeutic activity (1.85–5.5 GBq) that had been determined by radiosensitivity, to the rapid equilibration of the
dosimetry. radiolabeled antibodies within its extracellular fluid
In the US and EU countries, biodistribution studies are volume, and to the long retention kinetics in the
not required because of the lack of correlation between red bone marrow.
marrow dose and toxicity and considering that red marrow
toxicity is reversible. This choice (which seems to be dic-
tated mainly by practical/economic reasons) is arguable, as
cases in which only the scintigraphic images were able to 11.3.2.4 Trans-Arterial Radioembolization
predict toxicity of the treatment are not so rare [109]. of Liver Tumors
However, when using larger activities in clinical trials Trans-arterial radioembolization (TARE) is a locoregional
involving myeloablative doses with stem cell transplanta- treatment developed in order to release high-radiation doses
tion, personalized dosimetry estimates are required to avoid to unresectable primary hepatocarcinoma (HCC) [116] or
unpredictable toxicity. In fact, with bone marrow rescue, it is metastatic hepatic malignancies [117] by intra-arterial
not possible to identify a priori a “safe” maximal activity or administration of microspheres loaded with 90Y.
a unique critical organ for all patients, because of the wide Since the normal liver is mainly fed by portal blood and
intra-patient variability in doses to the vital organs. liver tumors by arterial blood, the administered radioactive
At present, neither 90Y-ibritumomab tiuxetan dosimetry agent is permanently trapped in the microcapillaries of the
outcomes in favor of its predictivity have been found nor cor- tumor where it delivers its radiation until complete physical
relation between red marrow dose and hematologic toxicity decay.
has been demonstrated [108–113]. This consideration has Currently, two medical devices are approved for clinical
discouraged the practice of red marrow dosimetry for use: (1) SIR-Spheres® made in plastic resin (SIRTex Medical,
Zevalin®. In this regard, toxicity in these patients is most Lane Cove, Australia) and (2) TheraSphere® made in glass
likely linked to involvement of bone marrow by the disease (MDS Nordion, Ottawa, Canada). The main difference
and to reduced bone marrow reserve due to previous aggres- between the two products is the number of administered par-
sive chemotherapy regimens [114]. ticles, which is ten times greater for the resin particles. This
The therapeutic activity of Zevalin® for the nonmyeloab- implies a truly embolic behavior, never observed in glass
lative therapy of rituximab-refractory or relapsed patients microspheres. The second difference is the approved indica-
with low-grade follicular non-Hodgkin’s lymphoma is pre- tion in the USA versus Europe. FDA approved the use of
scribed proportionally to the patient’s weight, with recom- resin spheres in patients with colorectal carcinoma meta-
mended administered activity of 15 MBq/kg up to a static to the liver, while glass spheres are approved for treat-
maximum limit of 1200 MBq [111]. Patients with platelet ment of HCC patients with or without portal vein thrombosis.
counts between 100,000 and 150,000 should receive In Europe and in various countries worldwide, both agents
11 MBq/kg [111]. Much higher activities, ranging from 29 to have regulatory approval for liver malignancies in general.
55 MBq/kg, have been used under myeloablative regimens, The amount of activity to be administered should gener-
followed by autologous stem cell transplantation (ASCT) ally be established taking into account the clinical and dosi-
[115]. metric factors which affect the outcome, i.e., baseline liver
function (Child Pugh, portal vein thrombosis, previous or
concomitant treatment), tumor burden, targeted organ frac-
tion, predicted absorbed dose to normal tissue, and absorbed
Key Learning Points dose to lesions. In practice, choice of the injected activity has
• Radioimmunotherapy (RIT) offers an important been based on empirical approaches or on approximate dosi-
option for the treatment of follicular low-grade non- metric evaluations.
Hodgkin’s lymphoma (NHL) refractory or relapsed In order to evaluate possible shunting of intra-arterially
after treatment with the current best practice. injected particles to the lungs (Lung Shunt Fraction, LSF) or
• The two radiolabeled antibodies that have been to the gastrointestinal (GI) tract, which would limit the
made commercially available for treatment (tositu- amount of activity that can safely be administered or would
momab and ibritumomab) constitute the radiola- even be an absolute contraindication to treatment, before
beled immunoglobulin components of Bexxar® and actual therapy a planar or SPECT scan of the trunk is per-
Zevalin®, respectively. formed after intra-arterial administration of 99mTc-albumin
macroaggregates (99mTc-MAA).
Thanks to the similar hepatic distribution patterns ity of this method in terms of tailored evaluation should not
obtained by 99mTc-MAA imaging and by bremsstrahlung be overemphasized. In fact, direct measurements in patients
imaging after administration of 90Y-microsphere, the acquisi- have provided evidence that the BSA-based estimates do not
tion of pre-therapeutic 99mTc-MAA images can be quantita- correlate with liver mass nor with tumor involvement.
tively analyzed for a preliminary dosimetric evaluation Comparison between the BSA-based and the dosimetry
[118]. However, the actual biodistribution of the therapeutic methods showed discrepancies in activities to be adminis-
microspheres is evaluated by post-therapy imaging obtained tered ranging from −30% to 35% [121].
by bremsstrahlung SPECT or, more accurately, by 90Y PET. In the partition method [122], the activity to be adminis-
Patient-specific dosimetry requires a detailed slice-by- tered is calculated within the MIRD equations, once a limit
slice volume of interest (VOI) drawing around tumor (T) and dose is prescribed for NT; lungs safety is also considered.
non-tumor tissue (NT). Using the mean dose approach, val- Liver involvement, tumor avidity as compared to the NTL,
ues of the absorbed doses per unit activity can easily be and the possible occurrence of lung shunt are taken into
derived from two simple equations. The first is the way to account by the following equation:
convert counts in the image into 90Y activity:
AT ´ M nL + AL ´ M nL
A = DnL = (11.12)

AVOI ( 90
)
Y : Atotal ( 90
)
Y = VOI counts : total counts (11.9)

49, 670 ´ AL (1 - LSF )
AVOI éë 90 Y ùû where A = activity, DnL(Gy) = absorbed dose (Gy) limit for

D [ Gy ] = 50 × (11.10)
M VOI [ kg ] NTL, AT = tumor activity (GBq), AL = liver activity (GBq), M
(g) = NTL mass, and LSF = lung counts/(lung + liver counts).
A more sophisticated evaluation can be made by the voxel Although this model was originally designed for single or
dosimetry approach or by Monte Carlo modeling that pro- discrete nodules, appropriate modification by a weighted T/
vides information on dose distribution and expected radio- NT ratio makes it easily applicable also to multiple lesions
biologic effects at the voxel level [119]. [107]. The review paper by Cremonesi et al. [102] provides
Producers indicate different approaches to choose the the absorbed dose limits applied by several authors, includ-
amount of activity to be injected for the two radiopharma- ing the most common recommendations [123] and more
ceuticals [118]. Three main methods are suggested [102] for conservative approaches [124].
resin spheres: the empirical method, a body surface area The model proposed for TARE with glass spheres relies
(BSA)-based method, and the partition method. on a simplified dosimetric equation, in which the dose aver-
The empirical method does not personalize the treatment aged on the injected liver (DL) is prescribed
for different liver sizes or tumor avidity, but it takes into
VL ´ 1.03
account the tumor involvement and attempts to lower the A = DL ´ (11.13)
radiation risks to the lungs and the NTL. With this method 50 (1 - LSF )

activities range from 1.2 to 3.0 GBq and the doses received
by lungs and by the NT are within 18 and 83 Gy, at most where A = activity, DL(Gy) = prescribed liver dose, VL = total
[120]. liver volume, and LSF = lung shunt fraction.
The BSA method, probably the most widely employed for Absorbed doses to tumor and NT are not separately calcu-
resin microspheres, has been proposed since in normal sub- lated, so no distinction is made for different tumor involve-
jects the liver mass correlates with BSA. The basic empirical ment or uptake ratios.
equation includes the patient body surface area [BSA (m2) = Several studies [123] have indicated prescribed radiation
0.20247 × H0.725 × W0.425, with H = height (m) and doses of at least 100 Gy for a maximal effect on tumors,
W = weight (kg)] and provides a maximum activity supposed based on the fact that no dose-limiting organ toxicity has
to be safe for NT and lungs, once applying the same restric- been observed in patients who had received nominal absorbed
tions for lung shunt as above: doses of up to 150 Gy. Mean cumulative doses to the treated
liver as high as 200 Gy or even 390 Gy have been described
MT
A = ( BSA - 0.2 ) + (11.11) as tolerated in patients affected by HCC with Okuda stage II
MT + ML and stage I who received multiple treatments [125].

Overall, these radiation doses are apparently quite far
where A = activity (GBq), MT = tumor mass, and ML from the limits recommended in the 90Y-resin approaches
(g) = total liver mass. In general, the activities calculated by and much farther than the tolerance doses set by external
the SBA method are more conservative than those derived beam radiotherapy (30–35 Gy) to avoid excessive risk of
from the empiric method and never exceed 2.5 GBq. radiation hepatitis [124]. This is explained with the nonuni-
Although including some individual parameters, the reliabil- formity of absorbed dose deposition at microscopic level,
which is markedly higher for the less numerous glass micro-

spheres [126]. Key Learning Points
Another basic point is the usually different administration • Trans-arterial radioembolization (TARE) is a
technique. From the USA experience following the different locoregional treatment developed in order to release
approved uses, glass spheres against HCC have been injected high-radiation doses to unresectable primary hepa-
into a lobe, while resin spheres against metastases were tocarcinoma (HCC) or metastatic hepatic malignan-
administered in the common hepatic artery. These are two cies by intra-arterial administration of microspheres
completely different situations, since the lobar approach loaded with 90Y.
exploits the organ reserve and the regeneration capability of • SIR-Spheres® made in plastic resin (SIRTex) and
the liver [127]. TheraSphere® made in glass are the two medical
The possibility of performing a real treatment planning in devices approved for clinical use.
radioembolization is emphasized by the result of three stud- • Before TARE, a planar or SPECT scan of the trunk
ies, which applied radiobiological models to analyze is performed after intra-arterial administration of
response and toxicity after standard administration of 99m
Tc-MAA in order to evaluate possible shunting
microspheres. of intra-arterially injected particles to the lungs.
Strigari et al. [128] derived the normal tissue complica- • The pre-therapeutic 99mTc-MAA images can be
tion probability (NTCP) and tumor control probability (TCP) quantitatively analyzed for a preliminary dosimetric
for HCC treated with resin spheres using post-therapy brems- evaluation.
strahlung SPECT. TCP(50%) was at 150 Gy, while a dose • The actual biodistribution of the therapeutic micro-
above 200 Gy demonstrated response in all cases. Liver tox- spheres is evaluated by post-therapy imaging
icity higher than G2 had an NTCP (50%) value of 52 Gy. obtained by bremsstrahlung SPECT or, more accu-
Flamen et al. [129] found a good correlation between the rately, by 90Y PET.
99m
Tc-MAA-based dose estimate and [18F]FDG response in
colorectal metastases treated with resins spheres.
Chiesa et al. [130] determined retrospectively with 99mTc-
MAA SPECT the NTCP curve for Child A patients treated 11.3.2.5 Therapy of Neuroendocrine Tumors
lobarly with glass microspheres, with a risk of radioinducing with 131I-MIBG
liver decompensation at 70 Gy if averaged over the whole Since the late 1980s, 131I-meta-iodobenzylguanidine
non-tumoral parenchyma, including the non-injected lobe. (131I-MIBG) has been used for the therapy of malignant neu-
This mean inclusive on the non-treated tissue is the simplest roectodermal tumors, including pheochromocytoma, para-
way to include the well-known liver volume effect (the ganglioma, carcinoid tumors, medullary thyroid cancer, and
smaller the irradiated fraction, the higher the tolerance). TCP neuroblastoma.
was dramatically dependent upon the lesion size, being As an analog of noradrenaline, at low concentrations
TCP(50%) = 250 Gy for masses less than 10 g and around 131
I-MIBG is taken up into cells of neuroectodermal origin
1300 Gy for larger tumors. by the noradrenaline transporter (NAT); once internalized, it
Garin et al. [131] were able to demonstrate a correlation is stored in intracellular storage granules. Transfer of MIBG
between lesion-absorbed dose and overall survival in HCC from the cytoplasm into neurosecretory granules is mediated
patient treated with glass spheres, with a cutoff of 205 Gy. by an ATPase-dependent proton pump; MIBG is not metabo-
This could means that therapy with glass spheres requires at lized but excreted unchanged [134].
least 205 Gy to stabilize the disease, while objective Treatment with 131I-MIBG is indicated for tumors show-
responses require a higher dose. Individualized dosimetry- ing adequate uptake and retention of radiolabeled MIBG on
based administrations with increase of administered activity the basis of a pre-therapy diagnostic scan, which is usually
with respect to the producer indication were also safe and performed using 123I-MIBG [135].
successful [132]. Most treatments are performed in children with neuro-
The unexplored way toward treatment planning in radio- blastoma arising in the nervous system or in the abdomen
nuclide treatment is open thanks to dosimetry in radioembo- near the adrenal glands [136].
lization, which is the less demanding and that with higher High-specific activity (up to 1.48 GBq/mg) is recom-
impact among all kinds of radiopharmaceutical type of mended for therapy, and single-administered activities range
dosimetry. As a confirmation, for the first time in the history between 3.7 and 11.2 GBq, according to tumor burden or
of nuclear medicine therapy, both 90Y microspheres produc- local legislation. Since several cycles of therapy with
ers have invested sizable funding in multicenter research 131
I-MIBG may be required to achieve objective response,
about dosimetry [133]. these treatments are often repeated at widely different inter-
vals, depending on blood count recovery; treatment is con- Table 11.4 Average of median values and ranges of organ absorbed
tinued until the maximum clinical response is achieved. doses values (Gy) during therapy with 131I-MIBG, per unit of activity
administered (MBq/kg)
Red bone marrow is the dose-limiting organ for therapy
Organ/tissue/region Absorbed doses (Gy)
with 131I-MIBG; therefore, activity reduction should be
Av. Median Range
considered in patients with myelosuppression or with
Red marrow 2.56 0.44–5.02
impaired renal function. In patients with neuroblastoma Liver 9.20 1.60–19.0
who have received prior intensive chemotherapy, the dose- Blood 1.29 0.20–3.53
limiting toxicity of 131I-MIBG therapy is myelotoxicity at Whole body 2.32 0.57–6.50
the 2-Gy whole-body dose, according to the pre-therapeu- Average values calculated from data reported by: Salvatori M,
tic MIBG scans. This threshold can be exceeded if bone Cremonesi M, Indovina L, Chianelli M, Pacilio M, Chiesa C, Zanzonico
marrow stem cell support is available. A total of 4.0 Gy P. Radiobiology and radiation dosimetry in nuclear medicine. In:
whole-body dose with stem cell rescue has been given with Strauss HW, Mariani G, Volterrani D, Larson M, eds. Nuclear
Oncology – From Pathophysiology to Clinical Applications. New York:
good tolerance and no other short-term, dose-limiting organ Springer; 2017:305–49
toxicity [137].
Administration of fixed activity “fractions” of 3.7–
S( WB¬ WB) = 1.34 ´ 10 -4 mp -0.921 (11.15)
11.1 GBq offers the advantages of simplicity and shorter iso-
lation and hospitalization period required for radioprotection;
this feature is non-negligible when treating very sick chil- where mP is the patient’s mass in kilograms. This equation
dren with poor prognosis. was generated by interpolating the S(WB←WB) values from the
Administration of high activities of 131I-MIBG based on MIRD phantoms for a newborn as well as for 1-year-,
body weight (555–777 MBq/kg), with stem cell support if 5-year-, 10-year-, and 15-year-old children, and finally for an
necessary, has also been reported. A maximum tolerated adult, each of whom has a specific mass. Table 11.4 reports
activity of 444 MBq/kg is reported, combined with myeloab- the absorbed doses to the organs calculated per unit of
lative chemotherapy and autologous stem cell transplanta- 131
I-MIBG activity administered.
tion [138]. An experimental MIBG protocol has been proposed in
However, when using fixed activities no dose-response which the aim was to deliver a total whole-body absorbed
assessment or optimization on the basis of absorbed radia- dose of 4.0 Gy in two fractions, in combination with topote-
tion dose is possible. Furthermore, the relatively LDR of can, after failure of induction chemotherapy [143]. Post-
radiation compared with (ultra-)high doses may be viewed as therapy dosimetry is performed after a first fixed fraction of
radiobiologically suboptimal. 444 MBq/kg, with calculation of the activity dose to be
The results obtained over more than two decades demon- administered with the subsequent second fraction in order to
strate that whole-body absorbed doses correlate both with achieve the desired total whole-body absorbed dose of 4 Gy
administered 131I-MIBG activity and with the subsequent [144].
development of hematologic toxicity [139, 140]. Recent studies in patients with high-risk neuroblastoma
An administered activity of 444–666 MBq/kg of analyzed the whole-body radiation dose (WBD) and tumor-
131
I-MIBG may deliver 50–700 cGy of whole-body absorbed absorbed dose from 131I-MIBG therapy in relation to tumor
dose; 80% of patients with advanced chemorefractory stage response and toxicities. These studies found that WBD cor-
III/IV neuroblastoma can develop grade 3 or 4 hematotoxic- related with 131I-MIBG activity, particularly with 131I-MIBG
ity at a whole-body absorbed dose of 2.5 Gy established administered per kilogram. These findings suggest that activ-
from a pre-therapy MIBG scan. ity prescriptions for 131I-MIBG therapies should be made
In relapsed or refractory neuroblastoma, the whole-body based on patient weight as opposed to a predetermined total
absorbed dose calculated according to the standard MIRD activity dose in order to achieve a targeted WBD. A proposed
schema can be prescribed accurately and allows higher activ- equation, which describes whole-body absorbed dose per
ities of 131I-MIBG to be administered within the safety limits unit of administered activity as a function of patient mass,
for bone marrow toxicity [141]. can be used as an alternative for prescriptions of activity on
The mean absorbed dose, D, is given by the product of the first administration when dosimetry data for the individual
cumulated activity and the MIRD whole body to whole-body patient are unknown [145]. However, no relationship between
S value, that is [142]: WBD and overall response was found, and no correlation
 was established between hematologic toxicity and WBD.
D( WB¬ WB) = A ´ S( WB¬ WB) (11.14) George et al. [146] evaluated the response, toxicity, and

long-term outcome of 131I-MIBG therapy in the treatment of
Value S(WB←WB) is often obtained using the following equa- refractory or relapsed neuroblastoma following a dosimetry-
tion (correction for body weight): based 3D individualized approach by using SPECT acquisi-
tions performed on consecutive days following the treatment. The pathophysiology of bone pain is not well understood,
The image-based 3D dosimetry application provided a and multiple mechanisms are postulated, including tumor-
tumor-absorbed dose map of consecutive therapies from induced cytokines, stimulating factors released by tumor
which dose volume histograms (DVH) were derived. Because cells, direct nerve injury, and infiltration of the bone trabecu-
dose heterogeneity could affect the outcome of subsequent lae and matrix by tumor cells causing osteolysis [150].
therapies, such 3D dosimetry approach where DVHs are The appropriate management of painful skeletal metasta-
taken together with tumor-absorbed dose as well as dose- sis includes the use of systemic analgesics, hormones, che-
limiting criteria could provide a reasonable method to help motherapeutic agents, steroids, external beam radiation
plan patient-specific treatment. therapy (EBRT), radiofrequency ablation, local surgery, and
A dosimetric approach based on patient’s pharmacokinet- radiopharmaceuticals [149]. These have shown their value in
ics may often result in the administration of higher activities the management of painful bone metastasis in clinical prac-
with excellent response rates while keeping toxicity within tice, although are infrequently used by many physicians,
acceptable limits. However, to date there are no published even those working in the fields of oncology and nuclear
randomized controlled trials of 131I-MIBG therapy for neuro- medicine [149, 151, 152].
blastoma at any stage of treatment [14]. To maximize the Phosphorus-32 (32P-ortophosphate) and strontium-89
therapeutic potential of 131I-MIBG therapy and to determine ( Sr-chloride) were the first bone-seeking radiopharmaceu-
89
its place within the patient pathway, well-designed clinical ticals approved for the treatment of painful bone metastases
trials incorporating dosimetry are needed. [153], while bisphosphonates labeled with either 153Sm
In the future, a more accurate assessment of the relationship [154], 186Re [155], or 188Re [156] are newer among the tradi-
between WBD, tumor-absorbed doses, and overall response tional bone-seeking radionuclides [157].
tumor dosimetry could be obtained by using 124I-MIBG with Although there are some differences in the physical half-
PET/CT imaging or 124I-MIBG-PET/MR imaging performed life, beta energy, penetration range, and biochemical
with an integrated PET/MRI system [147, 148]. characteristics among the various bone-seeking radiopharma-
ceuticals, no clear advantage in terms of increased response
rate has emerged. A uniform response rate of approximately
70% has been reported with all bone-seeking radiopharmaceu-
Key Learning Points
ticals, a response rate that appears to be inversely correlated
• Treatment with 131I-MIBG is indicated for malig-
with disease extent and Karnofsky index [149, 151, 152].
nant neuroectodermal tumors, including pheochro-
Several detailed dosimetric studies have been reported for
mocytoma, paraganglioma, carcinoid tumors,
bone-seeking radioisotopes. All the radiopharmaceuticals have
medullary thyroid cancer, and neuroblastoma.
shown an approximate tenfold therapeutic ratio between metas-
• Most treatments are performed in children with
tasis and bone, although there is an order of magnitude differ-
neuroblastoma arising in the nervous system or in
ence in absorbed dose calculations (5–50 Gy) between
the abdomen near the adrenal glands.
individual metastases [149, 152, 157]. Radiation dosimetry for
• A dosimetric approach based on patient’s pharma- 89
Sr-chloride, 153Sm-lexidronam (153Sm-EDTMP), and
cokinetics may often result in the administration of 186
Re-etidronate (186Re-HEDP) is reported in Table 11.5 [158].
higher activities with excellent response rates while
Hematological toxicity is the dose-limiting factor, usually
keeping toxicity within acceptable limits.
presenting as thrombocytopenia. Although there is a clear
• In the future, a more accurate assessment of the
relationship between metastatic burden and toxicity, the cor-
relationship between whole-body radiation dose
relation between thrombocytopenia and absorbed dose cal-
(WBD), tumor-absorbed doses, and overall response
culations is difficult to be noticed [149] even though there
tumor dosimetry could be obtained by using
have been some exceptions [159].
124
I-MIBG with PET/CT imaging or 124I-MIBG-
At present, the bone-seeking agent of choice has not yet
PET/MR imaging performed with an integrated
been determined. Since all the commonly used radiopharma-
PET/MRI system.
ceuticals have similar efficacy profiles, the agent should be
selected in a case-based fashion taking into consideration the
availability, toxicity, and goal of therapy.
It is intriguing that all studies that have searched for a
11.3.2.6 Treatment of Metastatic Bone Pain dose–response relationship have failed to show a correlation
Bone pain due to osseous metastases constitutes a relevant between absorbed dose to metastasis and clinical response.
clinical problem, especially in patients with breast and pros- Some of these calculations have shown clinical responses at
tate carcinoma that are responsible for more than 80% of the absorbed doses for which classical radiobiological para-
cases with bone metastases [149]. digms would predict no likelihood of response. Furthermore,
Table 11.5 Radiation dosimetry of bone-seeking radiopharmaceuti-

cals used for treatment of refractory metastatic bone pain • Recently, the use of the 223Ra-dichloride, a calcium
Absorbed dose per unit of radioactivity (mGy/ mimetic, short-lived alpha emitter (half-life
Organ/tissue MBq)
11.4 days), has been approved. The high LET of
89
Sr-Chloride 153
Sm-EDTMP 186Re-HEDP
alpha radiation gives rise to a greater biological
Bone surfaces 17.0 6.8 1.4
Red bone 11.0 1.5 1.3
effectiveness than beta radiation produced by other
marrow radiopharmaceuticals.
Lower bowel 4.7 0.01 0.57 • The use of 223Ra-dichloride in patients with symp-
Bladder 1.3 1.0 0.54 tomatic bone metastases by metastatic castration-
Kidneys 0.8 0.02 1.5 resistant (hormone-refractory) prostate cancer
Testes 0.8 0.005 0.008 (CRPC) showed a long life expectancy.
Ovaries 0.8 0.009 0.019
Reproduced with modifications from: Salvatori M, Cremonesi M,
Indovina L, Chianelli M, Pacilio M, Chiesa C, Zanzonico P. Radiobiology
and radiation dosimetry in nuclear medicine. In: Strauss HW, Mariani
G, Volterrani D, Larson M, eds. Nuclear Oncology – From
Pathophysiology to Clinical Applications. New York: Springer; References
2017:305–49
1. Amis ES Jr, Butler PF, Applegate KE, Birnbaum SB, Brateman
LF, Hevezi JM, Mettler FA, Morin RL, Pentecost MJ, Smith
no correlations have been found with primary endpoints, GG, Strauss KJ, Zeman RK. American College of Radiology
such as overall survival. white paper on radiation dose in medicine. J Am Coll Radiol.
Recently, 223Ra-dichloride (Xofigo© has been approved 2007;4:272–84.
2. Mettler FA Jr, Bhargavan M, Faulkner K, Gilley DB, Gray JE,
for patients with symptomatic bone metastases due to meta- Ibbott GS, Lipoti JA, Mahesh M, McCrohan JL, Stabin MG,
static castration-resistant (hormone-refractory) prostate can- Thomadsen BR, Yoshizumi TT. Radiologic and nuclear medicine
cer (CRPC) and no known visceral metastases. Using this studies in the United States and worldwide: frequency, radiation
bone-seeking agent, clinical studies showed a life expec- dose, and comparison with other radiation sources—1950–2007.
Radiology. 2009;253:520–31.
tancy longer than that of patients in the placebo group, delay- 3. Brenner DJ, Hall EJ. Computed tomography—an increasing
ing the start of bone fractures and pain [160]. source of radiation exposure. N Engl J Med. 2007;357:2277–84.
223
Ra is a calcium mimetic, half-life 11.4 days alpha emit- 4. Sgouros G. Dosimetry of internal emitters. J Nucl Med.
ter. The high LET of alpha radiation gives rise to a greater 2005;46(Suppl 1):18S.
5. Murray D, Mc Ewan AJ. Radiobiology of systemic radiation ther-
biological effectiveness than beta radiation and to cytotoxic- apy. Cancer Biother Radiopharm. 2007;22:1–23.
ity that is independent of dose rate, cell cycle growth phase, 6. Dale R, Carabe-Fernandez A. The radiobiology of conventional
and oxygen concentration [161]. The range of alpha particles radiotherapy and its application to radionuclide therapy. Cancer
(<100 μm) is much smaller than that of beta radiation, caus- Biother Radiopharm. 2005;20:47–51.
7. UNSCEAR 2000 Report. Volume I: Sources Annex D Medical
ing less hematological toxicity for a given absorbed dose to radiation exposures.
the bone surface than beta emitters [162]. 8. NCRP Report 160. Ionizing Radiation Exposure of the Population
Actually, several trials are ongoing for dose-escalation of the United States. Released in March 2009.
and combination therapy (e.g., with hormone therapy). 223Ra 9. Loevinger R, Budinger TF, Watson EE. MIRD primer for absorbed
dose calculations. Revised ed. New York, NY: The Society of
emits also a low number of photons useful for imaging; thus, Nuclear Medicine; 1991.
in vivo image-based dosimetry is feasible, allowing also 10. Snyder WS, et al. MIRD pamphlet no. 11—“S” absorbed dose
investigations on a macrodosimetric scale [163–165], as well per unit cumulated activity for selected radionuclides and organs.
as through the traditional microdosimetric approach [161]. New York, NY: Society of Nuclear Medicine; 1975.
11. Baechler S, Hobbs RF, Prideaux AR, et al. Extension of the biolog-
ical effective dose to the MIRD schema and possible implications
in radionuclide therapy dosimetry. Med Phys. 2008;35:1123–34.
Key Learning Points 12. Siegel J. MIRD pamphlet no 16 — techniques for quantita-
• Various bone-seeking radiopharmaceuticals can be tive radiopharmaceutical biodistribution data acquisition and
analysis for use in human radiation dose estimates. J Nucl Med.
used for the palliative treatment of bone pain due to 1999;40:37S–61S.
osseous metastases. 13. Bolch WE, et al. MIRD pamphlet no 17: the dosimetry of non-
• All the radiopharmaceuticals have shown an uniform activity distributions—radionuclide S values at the voxel
approximate tenfold therapeutic ratio between level. J Nucl Med. 1999;40:11S–36S.
14. Stabin MG. MIRDOSE: personal computer software for
metastasis and bone, although there is an order of internal dose assessment in nuclear medicine. J Nucl Med.
magnitude difference in absorbed dose calculations 1996;37:538–46.
(5–50 Gy) between individual metastases. 15. Harrison JD, Streffer C. The ICRP protection quantities, equiva-
lent and effective dose: their basis and application. Radiat Prot
Dosimetry. 2007;127(1–4):12–8.
16. Cristy M, Eckerman K. Specific absorbed fractions of energy at 35. Miller JC. Hybrid SPECT/CT: enhancing nuclear medicine.
various ages from internal photon sources (I-VII). Oak Ridge Radiol Rounds. 2010;8(7).
National Laboratory Report ORNL/TM-8381/V1-7. Springfield, 36. Brix G, Lechel U, Glatting G, Ziegler SI, Munzing W, Muller
VA: National Technical Information Service, Dept of Commerce; SP, Beyer T. Radiation exposure of patients undergoing whole-
1987. body dual-modality 18F-FDG PET/CT examinations. J Nucl Med.
17. Stabin MG, Stubbs JB, Toohey RE. Radiation dose estimates for 2005;46:608–13.
radiopharmaceuticals. Washington, DC: US Nuclear Regulatory 37. Groves AM, Owen KE, Courtney HM, Yates SJ, Goldstone KE,
Commission (NRC); 1996.. Report No.: NUREG/CR-6345. Blake GM, Dixon AK. 16-detector multislice CT: dosimetry esti-
18. Preston DL, Pierce DA, Shimizu Y, Cullings HM, Fujita S, mation by TLD measurement compared with Monte Carlo simula-
Funamoto S, Kodama K. Effect of recent changes in atomic bomb tion. Br J Radiol. 2004;77:662–5.
survivor dosimetry on cancer mortality risk estimates. Radiat Res. 38. Huda W, Vance A. Patient radiation doses from adult and pediatric
2004;162:377–89. CT. AJR Am J Roentgenol. 2007;188:540–6.
19. NRC. National Research Council, Committee on the biological 39. McBride WH, Chiang CS, Olson JL, et al. A sense of danger from
effects of ionizing radiation (BEIR V), health effects of exposure radiation. Radiat Res. 2004;162:1–19.
to low levels of ionizing radiations. Washington, DC: National 40. Erenpreisa J, Cragg MS. Mitotic death: a mechanism of survival?
Academy Press; 1990. A review. Cancer Cell Int. 2001;1:1.
20. NRC. National Research Council, Committee on the biological 41. De Nardo GL, Schlom J, Buchsbaum DJ, et al. Rationales, evi-
effects of ionizing radiation (BEIR VII-phase 2), health risks from dence, and design considerations for fractionated radioimmuno-
exposure to low levels of ionizing radiation. Washington, DC: therapy. Cancer. 2002;94(Suppl 4):1332.
National Academy Press; 2006. 42. Zanzonico P, Stabin MG. Quantitative benefit-risk analysis of
21. Preston DL, Ron E, Tokuoka S, Funamoto S, Nishi N, Soda M, medical radiation exposures. Semin Nucl Med. 2014;44:210–4.
Mabuchi K, Kodama K. Solid cancer incidence in atomic bomb 43. van Tinteren H, Hoekstra O, Smit EF, et al. Effectiveness of
survivors: 1958–1998. Radiat Res. 2007;168:1–64. positron emission tomography in the preoperative assessment of
22. Brenner DJ, Elliston CD, Hall EJ, Berdon WE. Estimates of the patients with suspected non-small-cell lung cancer: the PLUS
cancer risks from pediatric CT radiation are not merely theoreti- multicentre randomised trial. Lancet. 2002;359(9315):1388–92.
cal: comment on “point/counterpoint: in X-ray computed tomog- 44. Mansel RE, Fallowfield L, Kissin M, et al. Randomized mul-
raphy, technique factors should be selected appropriate to patient ticenter trial of sentinel node biopsy versus standard axillary
size against the proposition”. Med Phys. 2001;28:2387–8. treatment in operable breast cancer: the ALMANAC trial. J Natl
23. Treves ST, Davis RT, Fahey FH. Administered radiopharmaceuti- Cancer Inst. 2006;98:599–609.
cal doses in children: a survey of 13 pediatric hospitals in North 45. Gojer B, Williams KA. The role of scintigraphic perfusion imag-
America. J Nucl Med. 2008;49(6):1024–7. ing for predicting ischemic cardiac events in noncardiac surgery.
24. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of Am J Card Imaging. 1995;9:213–25.
radiation-induced fatal cancer from pediatric CT. AJR Am J 46. Kassis AI, Adelstein SJ. Radiobiologic principles in radionuclide
Roentgenol. 2001;176:289–96. therapy. J Nucl Med. 2005;46(Suppl 1):4S.
25. Einstein AJ, Henzlova MJ, Rajagopalan S. Estimating risk of can- 47. Roninson IB. Tumor senescence as a determinant of drug response
cer associated with radiation exposure from 64-slice computed in vivo. Drug Resist Updat. 2002;5:204.
tomography coronary angiography. JAMA. 2007;298:317–23. 48. Stabin MG, Sparks RB, Crowe E. OLINDA/EXM: the second-
26. Martin CJ. Effective dose: how should it be applied to medical generation personal computer software for internal dose assess-
exposures? Br J Radiol. 2007;80:639–47. ment in nuclear medicine. J Nucl Med. 2005;46:1023–7.
27. Ferrari M, De Marco P, Origgi D, Pedroli G. SPECT/CT radiation 49. Sgouros G. Treatment planning for internal emitter therapy: meth-
dosimetry. Clin Transl Imag. 2014;2:557–69. ods, applications, and clinical implications. In: Stelson A, Stabin
28. Brix G, Nekolla EA, Borowski M, Noßke D. Radiation risk and M, Sparks R, editors. Sixth international radiopharmaceutical
protection of patients in clinical SPECT/CT. Eur J Nucl Med Mol dosimetry symposium. Gatlinburg, TN: Oak Ridge Associated
Imaging. 2014;41(Suppl 1):S125–36. Unversities; 1999. p. 13–25.
29. Fahey FH, Palmer MR, Strauss KJ, Zimmerman RE, Badawi RD, 50. Tagesson M, et al. The SIMDOS Monte Carlo code for conversion
Treves ST. Dosimetry and adequacy of CT-based attenuation correc- of activity distributions to absorbed dose and dose-rate distribu-
tion for pediatric PET: phantom study. Radiology. 2007;243:96–104. tions. In: Stelson A, Stabin M, Sparks R, editors. Sixth interna-
30. Hulme KW, Kappadath SC. Implications of CT noise and arti- tional radiopharmaceutical dosimetry symposium. Gatlinburg,
facts for quantitative 99mTc SPECT/CT imaging. Med Phys. TN: Oak Ridge Associated Unversities; 1999. p. 425–40.
2014;41:1–10. 51. Brans B, Bodei L, Giammarile F, et al. Clinical radionuclide ther-
31. Buck AK, Nekolla S, Ziegler S, Beer A, Krause BJ, Herrmann K, apy dosimetry: the quest for the “Holy Gray”. Eur J Nucl Med
Scheidhauer K, Wester HJ, Rummeny EJ, Schwaiger M, Drzezga Mol Imaging. 2007;34:772–86.
A. SPECT/CT. J Nucl Med. 2008;49:1305–19. 52. Maxon HR, Smith HS. Radioiodine-131 in the diagnosis and treat-
32. Roach PJ, Schembri GP, Ho Shon IA, Bailey EA, Bailey ment of metastatic well differentiated thyroid cancer. Endocrinol
DL. SPECT/CT imaging using a spiral CT scanner for anatomi- Metab Clin North Am. 1990;19:685–718.
cal localization: impact on diagnostic accuracy and reporter confi- 53. Schlumberger M, Pacini F. Thyroid tumors. 5th ed. Paris: Editions
dence in clinical practice. Nucl Med Commun. 2006;27:977–87. Nucléon; 2003.
33. Sawyer LJ, Starritt HC, Hiscock SC, Evans MJ. Effective doses 54. Van Nostrand D, Atkins F, Yeganeh F, et al. Dosimetrically deter-
to patients from CT acquisitions on the GE Infinia Hawkeye: mined doses of radioiodine for the treatment of metastatic thyroid
a comparison of calculation methods. Nucl Med Commun. carcinoma. Thyroid. 2002;12:121–34.
2008;29:144–9. 55. Lassmann M, Hänscheid H, Chiesa C, et al. EANM Dosimetry
34. Montes C, Tamayo P, Hernandez J, et al. Estimation of the total Committee series on standard operational procedures for
effective dose from low-dose CT scans and radiopharmaceuti- pre-therapeutic dosimetry I: blood and bone marrow dosimetry
cal administrations delivered to patients undergoing SPECT/CT in differentiated thyroid cancer therapy. Eur J Nucl Med Mol
explorations. Ann Nucl Med. 2013;27:610–7. Imaging. 2008;35:1405–12.
56. Benua RS, Cicale NR, Sonenberg M, et al. Relation of radioio- 74. Sandström M, Garske-Román U, Granberg D, et al. Individualized
dine dosimetry to results and complications in treatment of meta- dosimetry of kidney and bone marrow in patients undergoing
static thyroid cancer. Am J Roentgenol Radium Ther Nucl Med. 177
Lu-DOTA-octreotate treatment. J Nucl Med. 2013;54:33–41.
1962;87:171–82. 75. Sandström M, Ilan E, Karlberg A, et al. Method dependence,
57. Maxon HR. Quantitative radioiodine therapy in the treatment of observer variability and kidney volumes in radiation dosimetry
differentiated thyroid cancer. Q J Nucl Med. 1999;43:313–23. of 177Lu-DOTATATE therapy in patients with neuroendocrine
58. Tuttle RM, Leboeuf R, Robbins RJ, Qualey R, Pentlow K, Larson tumours. EJNMMI Phys. 2015;2:24.
SM, et al. Empiric radioactive iodine dosing regimens frequently 76. Gustafsson J, Brolin G, Cox M, et al. Uncertainty propagation for
exceed maximum tolerated activity levels in elderly patients with SPECT/CT-based renal dosimetry in 177Lu peptide receptor radio-
thyroid cancer. J Nucl Med. 2006;47:1587–91. nuclide therapy. Phys Med Biol. 2015;60:8329–46.
59. Traino AC, Ferrari M, Cremonesi M, et al. Influence of total body 77. Ljungberg M, Celler A, Konijnenberg MW, Eckerman KF,
mass on the scaling of S-factors for patient-specific, blood-based Dewaraja YK, Sjögreen Gleisner K. MIRD Pamphlet No. 26:
red-marrow dosimetry. Phys Med Biol. 2007;52:5231–48. Joint EANM/MIRD guidelines for quantitative 177Lu SPECT
60. Sgouros G, Song H, Ladenson PW, et al. Lung toxicity in radio- applied for dosimetry of radiopharmaceutical therapy. J Nucl
iodine therapy of thyroid carcinoma: development of a dose-rate Med. 2016;57:151–62.
method and dosimetric implications of the 80-mCi rule. J Nucl 78. Dale R. Use of the linear-quadratic radiobiological model for
Med. 2006;47:1977–84. quantifying kidney response in targeted radiotherapy. Cancer
61. Song H, Prideaux A, Du Y, et al. Lung dosimetry for radioiodine Biother Radiopharm. 2004;19:363–70.
treatment planning in the case of diffuse lung metastases. J Nucl 79. Baechler S, Hobbs RF, Prideaux AR, Wahl RL, Sgouros
Med. 2006;47:1985–94. G. Extension of the biological effective dose to the MIRD schema
62. Luster M, Clarke SE, Dietlein M, Lassmann M, Lind P, Oyen and possible implications in radionuclide therapy dosimetry. Med
WJG, Tennvall J, Bombardieri E. Guidelines for radioiodine ther- Phys. 2008;35:1123–34.
apy of differentiated thyroid cancer. Eur J Nucl Med Mol Imaging. 80. Strigari L, Benassi M, Chiesa C, Cremonesi M, Bodei L, D’Andrea
2008;35:1941–59. M. Dosimetry in nuclear medicine therapy: radiobiology applica-
63. Freudenberg LS, Jentzen W, Stahl A, Bockisch A, Rosenbaum- tion and results. Q J Nucl Med Mol Imaging. 2011;55:205–21.
Krumme SJ. Clinical applications of 124I-PET/CT in patients 81. Barone R, Borson-Chazot F, Valkema R, et al. Patient-specific
with differentiated thyroid cancer. Eur J Nucl Med Mol Imaging. dosimetry in predicting renal toxicity with 90Y-DOTATOC: rel-
2011;38(Suppl 1):S48–56. evance of kidney volume and dose rate in finding a dose-effect
64. Sgouros G, Kolbert KS, Sheikh A, et al. Patient-specific dosim- relationship. J Nucl Med. 2005;46(Suppl 1):99S–106S.
etry for I-131 thyroid cancer therapy using I-124 PET and 82. Wessels BW, Konijnenberg MW, Dale RG, Breitz HB, Cremonesi
3-dimensional internal dosimetry (3D-ID) software. J Nucl Med. M, Meredith RF, et al. MIRD pamphlet no. 20: the effect of model
2004;45:1366–72. assumptions on kidney dosimetry and response-implications for
65. de Keizer B, Brans B, Hoekstra A, et al. Tumour dosimetry and radionuclide therapy. J Nucl Med. 2008;49:1884–99.
response in patients with metastatic differentiated thyroid cancer 83. Strigari L, Konijnenberg M, Chiesa C, et al. The evidence base for
using recombinant human thyrotropin before radioiodine therapy. the use of internal dosimetry in the clinical practice of molecular
Eur J Nucl Med. 2003;30:367–73. radiotherapy. Eur J Nucl Med Mol Imaging. 2014;41:1976–88.
66. Jentzen W, Bockisch A, Ruhlmann M. Assessment of simplified 84. Ilan E, Sandström M, Wassberg C, et al. Dose response of pancreatic
blood dose protocols for the estimation of the maximum tolerable neuroendocrine tumors treated with peptide receptor radionuclide
activity in thyroid cancer patients undergoing radioiodine therapy therapy using 177Lu-DOTATATE. J Nucl Med. 2015;56:177–82.
using 124I. J Nucl Med. 2015;56:832–8. 85. Minarik D, Sjögreen Gleisner K, Ljungberg M. Evaluation of
67. Freudenberg LS, Jentzen W, Görges R, Petrich T, Marlowe RJ, quantitative 90Y SPECT based on experimental phantom studies.
Knust J, et al. 124I-PET dosimetry in advanced differentiated thy- Phys Med Biol. 2008;53:5689–703.
roid cancer: therapeutic impact. Nuklearmedizin. 2007;46:121–8. 86. Minarik D, Ljungberg M, Segars P, Gleisner KS. Evaluation of
68. van Essen M, Krenning EP, Kam BL, et al. Peptide-receptor quantitative planar 90Y bremsstrahlung whole-body imaging. Phys
radionuclide therapy for endocrine tumors. Nat Rev Endocrinol. Med Biol. 2009;54:5873–83.
2009;5:382–93. 87. Fabbri C, Sarti G, Cremonesi M, et al. Quantitative analysis of 90Y
69. Bodei L, Cremonesi M, Kidd M, et al. Peptide receptor radionu- Bremsstrahlung SPECT-CT images for application to 3D patient-
clide therapy for advanced neuroendocrine tumors. Thorac Surg specific dosimetry. Cancer Biother Radiopharm. 2009;24:145–54.
Clin. 2014;24:333–49. 88. Cremonesi M, Botta F, Di Dia A, et al. Dosimetry for treatment
70. Garkavij M, Nickel M, Sjögreen-Gleisner K, Ljungberg M, with radiolabelled somatostatin analogues. A review. Q J Nucl
Ohlsson T, Wingårdh K, Strand SE, Tennvall J. 177Lu-[DOTA0,Tyr3] Med Mol Imaging. 2010;54:37–51.
octreotate therapy in patients with disseminated neuroendocrine 89. Sandström M, Garske U, Granberg D, et al. Individualized dosim-
tumors: analysis of dosimetry with impact on future therapeutic etry in patients undergoing therapy with 177Lu-DOTA-D-Phe1-
strategy. Cancer. 2010;116(4 Suppl):1084–92. Tyr3-octreotate. Eur J Nucl Med Mol Imaging. 2010;37:212–25.
71. Swärd C, Bernhardt P, Ahlman H, et al. [177Lu-DOTA0-Tyr3]- 90. Bodei L, Mueller-Brand J, Baum RP, et al. The joint IAEA,
octreotate treatment in patients with disseminated gastroen- EANM, and SNMMI practical guidance on peptide receptor
teropancreatic neuroendocrine tumors: the value of measuring radionuclide therapy (PRRNT) in neuroendocrine tumours. Eur J
absorbed dose to the kidney. World J Surg. 2010;34:1368–72. Nucl Med Mol Imaging. 2013;40:800–16.
72. Lassmann M, Chiesa C, Flux G, et al. EANM Dosimetry 91. Sarnelli A, Guerriero F, Botta F, et al. Therapeutic schemes in
Committee guidance document: good practice of clinical dosim- 177
Lu and 90Y-PRRT: radiobiological considerations. Q J Nucl
etry reporting. Eur J Nucl Med Mol Imaging. 2011;38:192–200. Med Mol Imaging. 2017;61:216–31.
73. Larsson M, Bernhardt P, Svensson JB, et al. Estimation of 92. Rolleman EJ, Melis M, Valkema R, Boerman OC, Krenning EP, de
absorbed dose to the kidneys in patients after treatment with Jong M. Kidney protection during peptide receptor radionuclide
177
Lu-octreotate: comparison between methods based on planar therapy with somatostatin analogues. Eur J Nucl Med Mol
scintigraphy. EJNMMI Res. 2012;2:49. Imaging. 2010;37:1018–31.
93. Bouchet LG, Bolch WE, Blanco HP, et al. MIRD Pamphlet No 19: with 90Y-radiolabelled ibritumomab tiuxetan (Zevalin). Eur J
absorbed fractions and radionuclide S values for six age-dependent Nucl Med Mol Imaging. 2007;34:616–22.
multiregion models of the kidney. J Nucl Med. 2003;44:1113–47. 112. Wiseman GA, Kornmehl E, Leigh B, et al. Radiation dosimetry
94. Green A, Flynn A, Pedley RB, Dearling J, Begent R. Nonuniform results and safety correlations from 90Y-ibritumomab tiuxetan
absorbed dose distribution in the kidney: the influence of organ radioimmunotherapy for relapsed or refractory non-Hodgkin’s
architecture. Cancer Biother Radiopharm. 2004;19:371–7. lymphoma: combined data from 4 clinical trials. J Nucl Med.
95. Bodei L, Cremonesi M, Ferrari M, et al. Long-term evaluation 2003;44:465–74.
of renal toxicity after peptide receptor radionuclide therapy with 113. Fisher DR, Shen S, Meredith RF. MIRD dose estimate report
90
Y-DOTATOC and 177Lu-DOTATATE: the role of associated risk No. 20: radiation absorbed-dose estimates for 111In- and
factors. Eur J Nucl Med Mol Imaging. 2008;35:1847–56. 90
Y-ibritumomab tiuxetan. J Nucl Med. 2009;50:644–52.
96. Imhof A, Brunner P, Marincek N, et al. Response, survival, and 114. Bodei L, Cremonesi M, Grana C, et al. Receptor radionuclide
long-term toxicity after therapy with the radiolabeled somatostatin therapy with 90Y-[DOTA]0-Tyr3-octreotide (90Y-DOTATOC)
analogue [90Y-DOTA]-TOC in metastasized neuroendocrine can- in neuroendocrine tumours. Eur J Nucl Med Mol Imaging.
cers. J Clin Oncol. 2011;29:2416–23. 2004;31:1038–46.
97. Cremonesi M, Ferrari M, Di Dia A, Botta F, De Cicco C, Bodei L, 115. Cremonesi M, Ferrari M, Grana CM, et al. High-
Paganelli G. Recent issues on dosimetry and radiobiology for pep- doseradioimmunotherapy with 90Y-Ibritumomab Tiuxetan: com-
tide receptor radionuclide therapy. Q J Nucl Med Mol Imaging. parative dosimetric study for tailored treatment. J Nucl Med.
2011;55:155–67. 2007;48:1871–9.
98. Cybulla M, Weiner SM, Otte A. End-stage renal disease after 116. Lewandowski RJ, Riaz A, Ryu RK, et al. Optimization of radio-
treatment with 90Y-DOTATOC. Eur J Nucl Med. 2001;28:1552–4. embolic effect with extended-shelf-life yttrium-90 microspheres:
99. Kwekkeboom DJ, Mueller-Brand J, Paganelli G, et al. Overview results from a pilot study. J Vasc Interv Radiol. 2009;20:1557–63.
of results of peptide receptor radionuclide therapy with 3 radio- 117. Kennedy AS, McNeillie P, Dezarn WA, et al. Treatment parame-
labeled somatostatin analogs. J Nucl Med. 2005;46(Suppl ters and outcome in 680 treatments of internal radiation with resin
1):62S–6S. 90
Y-microspheres for unresectable hepatic tumors. Int J Radiat
100. Bodei L, Cremonesi M, Grana CM, et al. Peptide receptor radio- Oncol Biol Phys. 2009;74:1494–500.
nuclide therapy with 177Lu-DOTATATE: the IEO phase I-II study. 118. Cremonesi M, Chiesa C, Strigari L, Ferrari M, Botta F, Guerriero
Eur J Nucl Med Mol Imaging. 2011;38:2125–35. F, et al. Radioembolization of hepatic lesions from a radiobiology
101. Kwekkeboom DJ, de Herder WW, Kam BL, et al. Treatment and dosimetric perspective. Front Oncol. 2014;4:210.
with the radiolabeled somatostatin analog [177Lu-DOTA0,Tyrd] 119. Gulec SA, Sztejnberg ML, Siegel JA, et al. Hepatic structural
octreotate: toxicity, efficacy, and survival. J Clin Oncol. dosimetry in 90Y microsphere treatment: a Monte Carlo mod-
2008;26:2124–30. eling approach based on lobular microanatomy. J Nucl Med.
102. Bodei L, Kidd M, Paganelli G, et al. Long-term tolerability of 2010;51:301–10.
PRRT in 807 patients with neuroendocrine tumours: the value 120. Kennedy AS, McNeillie P, Dezarn WA, Nutting C, Sangro B,
and limitations of clinical factors. Eur J Nucl Med Mol Imaging. Wertman D, et al. Treatment parameters and outcome in 680
2015;42:5–19. treatments of internal radiation with resin 90Y-microspheres
103. Kunikowska J, Królicki L, Hubalewska-Dydejczyk A, et al. for unresectable hepatic tumors. Int J Radiat Oncol Biol Phys.
Clinical results of radionuclide therapy of neuroendocrine 2009;74:1494–500.
tumours with 90Y-DOTATATE and tandem 90Y/177Lu-DOTATATE: 121. Ferrari M, Cremonesi M, Bartolomei M, et al. Different evalu-
which is a better therapy option? Eur J Nucl Med Mol Imaging. ations for safe 90Y-microspheres in selective internal radia-
2011;38:1788–97. tion therapy (SIRT). Eur J Nucl Med Mol Imaging. 2007;34(2
104. Seregni E, Maccauro M, Chiesa C, et al. Treatment with tandem Suppl):S178.
[90Y]DOTA-TATE and [177Lu]DOTA-TATE of neuroendocrine 122. Ho S, Lau WY, Leung TW, et al. Partition model for estimating
tumours refractory to conventional therapy. Eur J Nucl Med Mol radiation doses from yttrium-90 microspheres in treating hepatic
Imaging. 2014;41:223–30. tumours. Eur J Nucl Med. 1996;23:947–52.
105. Goldsmith SJ. Radioimmunotherapy of lymphoma: Bexxar and 123. Salem R, Thurston KG. Radioembolization with 90Yttrium micro-
Zevalin. Semin Nucl Med. 2010;40:122–35. spheres: a state-of-the-art brachytherapy treatment for primary
106. Kaminski MS, Tuck M, Estes J, et al. 131I-tositumomab therapy and secondary liver malignancies. Part 1: technical and meth-
as initial treatment for follicular lymphoma. N Engl J Med. odologic considerations. J Vasc Interv Radiol. 2006;17:1251–78..
2005;352:441–9. Erratum in: J Vasc Interv Radiol. 2006;17:1594.
107. Wahl RL. Tositumomab and 131I therapy in non-Hodgkin’s lym- 124. Cremonesi M, Ferrari M, Bartolomei M, et al. Radioembolisation
phoma. J Nucl Med. 2005;46(Suppl 1):128S–40S. with 90Y-microspheres: dosimetric and radiobiological investi-
108. Witzig TE, Flinn IW, Gordon LI, et al. Treatment with ibritu- gation for multi-cycle treatment. Eur J Nucl Med Mol Imaging.
momab tiuxetan radioimmunotherapy in patients with rituximab- 2008;35:2088–96.
refractory follicular non-Hodgkin’s lymphoma. J Clin Oncol. 125. Riaz A, Lewandowski RJ, Kulik LM, et al. Complications follow-
2002;20:3262–9. ing radioembolization with yttrium-90 microspheres: a compre-
109. Aricò D, Grana CM, Vanazzi A, et al. The role of dosimetry hensive literature review. J Vasc Interv Radiol. 2009;20:1121–30.
in the high activity 90Y-ibritumomab tiuxetan regimens: two 126. Walrand S, Hesse M, Chiesa C, Lhommel R, Jamar F. The low
cases of abnormal biodistribution. Cancer Biother Radiopharm. hepatic toxicity per Gray of 90Y glass microspheres is linked to
2009;24:271–5. their transport in the arterial tree favoring a nonuniform trap-
110. Wagner HN, Wiseman GA, Marcus CS, et al. Aministration ping as observed in post-therapy PET imaging. J Nucl Med.
guidelines for radioimmunotherapy of non-Hodgkin’s lymphoma 2014;55:135–40.
with 90Y-labeled anti-CD20 monoclonal antibody. J Nucl Med. 127. Gaba RC, Lewandowsi RJ, Kulik LM, Riaz A, Ibrahim SM,
2002;43:267–72. Mulcahy MF. Radiation lobectomy: preliminary findings of
111. Tennvall J, Fischer M, Bischof Delaloye A, et al. EANM proce- hepatic volumetric response to lobar yttrium-90 radioemboliza-
dure guideline for radio-immunotherapy for B-cell lymphoma tion. Ann Surg Oncol. 2009;16:1587–96.
128. Strigari L, Sciuto R, Rea S, Carpanese L, Pizzi G, Soriani A, et al. 145. Minguez P, Flux G, Genolla J, Guayambuco S, Delgado
Efficacy and toxicity related to treatment of hepatocellular carci- A, Fombellida JC, et al. Dosimetric results in treat-
noma with 90Y SIR spheres: radiobiological considerations. J Nucl ments of neuroblastoma and neuroendocrine tumors with
Med. 2010;51:1377–85. 131
I-metaiodobenzylguanidine with implications for the activity to
129. Flamen P, Vanderlinden B, Delatte P, et al. CORRIGENDUM: administer. Med Phys. 2015;42:3969–78.
multimodality imaging can predict the metabolic response of 146. George SL, Falzone N, Chittenden S, Kirk SJ, Lancaster D, Vaidya
unresectable colorectel liver metastases to radioembolization ther- SJ, et al. Individualized 131I-mIBG therapy in the management
apy with yttrium-90 labeled resin micropheres. Phys Med Biol. of refractory and relapsed neuroblastoma. Nucl Med Commun.
2014;59:2549. 2016;37:466–72.
130. Chiesa C, Mira M, Maccauro M, Romito R, Spreafico C, Morosi 147. Seo Y, Gustafson WC, Dannoon SF, Nekritz EA, Lee CL, Murphy
C, et al. Radioembolization of hepatocarcinoma with 90-Y glass ST, et al. Tumor dosimetry using [124I]m-iodobenzylguanidine
microspheres: development of an individualized treatment plan- microPET/CT for [131I]m-iodobenzylguanidine treatment of
ning strategy based on dosimetry and radiobiology. Eur J Nucl neuroblastoma in a murine xenograft model. Mol Imaging Biol.
Med Mol Imaging. 2015;42:1718–38. 2012;14:735–42.
131. Garin E, Lenoir L, Rolland Y, Edeline J, Mesbah H, Laffont 148. Huang SY, Bolch WE, Lee C, Van Brocklin HF, Pampaloni MH,
S. Dosimetry based on 99mTc-macroaggregated albumin SPECT/ Hawkins RA, et al. Patient-specific dosimetry using pretherapy
CT accurately predicts tumor response and survival in hepato- [123I]m-iodobenzylguanidine ([123I]mIBG) dynamic PET/CT
cellular carcinoma patients treated with 90Y-loaded glass micro- imaging before [131I]mIBG targeted radionuclide therapy for neu-
spheres: preliminary results. J Nucl Med. 2012;53:255–63. roblastoma. Mol Imaging Biol. 2015;17:284–94.
132. Garin E, Lenoir L, Edeline J, Laffont S, Mesbah H, Porée P, et al. 149. Paes FM, Serafini AN. Systemic metabolic radiopharmaceutical
Boosted selective internal radiation therapy with 90Y-loaded glass therapy in the treatment of metastatic bone pain. Semin Nucl Med.
microspheres (B-SIRT) for hepatocellular carcinoma patients: 2010;40:89–104.
a new personalized promising concept. Eur J Nucl Med Mol 150. Matthay KK, QuachA, Franc BL, et al. 131I-Metaiodobenzylguanidine
Imaging. 2013;40:1057–68. (131I-MIBG) double infusion with autologous stem cell rescue
133. Willowson KP, Tapner M, The QUEST Investigator Team, for neuroblastoma: a new approach to neuroblastoma therapy
Bailey DL. A multicentre comparison of quantitative 90Y PET/ (NANT) phase I study. J Clin Oncol. 2009;27:1020–5.
CT for dosimetric purposes after radioembolization with resin 151. Krishnamurthy GT, Krishnamurthy S. Radionuclides for meta-
microspheres – the QUEST phantom study. Eur J Nucl Med Mol static bone pain palliation: a need for rational re-evaluation in the
Imaging. 2015;42:1202–22. new millennium. J Nucl Med. 2000;41:688–91.
134. Giammarile F, Chiti A, et al. EANM procedure guidelines for 152. Finlay IG, Mason MD, Shelley M. Radioisotopes for the pallia-
131
I-meta-iodobenzylguanidine (131I-mIBG) therapy. Eur J Nucl tion of metastatic bone cancer: a systematic review. Lancet Oncol.
Med Mol Imaging. 2008;35:1039–47. 2005;6:392–400.
135. Monsieurs M, Brans B, Bacher K, et al. Patient dosimetry for neu- 153. Taylor AJ Jr. Strontium-89 for the palliation of bone pain due to
roendocrine tumours based on 123I-MIBG pretherapy scans and metastatic disease. J Nucl Med. 1994;35:2054.
131
I-MIBG post therapy scans. Eur J Nucl Med. 2002;29:1581–7. 154. Eary JF, Collins C, Stabin M, et al. Samarium-153-EDTMP
136. Fielding SL, Flower MA, Ackery D, et al. Dosimetry of iodine- biodistribution and dosimetry estimation. J Nucl Med.
131metaiodobenzylguanidine in resistant neuroblastoma: results 1993;34:1031–6.
of a UK Study. Eur J Nucl Med. 1991;18:308–16. 155. Graham MC, Scher HI, Liu GB, et al. Rhenium-186-labeled
137. Buckley SE, Chittenden S, Saran FH, et al. Whole-body dosimetry hydroxyethylidene diphosphonate dosimetry and dosing guide-
for individualized treatment planning of 131I-MIBG radionuclide lines for the palliation of skeletal metastases from androgen-
therapy for neuroblstoma. J Nucl Med. 2009;50:1518–24. independent prostate cancer. Clin Cancer Res. 1999;5:1307–18.
138. Matthay KK, Jessica C, Villablanca J, et al. Phase I dose escala- 156. Liepe K, Kropp J, Runge R, et al. Therapeutic efficiency of
tion of iodine-131-metaiodobenzylguanidine with myeloablative rhenium-188-HEDP in human prostate cancer skeletal metastases.
chemotherapy and autologous stem-cell transplantation in refrac- Br J Cancer. 2003;89:625–9.
tory neuroblastoma: a new approaches to neuroblastoma therapy 157. Lewington VJ. Bone-seeking radiopharmaceuticals. J Nucl Med.
consortium study. J Clin Oncol. 2006;24:500–6. 2005;46:38S–47S.
139. Monsieurs MA, Thierens HM, Vral A, et al. Patient dosim- 158. Bodei L, Lam M, Chiesa C, Flux G, Brans B, Chiti A, Giammarile
etry after I-131-MIBG therapy for neuroblastoma and carcinoid F. EANM procedure guideline for treatment of refractory meta-
tumours. Nucl Med Commun. 2001;22:367–74. static bone pain. Eur J Nucl Med Mol Imaging. 2008;35:1934–40.
140. DuBois SG, Messina J, Maris JM, et al. Haematological toxic- 159. Pacilio M, Ventroni G, Basile C, et al. Improving the dose–
ity of high-dose iodine-131-metaiodobenzylguanidine therapy for myelotoxicity correlation in radiometabolic therapy of bone
advanced neuroblastoma. J Clin Oncol. 2004;22:2452–60. metastases with 153Sm-EDTMP. Eur J Nucl Med Mol Imaging.
141. Matthay KK, Panina C, Huberty J, et al. Correlation of tumour 2014;41:238–52.
and whole-body dosimetry with tumour response and toxicity in 160. Parker C, Nilsson S, Heinrich D, Helle SI, O’Sullivan JM, Fosså
refractory neuroblastoma treated with 131I-MIBG. J Nucl Med. SD, et al. Alpha emitter radium-223 and survival in metastatic
2001;42:1713–21. prostate cancer. N Engl J Med. 2013;369:213–23.
142. Flower MA, Fielding SL. Radiation dosimetry for 131I-mIBG ther- 161. Sgouros G, Roeske JC, McDevitt MR, Palm S, Allen BJ, Fisher
apy of neuroblastoma. Phys Med Biol. 1996;41:1933–40. DR, et al. MIRD pamphlet no. 22 (abridged): radiobiology and
143. Gaze MN, Chang YC, Flux GD, et al. Feasibility of dosimetry- dosimetry of α-particle emitters for targeted radionuclide therapy.
based high-dose I-131-meta-iodobenzylguanidine with topotecan J Nucl Med. 2010;51:311–28.
as a radiosensitizer in children with metastatic neuroblastoma. 162. Hobbs RF, Song H, Watchman CJ, Bolch WE, Aksnes AK, Ramdahl
Cancer Biother Radiopharm. 2005;20:195–9. T, et al. A bone marrow toxicity model for Ra-223 α-emitter radio-
144. Koral KF, Huberty JP, Frame B, Matthay KK, Maris JM, et al. pharmaceutical therapy. Phys Med Biol. 2012;57:3207–22.
Hepatic absorbed radiation dosimetry during I-131 metaiodoben- 163. Chittenden S, Hindorf C, Parker CC, Lewington VJ, Pratt BE,
zylguanadine (MIBG) therapy for refractory neuroblastoma. Eur J Johnson B, et al. Phase 1, open-label study of the biodistribution,
Nucl Med Mol Imaging. 2008;35:2105–12. pharmacokinetics and dosimetry of Radium-223 dichloride (223Ra
dichloride) in patients with hormone refractory prostate cancer 165. Pacilio M, Ventroni G, Cassano B, Ialongo P, Lorenzon L, Di
and skeletal metastases. J Nucl Med. 2015;56:1304–9. Castro E, et al. A case report of image-based dosimetry of bone
164. Pacilio M, Ventroni G, De Vincentis G, Cassano B, Pellegrini R, metastases with Alpharadin (223Ra-dichloride) therapy: inter-
Di Castro E, et al. Dosimetry of bone metastases in targeted radio- fraction variability of absorbed dose and follow-up. Ann Nucl
nuclide therapy with alpha-emitting 223Ra-dichloride. Eur J Nucl Med. 2016;30:163–8.
Med Mol Imaging. 2016;43:21–33.
Radiation Protection for Patients
12
Lawrence T. Dauer, Sören Mattsson, Eliseo Vañó,
and Yoshiharu Yonekura
Contents
12.1 Radiation Protection Concepts 262
12.1.1 Principles of Radiation Protection for Patients 262
12.1.2 Radiation Protection Culture 262
12.1.3 Responsibilities 263
12.1.4 Radiation Protection Implementation 263
12.2 Medical Exposure 264
12.2.1 Justification of Medical Exposure 264
12.2.2 Optimization of Protection 264
12.2.3 Pregnant and Lactating Women 265
12.2.4 Pediatric Considerations 267
12.2.5 Hybrid Imaging Considerations 267
12.2.6 Biomedical Research 268
12.2.7 Management of the Therapy Patient 268
12.3 Potential Exposure 270
12.3.1 Error Prevention and Safety Assessment 270
12.3.2 Emergency Procedures 271
12.3.3 Incident Response Investigation and Improvement 271
References 271
Learning Objectives standards with regard to radiation protection of the nuclear

• Understand the fundamental principles of radiation pro- medicine patient.
tection and how they are applied to patients receiving • Delineate key responsibilities and elements for ensuring
administrations of radiopharmaceuticals for nuclear med- radiation protection of the nuclear medicine patient.
icine diagnosis or therapy. • Explain how justification of medical exposures applies to
• Recognize the importance of a radiation protection cul- radiation protection of the nuclear medicine patient.
ture and implementation of the international basic safety • Explain factors that need to be considered and adjusted
according to the principle of optimization of protection
L. T. Dauer (*) for the nuclear medicine patient.
Department of Medical Physics, Memorial Sloan Kettering Cancer • Identify aspects of quality assurance programs and topics
Center, New York, NY, USA for standard operating procedures with regard to nuclear
Department of Radiology, Memorial Sloan Kettering Cancer medicine facilities for both diagnostic and therapeutic
Center, New York, NY, USA procedures.
e-mail: dauerl@mskcc.org
• Recognize the importance of considering radiation pro-
S. Mattsson tection precautions associated with conception, preg-
Medical Radiation Physics, Lund University, Malmö, Sweden
nancy, children, breast-feeding, and release of the
E. Vañó patient.
Department of Radiology, Complutense University of Madrid,
• List elements of error prevention (i.e., prevention of unin-
Madrid, Spain
tended and accidental exposures), emergency procedures,
Y. Yonekura
and incident response investigation and improvement.
National Institute of Radiological Sciences, Chiba, Japan

262 L. T. Dauer et al.
12.1 Radiation Protection Concepts the likelihood of incurring exposures, the number of people
exposed, and the magnitude of their individual doses should
12.1.1 Principles of Radiation Protection all be kept as low as reasonably achievable (ALARA), taking
for Patients into account economic and societal factors), and the princi-
ple of limitation of doses (such that doses to any individual
12.1.1.1 Nuclear Medicine Exposures from regulated sources in planned exposure situations other
The United Nations Scientific Committee on the Effects of than medical exposure of patients should not exceed the
Atomic Radiation (UNSCEAR) has identified that medical appropriate limits recommended by the ICRP) [4]. In a
exposures represent the largest human-made source of radia- nuclear medicine facility, staff occupational and public expo-
tion exposure, accounting for more than 95% of radiation sures are subject to all three principles, while medical expo-
exposure [1]. Traditional nuclear medicine procedures (both sures of patients are subject to the first two only. While there
imaging and therapy) represent about 1–5% or more of all are no specific limits on patient medical exposures, the use of
medical imaging depending on the country [1, 2] and con- radiation in nuclear medicine procedures for patients (or in
tribute significantly to the overall collective effective dose biomedical research) needs to be justified (i.e., benefits out-
from diagnostic procedures. With the advent of hybrid imag- weigh risks) and optimized. This requires careful attention
ing technologies, such as positron-emission tomography/ and balance to ensure that for specific procedures and
computed tomography (PET/CT), single photon emission patients, radiation doses are maintained as low as reasonably
computed tomography/computed tomography (SPECT/CT), achievable, while image quality and specificity or therapeu-
and most recently magnetic resonance imaging paired with tic outcomes are maintained as high as clinically intended.
PET (PET/MRI), the numbers of procedures employing
radiopharmaceuticals will likely increase. Currently, the
total number of nuclear medicine examinations is estimated 12.1.2 Radiation Protection Culture
at more than 35 million per year [3]. Along with broader
usage and larger numbers of procedures comes an increasing The International Atomic Energy Agency (IAEA) has
responsibility for ensuring the many beneficial uses of radia- developed radiation safety standards, based on the recom-
tion while preventing or minimizing detrimental radiation mendations of the ICRP and the scientific information pro-
effects. The aim is to achieve the clinical purpose of the med- vided by UNSCEAR and international consensus for the
ical imaging and therapy with constant vigilance on overall purpose of protection. As such, they represent an important
safety. In nuclear medicine facilities, such conditions demand foundational set of minimal standards for nuclear medicine
rigorous radiation protection programs that extend to facilities. Of course, national and local regulatory require-
patients, medical staff, and members of the public, as well as ments may have additional or slightly different require-
caretakers and comforters of the patients undergoing proce- ments. The IAEA has compiled these into Radiation
dures. This chapter presents important considerations for Protection and Safety of Radiation Sources: International
radiation protection of the nuclear medicine patient. Basic Safety Standards [7], commonly referred to as the
Basic Safety Standards (BSS). It is important that the
12.1.1.2 B asic Principles of Radiation implementation of radiation protection in a nuclear medi-
Protection cine facility be generally based on the IAEA BSS (as well
The International Commission on Radiological Protection as local regulatory requirements and oversight), comple-
(ICRP) recommends a consistent system of radiological pro- mentary to the systems for implementing quality medical
tection that covers all possible exposure situations [4], includ- practice in the facility, and considers the most recent evi-
ing planned exposure situations, emergency exposure dence-based recommendations [8].
situations, and existing exposure situations. Nuclear medicine Typically, an appropriate authorization (or license) from
procedures fall under the first situation, a planned exposure, the appropriate radiation protection regulatory body (or
under national or local regulatory control [5, 6]. Medical agency) is required and based on meeting minimum require-
exposures refer primarily to exposure incurred by patients for ments and demonstrated compliance. Beyond administrative
the purpose of medical diagnosis or treatment and are policies, procedures, and personnel, the most important aspect
extended to exposures incurred by individuals helping in the of a quality and safe nuclear medicine facility is the establish-
support and comfort of patients undergoing such procedures ment of a radiation protection culture where a safety-based
(as well as volunteers in biomedical research programs). attitude is engendered in every individual, such that protection
The ICRP system of radiological protection is built upon and accident prevention are regarded as a natural part of daily
three fundamental principles of radiological protection: the duties [3]. This objective requires education, training, and
principle of justification (whereby any decision that alters encouragement of a questioning and learning attitude, as well
the radiation exposure situation should do more good than as the cooperation and support (with resources, personnel, and
harm), the principle of optimization of protection (whereby money) from administration/management.
12 Radiation Protection for Patients 263
12.1.3 Responsibilities s tandards. A radiation protection program (RPP) is a compre-

hensive implementation tool that should cover all relevant
The licensee of a nuclear medicine facility has the prime aspects of protection of the worker, the general public, and the
responsibility for ensuring radiation protection of the nuclear patient. The licensee needs to provide support and resources
medicine patient by applying the relevant regulations. The necessary to comply with the RPP as well as ongoing effective
licensee must ensure that the necessary personnel (as out- implementation. In addition to specific radiation protection
lined below) are appointed and that these individuals have requirements, the RPP should include at a minimum adminis-
the necessary education, training (initial and continuous), trative requirements (authorizations, inspections, accredita-
and competence to perform assigned duties as well as in tion), managerial requirements (commitment, organization
radiation protection. In addition to an overall medical quality and responsibilities, quality assurance, and human factor con-
assurance program, a specific radiation protection program siderations), safety of radioactive materials, design of facilities
must be established, documented, and understood. Several (and associated shielding), operating procedures, security and
specific responsibilities are briefly discussed below (more control of sources, and emergency plans. Specific guidance for
details are likely available from local regulatory agencies RPP and standards in nuclear medicine is available [9].
and professional societies with regard to specific training and
qualifications for such individuals). 12.1.4.2 Radiation Protection Committee
Nuclear medicine specialist (physician or other qualified Compliance with the RPP (and ultimately the BSS) should
individuals) has the responsibility for the overall radiation be supervised by a facility radiation protection committee,
protection of their patient, including justifying a given nuclear made up of members including an administrator/manage-
medicine procedure, in collaborative discussions with the ment representative, the chief and other representative
referring medical practitioner, and ensuring optimization of nuclear medicine physicians (imaging, therapy, and biomed-
protection in the performance of the procedure. A nuclear ical research), a medical physicist (with nuclear medicine
medicine technologist, typically specifically trained on both training), the RSO, a nuclear medicine technologist, a nurse
radiopharmaceuticals and associated equipment, is uniquely (especially for patients undergoing therapy with radiophar-
positioned to implement the optimization of the patient’s maceuticals), a radiopharmacist, and others as appropriate
exposure. A medical physicist in nuclear medicine should (research radiochemist, maintenance engineer, etc.). The
have a comprehensive knowledge of the imaging equipment radiation protection committee should meet regularly, espe-
used including performance specifications, physical limita- cially to review program changes and staffing aspects and
tions of the equipment, calibration, quality control, and image audit results, new nuclear medicine applications, and bio-
quality [3]. Other personnel that have responsibilities in radi- medical research protocols. The committee should conduct
ation protection in nuclear medicine include radiopharma- overall RPP audits and evaluations in order to ensure ongo-
cists, nurses, and appropriate radiation protection specialists. ing compliance and to identify opportunities for continuous
In most nuclear medicine facilities, licensees are required improvement. These results, along with any committee rec-
to appoint a person to oversee and implement radiation pro- ommendations, should be communicated to the medical
tection activities, the radiation safety officer (RSO) (or other facility administration.
equivalent positions, e.g., radiation protection officer, radia-
tion protection expert, authorized medical physicist). The
RSO should be specially trained with both theoretical and Key Learning Points
practical knowledge of the properties and hazards of ionizing • Medical imaging accounts for more than 95% of
radiation, the means of protection, as well as national and human-made radiation exposure. Nuclear medicine
local regulatory requirements. procedures represent about 1–5% or more of all
All personnel on whom protection and safety depend medical imaging depending on the country but sig-
must be appropriately trained and qualified so that they nificantly contribute to the overall collective effec-
understand their responsibilities and perform their duties tive dose from diagnostic procedures. The advent of
with appropriate judgment and according to defined site pro- hybrid imaging, broader usage, and novel applica-
cedures. Specific levels of training are given by IAEA [9]. tions is expected to increase the responsibility for
ensuring the many beneficial uses of radiation while
preventing or minimizing detrimental radiation
12.1.4 Radiation Protection Implementation effects.
• The ICRP has provided guidelines for a system of
12.1.4.1 Program radiological protection that extend to medical expo-
The BSS requires a nuclear medicine facility to develop, sures that refer primarily to exposure incurred by
implement, and document a protection and safety program to patients for the purpose of medical diagnosis or
ensure compliance with applicable radiation protection
The nuclear medicine specialist (e.g., physician) has the

treatment (or consented biomedical research) as ultimate responsibility for the control of all aspects of the
well as to supporters, comforters, and caretakers. conduct and extent of nuclear medicine applications, includ-
• Two fundamental radiation protection principles ing the justification of the given procedure for a given patient.
apply to nuclear medicine patients: justification In justifying nuclear medicine procedures, relevant interna-
(whereby any decision that alters the radiation tional or national guidelines on appropriateness of examina-
exposure situation should do more good than harm) tions should be taken into account by the nuclear medicine
and optimization of protection (whereby the magni- specialist (e.g., the current guidelines of the Society of
tude of individual doses should be kept as low as Nuclear Medicine and Molecular Imaging [SNMMI] and the
reasonably achievable, taking into account eco- European Association of Nuclear Medicine [EANM] should
nomic and societal factors). be specifically reviewed and consulted by the nuclear medi-
• A radiation protection culture and implementation cine specialist).
of the international basic safety standards with
regard to radiation protection of the nuclear medi-
cine patient are essential elements of safety. Key 12.2.2 Optimization of Protection
responsibilities and program elements need to be
implemented for ensuring radiation protection of Nuclear medicine procedures that have been justified must
the nuclear medicine patient. then be performed according to the principle of optimization
of protection. For diagnostic nuclear medicine procedures,
the patient exposure should be the minimum necessary to
achieve the clinical purpose of the procedure, taking into
12.2 Medical Exposure account acceptable image quality established by appropriate
professional bodies. Appropriate clinical patient dosimetry
The detailed requirements given in Section 3 of the BSS [7] (organ and effective doses) should be performed where nec-
are applicable to medical exposure in nuclear medicine facil- essary by a medical physicist and documented, including for
ities. The IAEA also describes strategies to involve other diagnostic nuclear medicine representative typical patient
organizations such as professional societies (nuclear medi- doses for common procedures. Table 12.1 summarizes the
cine physicians, medical physicists, nuclear medicine tech- effective dose to adults following several commonly per-
nologists, radiopharmacists), whose cooperation is essential formed diagnostic nuclear medicine procedures [10, 11].
to ensure compliance with the overall RPP for medical expo- Comprehensive data on such doses (including specific organ
sures [9]. A summary of important issues with regard to doses and doses for other ages) have been compiled for diag-
radiation protection in medical exposures as applicable to nostic nuclear medicine and should be consulted [12–16].
nuclear medicine is given in this section. The concept of a diagnostic reference level (DRL) pro-
vides a tool for the optimization of protection in medical
exposure [4, 6, 17, 18] in that even though no dose limits are
applied to medical exposure, the process of optimization
12.2.1 Justification of Medical Exposure should result in about the same administered activity ranges
for the same type of examination and for the same size and/
In relation to radiation protection of medical exposures, the or weight of the patient. In the case of nuclear medicine, the
BSS [7] states that “Medical exposures shall be justified by DRL is given as an administered activity for a certain type of
weighing the expected diagnostic or therapeutic benefits… examination and for a normal-sized patient. When recog-
that they yield against the radiation detriment that they might nized and utilized, a DRL can assist in optimization of pro-
cause, with account taken of the benefits and the risks of tection to avoid unnecessarily high activities to the patient or
available alternative techniques that do not involve medical too low activities to provide useful diagnostic information.
exposure.” The principle of justification therefore includes a DRLs are normally set at a national level as a result of con-
case-by-case application that any examination (or therapy or sultation between the health authority, relevant professional
research protocol) be based upon a correct assessment of the bodies, and the radiation protection regulatory body.
indications for the examination, the actual clinical situation, For therapeutic nuclear medicine, clinical patient dosime-
the expected diagnostic and therapeutic yields, and the way try should be evaluated for each individual patient and
in which the results are likely to influence the diagnosis and includes absorbed doses to relevant organs or tissues. The use
the medical care of the patient [3]. of radiopharmaceuticals for therapy requires more detailed
Table 12.1 Effective dose to adults following several commonly performed diagnostic nuclear medicine proceduresa
Effective dose per unit activity “Typical activity” employed Effective dose per investigation
Radiopharmaceutical administered (mSv/MBq) (MBq) (mSv)
[11C]acetate 3.5E−03 500–1000 1.8–3.5
[11C]choline 4.7E−03 400 1.9
[15O]water 1.1E−03 400 0.44
[18F]FDG 1.9E−02 250–400 4.8–7.6
67
Ga-citrate 1.0E−01 150 15
99m
Tc-DMSA 8.8E−03 100 0.88
99m
Tc-DTPA 4.9E−03 300 1.5
99m
Tc-furifosmin (rest/exercise) 1.0E−02/8.9E−03 300–900 2.7–9.0
99m
Tc-MAA 1.1E−2 100–200 1.1–2.2
99m
Tc-MAG3 7.0E−03 70–200 0.49–1.4
99m
Tc-phosphates, Tc-phosphonates 4.9E−03 600 2.9
99m
Tc-sestamibi (rest/exercise) 9.0E−03/7.9E−03 300–900 2.4–8.1
123
I-iodide (35% thyroid uptake) 3.0E−01 2–20 0.6–6
123
I-iodide (1% thyroid uptake) 3.7E−02 200–400 7.4–14.8
123
I-MIBG 1.3E−02 400 5.2
131
I-iodide (35% thyroid uptake) 2.9E+01 0.2 5.8
131
I-iodide (1% thyroid uptake) 2.8E−01 4 1.1
201
Tl-chloride 1.4E−01 80 11
See [12–16] for specific organ doses and doses for other ages for these and many other radiopharmaceuticals
a
as the overall medical facility quality management system

Table 12.2 Aspects of a QA program for medical exposures (adapted with direct oversight and support by hospital administration
from [3]) [19–21]. Table 12.2 lists several specific aspects that should
Measurements by, or under the oversight of, a medical physicist of be included in such a QA program for medical exposures.
the physical parameters of medical radiological equipment at the Regular and independent audits of the QA program and the
time of acceptance and commissioning prior to clinical use on
patients, periodically thereafter, and after any major maintenance RPP shall be implemented.
that could affect patient protection
Implementation of corrective actions if measured values of the 12.2.2.2 D ocumented Standard Operating
physical parameters are outside established tolerance limits Procedures
Verification of the appropriate physical and clinical factors used in Written local rules and standard operating procedures for all
patient diagnosis or treatment
applications of nuclear medicine should be developed,
Records of relevant procedures and results
Periodic checks of the appropriate calibration and conditions of implemented, reviewed, and improved over time. These local
operation of dosimetry and monitoring of equipment rules should be acknowledged, known, and used by all
involved in medical exposures using nuclear medicine.
Table 12.3 lists minimum topics for such local rules and
and patient-specific dosimetry planning, including both tumor standard operating procedures.
and normal tissues. The appropriate radiopharmaceutical and
activity are selected and administered in such a way to pri-
marily localize the activity in the organ(s) of interest, while 12.2.3 Pregnant and Lactating Women
the activity in the rest of the body is kept ALARA.
The ICRP notes that special consideration should be given to
12.2.2.1 Quality Assurance Program pregnant women exposed to ionizing radiation due to the larger
Licensees of a nuclear medicine facility must have a compre- probability of inducing radiation effects in individuals exposed
hensive quality assurance (QA) program for medical expo- in utero compared to exposed adults [4, 22–24]. Therapeutic
sures, with active participation of the nuclear medicine administrations of radiopharmaceuticals are routinely contrain-
specialists, medical physicists, nuclear medicine technolo- dicated in the case of pregnancy or breast-feeding as they may
gists, and radiopharmacists. Such a QA program should be result in very high fetal doses. In addition, beyond 10–13 weeks
complementary to and part of the wider RPP (that also of gestation, the fetal thyroid may receive extremely high doses
includes occupational, public, and patient exposures) as well in cases of therapy using 131I-iodide or 131I-labeled compounds
Table 12.3 Topics covered by local rules and standard operating pro- for an appropriate period (see IAEA Safety Reports Series
cedures (adapted from [3])
No. 40 for additional information [9]). Therapy with a radio-
Routines for patient identification and information active substance is typically contraindicated and should only
Prescribed radiopharmaceutical and activity for adults and children be considered when the application may be a lifesaving pro-
for different types of examination, including methods used to adjust
the activity to the single patient and routes of administration cedure. Importantly, therapy of hyperthyroidism with
Management of patients that are pregnant or might be pregnant
131
I-iodide in a pregnant woman is strictly contraindicated.
Management of breast-feeding patients This is partly due to the possibility of external irradiation of
Routines for safe preparation and administration of the fetus, but mostly due to radioactive iodide crossing the
radiopharmaceuticals including activity measurements placenta into the fetal circulation with subsequent uptake by
Procedures in case of misadministration of the radiopharmaceutical its thyroid. The gland may well be destroyed by beta radiation
or other errors
from the 131I taken up. Therefore, other methods of treatment
Detailed procedure manuals for every type of examination including
handling of equipment should be employed, if possible, until delivery. When metas-
Detailed protocols (and associated procedures) for nuclear medicine tases from thyroid cancer are diagnosed in a pregnant woman,
therapy, including dosimetry, treatment planning, administration, treatment with 131I-iodide, if it cannot be delayed until after
and patient follow-up delivery, might not be compatible with continuation of the
pregnancy [30]; nevertheless, the choice whether or not to
delay therapy until after delivery must be based on estimation
(e.g., 131I-metaiodobenzylguanidine). As a general rule, it is of the absorbed dose to the uterus (see also below).
also recommended that diagnostic procedures of women likely Exposure of a pregnant patient (either at a time when the
to be pregnant be avoided unless there are strong clinical indi- pregnancy was not known or when evaluated as required)
cations. Comprehensive compilations of doses to the embryo often leads to her apprehension because of concern about the
and fetus from substances in ionic form and some radiophar- possible effects on the fetus. The expected fetal dose should
maceuticals have been published [25–28] and should be con- be carefully estimated by a qualified medical physicist (espe-
sulted in conjunction with a qualified medical physicist when cially if the fetal dose is suspected to be higher than 10 mGy),
necessary. and the medical specialist should inform the patient about the
A woman of childbearing age should be evaluated regard- possible risks. Such exposures may lead to a discussion
ing possible pregnancy or a missed period. This should be regarding termination of pregnancy due to the radiation risks.
performed prior to the procedure or treatment. It is recom- Many misunderstandings and lack of knowledge, also among
mended [29] to place a poster in the waiting area requesting physicians, have probably resulted in unnecessary termina-
a woman to notify the staff if she is or thinks she is pregnant. tion of pregnancies [3]. It is generally considered that for a
Depending on the type of procedure, other facilities may also fetal dose of less than 100 mGy, as in most diagnostic proce-
require the patient to fill out a specific questionnaire on the dures, termination of pregnancy is not justified from the point
subject (especially for higher-dose procedures and therapeu- of view of radiation risks [3, 31]. At higher doses, individual
tic administrations where additional tests may be performed circumstances should be taken into account with careful eval-
for confirmation). If pregnancy is confirmed, careful consid- uation by the whole nuclear medicine team and the patient.
eration should be given to other methods for diagnosis or Nuclear medicine examinations on women who are
postponement of the examination until after delivery. In the breast-feeding present a potential radiation hazard to the
case where it is deemed that a nuclear medicine diagnostic baby. This is due to uptake of some radiopharmaceuticals in
procedure is required, then the examination should be per- breast glandular tissue followed by excretion into the breast
formed only after careful optimization of protection that also milk. The dose to the baby depends on various factors (radio-
considers protection of the embryo/fetus (perhaps reducing pharmaceutical, amount of milk, time of the feeding of the
the administered activity with longer image acquisition child after administration, and geometric factors as the
times, frequent voiding). mother represents a source of external exposure). It is recom-
Hybrid imaging diagnostic procedures such as PET/CT or mended [32] to place a poster in the waiting area requesting
SPECT/CT examinations are of special concern, as routine a woman to notify the staff if she is breast-feeding. Some
diagnostic CT examinations of the pelvic region with and restrictions on breast-feeding and advice to the mother are
without contrast can lead to a dose of 50 mSv to the uterus necessary to minimize the exposure of the baby to an accept-
(about equivalent to the fetal dose in early pregnancy), able level [33]. It is the responsibility of the nuclear medi-
depending on scan techniques, the scanned region, and the cine specialist in cooperation with the medical physicist to
patient. In such a case, it is important to use lower-dose CT establish local rules regarding breast-feeding and close con-
protocols and to reduce the scanning area to a minimum. tact between the mother and the child after nuclear medicine
Following treatment with a therapeutic activity of a radio- administrations. Guidance on such rules has been promul-
nuclide, female patients should be advised to avoid pregnancy gated by the IAEA [9] and should be specifically consulted.
It should be emphasized that after 131I therapy (and also after that have shown that for a given radiation dose, children are
diagnostic investigations using 123/124/125/131I), mothers must generally at more risk of tumor induction than are adults and
cease breast-feeding immediately [15, 34]. that radiogenic tumor incidence in children is more variable
Female patients should be advised that breast-feeding is than in adults (and depends on the tumor type, age, and gen-
contraindicated after therapeutic administration of radionu- der). UNSCEAR noted that for leukemia and thyroid, skin,
clides, and women as well as men should be advised con- breast, and brain cancer induction, children were clearly
cerning the avoidance of conception after therapeutic more radiosensitive than adults. Health effects and risks are
administrations. Pregnancy is a strong contraindication to also dependent on a number of physical factors. Regarding
unsealed radionuclide therapy, unless the therapy is lifesav- internal exposure, because of the smaller size of infants and
ing. This advice is all the more valid for radioiodine therapy children and thus because their organs are closer together,
and for other radionuclides with the potential to impart radia- radionuclides concentrated in one organ irradiate other
tion doses to the fetus in the range of a few mSv. Therefore, organs of children’s bodies more than occurs in adults.
where treatment is likely or anticipated, the patient should be Because children have smaller body diameters and there is
advised to take appropriate contraceptive measures in the less shielding by the overlying and adjacent tissues, the dose
time prior to therapy. Some radiopharmaceuticals, including to their internal organs will be larger than for an adult for a
131
I as iodide and 32P as phosphate, rapidly cross the placenta; given external exposure. There are also many other age-
therefore, the possibility of pregnancy should be carefully related factors involving metabolism and physiology that
excluded before administration. The ICRP has given detailed make a substantial difference in dose at different ages. In
guidance [23, 24]. diagnostic medical exposure, children may receive signifi-
Following treatment with a therapeutic activity of a radio- cantly higher doses than adults for the same examination if
nuclide, female patients should also be advised to avoid the technical parameters for delivering or administering the
pregnancy for an appropriate period. The ICRP suggests that dose are not specifically adapted.
women should not become pregnant for some time after As in all nuclear medicine exposures, it is important to
radionuclide therapy (e.g., 6 months for radioiodine, the implement optimization of protection for an examination of a
most common radionuclide used) [6]. Various shorter or lon- child. Typically this is implemented through the optimization
ger times for this and other radionuclides are given by ICRP of the administered activity to the minimum consistent with
and the IAEA [9, 33] which identifies periods of 3 months, obtaining a diagnostic result. For children or young persons,
4 months, and 24 months for 32P, 131I, and 89Sr treatments, body weight should always be measured (and perhaps the
respectively. Some practitioners use a 6–12-month gap for height or body surface area determined as well), and adult
131
I, with a view to providing further confidence in this administered activity be scaled down. Local procedures
regard. The IAEA [35] gives additional information on pre- should be developed for each nuclear medicine examination,
caution times for female avoidance of conception for specific trained, and implemented along the continuum of care. In
radionuclide therapies. hybrid imaging, the specific CT protocol should also be opti-
The administration of therapeutic activities of relatively mized (perhaps by reducing the current-time product [mAs]
long-lived radionuclides in ionic chemical forms to males is and tube potential [kV] without compromising the diagnostic
also a possible source of concern because of the appearance quality of the images). It is important to use individual proto-
of larger quantities of these radionuclides in ejaculate and in cols based on the size and/or weight of the child. Specific
sperm. It is widely recommended in practice, on the basis of imaging protocols should be developed in conjunction with
prudence, that male patients take steps to avoid fathering qualified medical physicists and the nuclear medicine special-
children during the months immediately following therapy. ist. Several professional organizations and researchers have
However, there is no strong evidence base to support this offered guidance for the optimization of protection for pediat-
view [35]. Some have suggested that it may be prudent to ric applications of nuclear medicine [37–40].
advise sexually active men who have been treated with 131I
(iodide), 32P (phosphate), or 89Sr (strontium chloride) to
avoid fathering children for a period of 4 months after treat- 12.2.5 Hybrid Imaging Considerations
ment, a period suggested as it is longer than the life of a
sperm cell. Hybrid imaging diagnostic procedures, such as PET/CT or
SPECT/CT scans, have become indispensable modalities for
the diagnosis, staging, and monitoring of therapy response of
12.2.4 Pediatric Considerations a broad range of malignancies and other diseases. Regardless
of acquisition circumstances, the patient receives dose from
UNSCEAR has specifically evaluated the risks and effects of both the PET or SPECT administrations of radiopharmaceu-
radiation exposure of children and pediatric populations [36] ticals for nuclear medicine diagnosis and from the CT (that
can be for attenuation correction, low-dose co-registration, or of the deterministic or stochastic biological effects. There is
even full diagnostic imaging). Justification and optimization a need for a multidisciplinary approach for such procedures.
apply for hybrid imaging as well. Increased awareness of the In radiation therapy, the dose to the patient is intentional,
risk of exposure to ionizing radiation has resulted in efforts to and its potentially cell-killing properties are the very purpose
minimize radiation dose incurred during X-ray and nuclear of the treatment. In such cases, optimization of protection
medicine imaging tests [41]. Implementation of the optimiza- becomes an effort in minimizing doses (and/or their deleteri-
tion of protection strategies depends critically on accurate ous effects) to surrounding tissues without compromising
dose measurement/dosimetry to maximize the benefit/risk the predetermined and intentionally lethal dose and effect on
ratio from imaging tests. Careful evaluation (by the qualified the target region. This is an important balance for if the dose
medical physicist and nuclear medicine specialist and radiol- to the target tissue is too low, the therapy will be ineffective,
ogists) of CT protocols used in hybrid imaging is important and the exposure would not have been justified. At the same
because when protocols other than low-dose attenuation cor- time, excess radiation delivered to normal (or healthy) tis-
rection only are utilized, they can result in radiation doses at sues is associated with various undesired or deleterious
least as high as received from the nuclear medicine or even effects and perhaps some risk of second malignancies.
higher (as in the case when diagnostic full-body CT scans are Dosimetry should be performed for each treatment (particu-
coupled with the nuclear medicine diagnostic imaging) [42]. larly for children and young people).
12.2.7.2 C are and Support of the Patient

12.2.6 Biomedical Research (Comforters and Caretakers)
A patient who has undergone a therapeutic nuclear medicine
An exposure as part of biomedical research is treated on the procedure is a source of radiation that can lead to the expo-
same basis as a medical exposure and, therefore, is not subject sure of other persons that come into the proximity of the
to dose limits. However, in all investigations involving expo- patient. External irradiation of the persons close to the patient
sure of humans, a careful estimation of the radiation dose to depends upon the radionuclide used and its emissions, half-
the volunteer should be made. Associated risk should then be life, and biokinetics. Excretion results in the possibility of
weighed against the benefit for future patient groups and the contamination of the patient’s environment and of inadver-
society. With regard to justification of medical exposure, any tent ingestion by other persons. Individuals that work with
nuclear medicine procedure that occurs as part of a biomedical therapeutic nuclear medicine patients are subject to the sys-
research project (or protocol) is considered justified only if the tem of radiation protection for occupationally exposed per-
project has been approved by both the radiation protection sons. If the person, other than occupationally, is knowingly
committee and an ethics committee. Specifically, the radiation and voluntarily providing care, then their exposure is consid-
exposure of humans for biomedical research is deemed not to ered part of medical exposure.
be justified unless it is in accordance with the provisions of the Children, the elderly or the infirm, may have difficulty
Helsinki Declaration [43] and follows the guidelines for its during a nuclear medicine procedure, and adults associated
application prepared by the Council for International with the patient may knowingly and voluntarily choose to
Organizations of Medical Sciences [44] and any other local care, support, or comfort such patients. In such circum-
regulatory requirements. Both the ICRP [4] and the IAEA [7] stances, the BSS notes that the dose to these persons should
give recommendations with regard to the use of dose con- be constrained so that is unlikely that it will exceed 5 mSv
straints, on a case-by-case basis, in the process of optimiza- during the period of a patient’s procedure. The dose to chil-
tion. The ICRP has also provided additional guidance on dren (as well as all in the general public) encountering
radiological protection in biomedical research [45]. patients who have been administered radioactive materials
should be similarly constrained to less than 1 mSv. Special
concern should be given to members of the family of a
12.2.7 Management of the Therapy Patient patient who has received radionuclide therapy. Procedures
for advising caretakers and comforters should be in place in
12.2.7.1 Optimization Considerations consultation with the RSO. Registrants and licensees should
for Nuclear Medicine Therapy ensure that caretakers and comforters of patients during the
Radionuclide therapy is a complex procedure, encompassing course of treatment with radionuclides (e.g., 131I for hyper-
a wide range of radionuclides, different targeting mecha- thyroidism and thyroid carcinoma, 89Sr and 186Re for pain
nisms, and various methods of administration. Each radio- palliation) receive sufficient written instructions on relevant
therapeutic procedure presents a unique set of challenges for radiation protection precautions (e.g., time and proximity to
dosimetry calculation, related either to quantitative imaging, the patient). Example methodologies for evaluating precau-
the absorbed dose calculations themselves, or considerations tion time requirements have been published [46, 47].
12.2.7.3 Release of the Patient achievable and information on the risks of ionizing

While precautions for the public are rarely required after radiation.
diagnostic nuclear medicine procedures, some therapeutic Patients traveling after radioiodine therapy rarely present a
nuclear medicine procedures, particularly those involving hazard to other passengers if travel times are limited to a few
131
I, can result in significant exposure to other people. hours. Travel for 1–2 h immediately posttreatment in a private
However, patients do not need to be hospitalized automati- automobile large enough for the patient to maintain a distance
cally after all radionuclide therapies. Relevant national dose of 1 m or greater from the other vehicle occupant(s) is gener-
limits must be met, and the principle of optimization of pro- ally permissible. A case-by-case analysis is necessary to
tection must be applied, including the use of relevant dose determine the actual travel restrictions for each patient, espe-
constraints. The dose to members of the general public, as cially for longer trips and for travel by public transport. The
well as children, encountering patients who have been ICRP has evaluated the use of holding tanks for the storage of
administered radioactive materials, should be constrained to urine for patients following nuclear medicine therapy and
less than 1 mSv [3, 4]. The decision to hospitalize or release noted that their use is generally unnecessary [49].
a patient should be determined by the nuclear medicine spe- International security measures, such as those in place at
cialist on an individual basis. In addition to residual activity airports and border crossing points, can include extremely
in the patient, the decision should take many other factors sensitive radiation detectors. It is quite possible that patients
into account including the patient’s wishes, family consider- treated with gamma-emitting radionuclides could trigger
ations (particularly the presence of children), environmental these alarms, particularly in the period immediately follow-
factors, and existing guidance and regulations. Hospitalization ing discharge. Environmental or other radiation detection
will reduce exposure to the public, but will increase exposure devices are able to detect patients who have had radioiodine
to hospital staff. Hospitalization often involved a significant therapy and some diagnostic procedures for several weeks
psychological burden as well as monetary and other costs after treatment [50, 51]. Triggering of an alarm does not
that should be analyzed and justified. mean that a patient is emitting dangerous levels of radia-
It should be noted that the dose to adults from patients is tion—the detectors are designed to detect levels of radioac-
mainly due to external exposure. Internal contamination of tivity far below those of concern to human health. The
family members is most likely in the first 7 days after security authorities are well aware of this possibility, and if a
treatment. In most circumstances, the risks from internal patient is likely to travel soon after discharge, the hospital or
contamination of others are less significant than those from the patient’s doctor should provide a written statement of the
external exposure [35]. In general, contamination of adults is therapy and radionuclide used, for the patient to carry.
less important than external exposure. However, contamina-
tion of infants and children with saliva from a patient could 12.2.7.4 Changes in Medical Status
result in significant doses to the child’s thyroid (in the case of If the medical condition of a patient deteriorates such that
131
I therapies) [6]. Therefore, it is important to avoid con- intensive nursing care becomes necessary, urgent medical
tamination of infants, young children, and pregnant women. care is a priority and should not be delayed. In the event the
For both external and internal exposure avoidance, written patient’s medical condition deteriorates, frequent or contin-
instruction to the patient concerning contact with other per- ual monitoring of the patient may be necessary (e.g., septic
sons and relevant precautions for radiation protection must shock, pulmonary edema, stroke, or myocardial infarction).
be provided and should be based on calculated estimates of Lifesaving efforts shall take precedence over considerations
exposure to others. Examples of such calculations are found of radiation exposures received by medical personnel.
in the literature [47, 48] and should be performed with the Medical personnel should, therefore, proceed with emer-
help of a qualified medical physicist. gency care while taking precautions against the spread of
When deciding on the appropriate discharge activity and contamination and minimizing external exposure. The advice
instructions for a particular patient, the licensee should take of the RSO should be sought immediately.
into account the transport and the living conditions of the The care of patients receiving radiopharmaceutical ther-
patient, such as the extent to which the patient can be iso- apy and who are on dialysis may require additional consider-
lated from other family members and the requirement to ation. In general, for systemic treatments, these patients will
dispose safety of the patient’s contaminated excreta. In not biologically clear radioactive materials as quickly as
some cases, such as for the elderly or child patient, it may be typical patients since the clearance is highly dependent on
necessary to discuss the precautions to be taken with other the schedule of the dialysis session. It may be necessary to
family members. When required, the patient or legal guard- reduce or otherwise adjust the activity required for a therapy.
ian shall be provided with written (and also a verbal expla- The decision as to the activity required for such patients
nation of) instructions with a view to the restriction of doses should be based on either a trace trial administration of activ-
to persons in contact with the patient as far as reasonably ity and the observed elimination rate or a careful review of
the available literature for similar patient administrations. 12.3 Potential Exposure
Typically, the largest amount of radioactivity will be elimi-
nated during the first dialysis session following radiophar- 12.3.1 Error Prevention and Safety
maceutical therapy. Assessment
Therapeutic amounts of radioactive materials are typically
not administered to critically ill patients unless there are cir- Unintended and accidental exposures that occur due to
cumstances where the palliative use of radioactive materials human error, equipment failure, or other combinations of
in terminal patients will significantly improve the quality of errors are called potential exposures. It is the responsibility
life of the patient. However, should the patient die in the of the licensee and all involved in nuclear medicine applica-
period immediately following therapy, special consideration tions to prevent such events as far as possible and, in the
may need to be given to the treatment of the corpse. In radio- event that they do occur, to mitigate their consequences and
nuclide therapy, emergency plans have to be available on how implement after action careful safety assessments. A well-
to handle the cadaver and should be developed in conjunction established RPP is fundamental in accident prevention
with the qualified medical physicist. Since it is a sensitive together with a high level of radiation protection and safety
issue, depending on ethical and religious rules and traditions, culture in the organization and among the people working in
advice should be available from the national authorities. a nuclear medicine facility. Nuclear medicine facilities must
conduct a prospective safety assessment applied to all stages
of the design and operation of the nuclear medicine facility,
Key Learning Points including a systematic critical review and identification of
• The nuclear medicine specialist has the ultimate possible events leading to unintended or accidental exposure
responsibility for the control of all aspects of the and to develop effective and multiple barriers to their occur-
conduct and extent of nuclear medicine applica- rence (including forcing functions, procedures, training, and
tions, including justifying nuclear medicine proce- culture, among a comprehensive safety program). Working
dures. Relevant international or national guidelines procedures should require key decisions, especially in radia-
on appropriateness of examinations should be taken tion therapy, to be subject to independent confirmation.
into account by the nuclear medicine specialist. Effective communication between all the staff involved is a
• For diagnostic nuclear medicine procedures, the vital part of the process.
patient exposure should be the minimum necessary Examples of what can potentially go wrong in a nuclear
to achieve the clinical purpose of the procedure, medicine facility with respect to patient safety may include
taking into account acceptable image quality estab- (but not be limited to) wrong patient scheduled, wrong
lished by appropriate professional bodies. A diag- patient identified, missed pregnancy or breast-feeding infor-
nostic reference level (DRL) can assist in mation, misadministration (wrong patient, wrong activity,
optimization of protection to avoid unnecessarily wrong radiopharmaceutical), unexpected contamination,
high activities to the patient or too low activities to inconclusive equipment results, or the need for repeated
provide useful diagnostic information. scans.
• For therapeutic nuclear medicine, clinical patient Prior to any administration, the following should be
dosimetry should be evaluated for each individual verified:
patient and includes absorbed doses to relevant
organs or tissues. –– Dose on the radiopharmaceutical label matches the
• A comprehensive quality assurance program for prescription.
medical exposures is an essential component of a –– Identification of the patient by two independent means.
healthy radiation protection program and should be –– Identity of the radionuclide.
developed and implemented with active participa- –– Identity of the radiopharmaceutical.
tion of all personnel associated with procedures. –– Total activity to be administered.
• Special radiation protection precautions are essen- –– Total activity in the dose being administered.
tial when dealing with conception, pregnancy, chil- –– Date and time of administration.
dren, breast-feeding, and release of the patient. –– Patients have been given information about their own
safety.
12.3.2 Emergency Procedures References
Nuclear medicine facilities need to prepare emergency pro- 1. UNSCEAR. Effects of ionizing radiation. UNSCEAR 2006
report to the general assembly with scientific annexes. Volume 1.
cedures on the basis of events identified in the safety assess- New York, NY: United Nations Scientific Committee on the Effects
ment (see above) and hazard vulnerability assessments. Such of Atomic Radiation; 2008.
procedures should be clear, concise, specific, and available in 2. NCRP. Ionizing radiation exposure of the population of the United
places where their need is anticipated. They should include, States. NCRP report no. 160. Bethesda, MD: National Council on
Radiation Protection and Measurements; 2009.
as a minimum, descriptions of possible incidents and acci- 3. Baily D, Humm JL, Todd-Pokropek A, van Answegan A, editors.
dents (and measures for handling them), responsible actions Nuclear medicine physics: a handbook for teachers and students.
of responding individuals, equipment and tools available, Vienna: International Atomic Energy Agency; 2014.
training and periodic exercises/drills, immediate measures to 4. ICRP. The 2007 recommendations of the International Commission
on Radiological Protection. ICRP publication 103. Ann ICRP.
avoid unnecessary radiation doses to patients (as well as staff 2007;37(2-4):1.
and the public), measures to prevent spread of contamina- 5. ICRP. Radiological protection in medicine. ICRP publication 73.
tion, and any mitigating actions for the patient. Ann ICRP. 1996;26(2):1.
6. ICRP. Radiological protection in medicine. ICRP publication 105.
Ann ICRP. 2008;37(6):1.
7. IAEA. Radiation protection and safety of radiation sources: inter-
12.3.3 Incident Response Investigation national basic safety standards. IAEA safety standards series no.
and Improvement GSR Part 3. Vienna: International Atomic Energy Agency; 2014.
8. IAEA. Radiation protection of patients. Information for health
professionals. Vienna: International Atomic Energy Agency;
In the unfortunate case where an unintended or accidental 2017. https://rpop.iaea.org/RPOP/RPoP/Content/InformationFor/
medical exposure occurs, the licensee is required to deter- HealthProfessionals/3_NuclearMedicine/index.htm. Accessed 11
mine the patient doses involved, identify any corrective Sep 2017.
actions needed to prevent recurrence, and implement the cor- 9. IAEA. Applying radiation safety standards in nuclear medicine,
IAEA safety reports series no. 40. Vienna: International Atomic
rective measures. Licensees should examine both immediate Energy Agency; 2005.
causes and any latent system causes. There may be a require- 10. Mattsson S, Hoeschen C. Radiation protection in nuclear medicine.
ment to report the event to the regulatory body (and the RSO Heidelberg: Springer; 2013.
should be notified and consulted to evaluate). 11. Martin CJ, Sutton DG, editors. Practical radiation protection in
healthcare. 2nd ed. Oxford: Oxford University Press; 2015.
In the case of a misadministration or an accident in radio- 12. ICRP. Radiation dose to patients from radiopharmaceuticals. ICRP
nuclide therapy, the responsible nuclear medicine specialist publication no. 53. Ann ICRP. 1987;18(1-4):1.
should be promptly informed, who should then inform the 13. ICRP. Radiation dose to patients from radiopharmaceuticals.

referring physician and the patient. Medical physicists Addendum 2 to ICRP publication 53. ICRP publication no. 80. Ann
ICRP. 1998;28(3):1.
should make dose calculations, and the staff involved in the 14. ICRP. Radiation dose to patients from radiopharmaceuticals -

accident should be consulted. Reports may need to be sub- addendum 3 to ICRP publication 53. ICRP publication 106. Ann
mitted to the regulatory body. The initiating event and the ICRP. 2008;38(1/2):1.
contributing factors should always be identified and should 15. ICRP. Radiation dose to patients from radiopharmaceuticals: a

compendium of current information related to frequently used sub-
be used in developing improvement actions, future training, stances. ICRP publication no. 128. Ann ICRP. 2015;44(2S):1.
and communications. 16. Stabin M, Siegel JA. RADAR dose estimate report: a compen-
dium of radiopharmaceutical dose estimates based on Olinda/Exm
Version 2.0. J Nucl Med. 2018;59(1):154.
17. ICRP. Diagnostic reference levels in medical imaging: review and
additional advice. Ann ICRP. 2017;31(4):33. http://www.icrp.org/
Key Learning Point docs/DRL_for_web.pdf. Accessed 11 Sep 2017.
• It is important to develop, train on, and implement 18. Vañó E, et al. Diagnostic reference levels in medical imaging. ICRP
standard operating procedures, as well as error pre- publication no. 135. Ann ICRP. 2017;46:1–144.
vention techniques, emergency procedures, and 19. WHO. Quality assurance in nuclear medicine. Geneva: World

Health Organization; 1982.
incident response investigation and improvement 20. IAEA. Nuclear medicine resources manual. Vienna: International
activities. Atomic Energy Agency; 2006.
21. IAEA. Quality assurance for radioactivity measurement in nuclear 37. Fahey FH, Treves ST, Adelstein SJ. Minimizing and communi-
medicine. IAEA Technical Report Series No. 454. Vienna: cating radiation risk in pediatric nuclear medicine. J Nucl Med.
International Atomic Energy Agency; 2006. 2011;52(8):1240–51.
22. ICRP. Age-dependent doses from intakes of radionuclides. ICRP 38. EANM. Paediatric dosage card. Vienna: European Association

publication no. 72. Ann ICRP. 1996;26(1):1. of Nuclear Medicine; 2016. http://www.eanm.org/content-eanm/
23. ICRP. Pregnancy and medical radiation. ICRP publication 84. Ann uploads/2017/01/EANM_Dosage_Card_040214.pdf. Accessed 11
ICRP. 2000;30(1):1. Sep 2017.
24. ICRP. Biological effects after prenatal irradiation (embryo and
39. Fahey FH, Goodkind A, MacDougall RD, Oberg L, Ziniel SI,
fetus). ICRP publication 90. Ann ICRP. 2003;33(1-2):1. Cappock R, et al. Operational and dosimetric aspects of pediatric
25. ICRP. Doses to the embryo and fetus from intakes of radionuclides PET/CT. J Nucl Med. 2017;58(9):1360–6.
by the mother. ICRP publication no. 88. Ann ICRP. 2001;31(1-3):1. 40. Treves ST, Gelfand MJ, Fahey FH, Parisi MT. 2016 Update of the
26. ANSI/HPS. Fetal radiation dose calculations. ANSI/HPS N13.54. North American Consensus Guidelines for pediatric administered
McLean, VA: Health Physics Society; 2008. radiopharmaceutical activities. J Nucl Med. 2016;57(12):15N–8N.
27. Stabin MG. Radiation dose and risks to fetus from nuclear medi- 41. Brink JA, Amis ES Jr. Image Wisely: a campaign to increase

cine procedures. Phys Med. 2017;43:190–8. awareness about adult radiation protection. Radiology.
28. Zanotti-Fregonara P, Stabin MG. New fetal radiation doses for 18F- 2010;257(3):601–2.
FDG based on human data. J Nucl Med. 2017;58(11):1865–6. 42. Quinn B, Dauer Z, Pandit-Taskar N, Schoder H, Dauer LT. Radiation
29. Valentin J. Pregnancy and medical radiation. ICRP publication no. dosimetry of 18F-FDG PET/CT: incorporating exam-specific
84. Ann ICRP. 2000;30(1):1. parameters in dose estimates. BMC Med Imaging. 2016;16(1):41.
30. ICRP. Radiation and your patient - a guide for medical practitio- https://doi.org/10.1186/s12880-016-0143-y.
ners. Ann ICRP. 2001;31(4):1. 43. WMA. 18th World Medical Assembly, Helsinki, 1974, as amended
31. Dauer LT, Thornton RH, Miller DL, Damilakis J, Dixon RG, Marx by the 59th World Medical Assembly. Seol: World Medical
MV, et al. Radiation management for interventions using fluoro- Association; 2008.
scopic or computed tomographic guidance during pregnancy: a 44. WHO. International ethical guidelines for biomedical research

joint guideline of the Society of Interventional Radiology and the involving human subjects. Council for International Organizations
Cardiovascular and Interventional Radiological Society of Europe of Medical Sciences. Geneva: World Health Organization; 2002.
with Endorsement by the Canadian Interventional Radiology 45. ICRP. Radiological protection in biomedical research. ICRP publi-
Association. J Vasc Interv Radiol. 2012;23(1):19–32. cation no. 62. Ann ICRP. 1992;22(3):1.
32. ICRP. Protection of the patient in nuclear medicine (and statement 46. Zanzonico PB, Siegel JA, St Germain J. A generalized algorithm
from the 1987 Como meeting of ICRP). ICRP publication no. 52. for determining the time of release and the duration of post-release
Ann ICRP. 1987;17(4):1. radiation precautions following radionuclide therapy. Health Phys.
33. Valentin J. Doses to infants from ingestion of radionuclides in moth- 2000;78(6):648–59.
er’s milk. ICRP publication no. 95. Ann ICRP. 2004;34(3–4):1–27. 47. NCRP. Management of radionuclide therapy patients. NCRP report
34. Leide-Svegborn S, Ahlgren L, Johansson L, Mattsson S. Excretion no. 155. Bethesda, MD: National Council on Radiation Protection
of radionuclides in human breast milk after nuclear medicine and Measurements; 2006.
examinations. Biokinetic and dosimetric data and recommenda- 48.
USNRC. Radioactive effluents from nuclear power plants
tions on breastfeeding interruption. Eur J Nucl Med Mol Imaging. annual report 2008. Washington, DC: U.S. Nuclear Regulatory
2016;43(5):808–21. Commission; 2008.
35. IAEA. Release of patients after radionuclide therapy with contri- 49. ICRP. Release of patients after therapy with unsealed radionuclides.
butions from the ICRP, IAEA safety report series no. 63. Vienna: ICRP publication 94. Ann ICRP. 2004;34(2):1.
International Atomic Energy Agency; 2009. 50. Dauer LT, Williamson MW, St Germain J, Strauss HW. Tl-201 stress
36. UNSCEAR. Sources, effects and risks of ionizing radiation. United tests and homeland security. J Nucl Cardiol. 2007;14(4):582–8.
Nations Scientific Committee on the Effects of Atomic Radiation. 51. Dauer LT, Williamson MW, St Germain J, Strauss HW. Response to
UNSCEAR 2013 Report to the General Assembly with Scientific nuclear granny. J Nucl Cardiol. 2007;14:904–5.
Annexes, Volume II, Scientific Annex B, Effects of radiation expo-
sure of children. New York, NY: United Nations, 2013.
Radiation Protection for Personnel
and the Environment 13
Antonio C. Traino
Contents
13.1 General Concepts in Radiation Protection 273
13.2 adiation Protection Units
R 275
13.2.1 Average Absorbed Dose 275
13.2.2 Equivalent Dose 275
13.2.3 Effective Dose 275
13.2.4 Committed Dose 276
13.3 R adiation Protection in Nuclear Medicine 276
13.3.1 Introduction 276
13.3.2 Formal Authorizations and Responsibilities 276
13.4 Facility Design 277
13.4.1 Location and General Layout 277
13.4.2 Storage of Radioactive Sources 277
13.4.3 Structural Shielding 277
13.4.4 Workplace Classification 277
13.4.5 Waste Management Operations 278
13.5 Exposure of Workers 278
13.5.1 Sources of Exposure 278
13.5.2 Justification, Optimization, and Dose Limitation 279
13.5.3 Pregnant Workers 279
13.5.4 Protective Clothing 279
13.5.5 Personal Monitoring 279
13.5.6 Local Rules and Supervision 280
References 280
Learning Objectives • Identify and manage the main sources of radiation

• Understand the fundamental principles of radiation protec- exposure.
tion applied to the nuclear medicine facilities workers. • Understand how to manage the solid and liquid waste pro-
• Recognize the importance of a radiation protection cul- duced in a nuclear medicine facility.
ture and implementation of the international basic safety • Define the classification of working areas and workers.
standards with regard to radiation protection of the nuclear
medicine staff.
• Define key responsibilities and main issues for ensuring
radiation protection of the nuclear medicine staff. 13.1 G
eneral Concepts in Radiation
• Identify and learn the main measurement units used in Protection
radiation protection.
• Define the main factors for planning/implementing a Nuclear medicine is the medical specialty based on the use
nuclear medicine facility. of radionuclides administered in the form of unsealed
compounds for diagnostic and therapeutic purposes to
A. C. Traino (*) patients with a wide variety of diseases.
Health Physics Unit, University Hospital of Pisa, Pisa, Italy
e-mail: c.traino@ao-pisa.toscana.it

274 A. C. Traino
The extensive use of radioactive compounds, especially in Table 13.1 Annual dose limits for workers and public (from ICRP)
the unsealed form, requires a careful observation of the Occupational Public
radiation protection principles both for patients and for per- Effective dose (mSv) 20 1
sonnel. The objectives of radiation protection continue to Equivalent dose (mSv) to
evolve. While national regulatory bodies are present in Eye lens 20 15
Skin 500 50
almost each country, a reasonably consistent approach can
Hands and feet 500
be considered to have general validity and is commonly
applied throughout the world: this system is called “system
of radiological protection,” as defined by the International This chapter deals only with occupational exposure, for
Commission on Radiological Protection (ICRP) which takes which ICRP introduced three fundamental radiation protec-
into account the scientific information coming from the tion principles:
United Nations Scientific Committee on the Effects of
Atomic Radiation (UNSCEAR) [1]. This system will be 1. Justification: any decision that alters the radiation expo-
described in this chapter, paying particular attention to its sure situations should do more good than harm.
application to nuclear medicine occupational exposure. 2. Optimization of protection: the likelihood of incurring
Whereas the detailed formulation of the radiation protection exposures, the number of people exposed, and the magni-
principles can be found in the ICRP publications, this chap- tude of their individual doses should all be kept as low as
ter describes a concise and simplified summary of these reasonably achievable (ALARA), taking into account
principles. economic and societal factors.
With its Publication 103 [2], the ICRP recommends a sys- 3. Limitation of doses: the total dose to any individual from
tem of radiological protection to cover all possible exposure regulated sources in planned exposure situations other
situations, by considering three possible conditions: (1) than medical exposure of patients should not exceed the
planned exposure, (2) emergency exposure, and (3) existing appropriate limits recommended by the ICRP.
exposure situations. Regarding nuclear medicine, only the
first condition can be taken into consideration. In fact, the All these three principles must be considered in radia-
use of radiation in nuclear medicine is a planned exposure tion protection of workers (occupational radiation protec-
situation under a mandatory regulatory control, for which tion). Recommended dose limits are given in Table 13.1
appropriate authorization from the regulatory body is neces- for workers and public. These limits are derived from
sary before starting operations. The so-called potential expo- ICRP 103 [2] (effective and equivalent dose to the skin and
sure (accidental spills of radioactive compounds and other hands and feet), from ICRP 118 [3], and from International
incidents in the workplace) can occur, but these remain part Atomic Energy Agency (IAEA) [4] documents (for the eye
of the planned exposure situation, and they must be consid- lens).
ered and described, with the corresponding planned solu- It should be noted that the limits on effective dose are
tions, in the application to the regulatory body for for the sum of the effective doses from external exposure
authorization. in the specified time period and the committed effective
ICRP defines as a “practice” a planned exposure in a dose from intakes of radionuclides in the same period. For
nuclear medicine facility, because “practice” means every adults, the committed effective dose is calculated for a
situation that can improve/minimize exposure to ionizing 50-year period after intake, whereas for children it is com-
radiation by optimizing the operating procedures. Regarding puted for the period up to reaching 70 years of age; addi-
nuclear medicine, ICRP classifies radiation exposure of indi- tional restrictions apply to occupational exposure of
viduals into three categories: pregnant women.
1. Medical exposure, i.e., exposure occurring in patients

treated with radionuclides for diagnosis or therapy and in Key Learning Points
individuals helping to support and comfort these patients • Knowledge of the radiation protection system: role
(caretakers) of ICRP, UNSCEAR, and IAEA.
2. Occupational exposure, i.e., exposure of workers
• Concept of “practice” as an activity suitable to
employed in nuclear medicine facilities improve the exposure to ionizing radiations.
3. Public exposure, i.e., exposure incurred by members of • Radiation protection principles: justification, opti-
the general public (e.g., inadvertent exposure of individu- mization, and limitation of radiation doses.
als staying close to the patient after discharge from the • Annual limits of radiation dose for workers and
nuclear medicine department after having been treated public.
with a radioactive compound, etc.)
13 Radiation Protection for Personnel and the Environment 275
13.2 Radiation Protection Units more closely reflects the biological effect on the irradiated
tissue or organ. This quantity is called the equivalent dose
13.2.1 Average Absorbed Dose HT, which is defined as:
H T = wR DT (13.2)
The fundamental quantity to be used in radiation protection
is the average absorbed dose DT in a tissue/organ. It can be where wR is the radiation weighting factor. The wR factor
expressed by: changes, depending on the type and quality of the radiation
interacting with the target. Table 13.2 reports the wR values
eT
DT = (13.1) for various types and quality of ionizing radiation.
mT The SI unit of equivalent dose is called sievert (J/kg).

Exposure from different types of radiation results in a total
where εT is the total energy imparted to a tissue/organ T and equivalent dose that is the sum of the equivalent doses
mT is the mass of that organ. DT is a physical quantity that from each type of radiation. It should be noted that HT can-
can be measured, and its measurement unit in the not be considered as a physical quantity, because it cannot
International System of Units (SI) is called gray (Gy): be directly measured. In fact, factor wR is an empirical
1 Gy = 1 J/1 kg. factor.
For low absorbed doses (when stochastic effects occur, a
condition that is typical of occupational radiation protec-
tion), the biological effects of ionizing radiations depend not 13.2.3 Effective Dose
only on the absorbed dose but also on the type (and quality)
of the particles interacting with the organ/tissue and on the The probability of stochastic effects depends not only on
radiosensitivity of the same organ/tissue. In order to take into the equivalent dose but also on radiosensitivity of the
account these two facts, two radiation protection quantities organ/tissue irradiated. Radiosensitivity of the organs of
that cannot be measured directly were introduced: the equiv- the human body is quantified through the tissue weighting
alent dose and the effective dose. factors wT. WT represents the relative contribution of an
For internal radiation exposure from radionuclides, the organ or tissue, T, to the total damage due to the stochastic
equivalent and the effective dose depend also on the biologi- effects resulting from a uniform irradiation of the
cal turnover and retention of the radionuclide. This is taken whole body.
into account in the committed dose quantities (equivalent The total tissue-weighted equivalent dose, E, is called
and effective). effective dose and is defined from the equation:
E = å wT H T (13.3)

13.2.2 Equivalent Dose where the sum is obtained by taking into account all the
organs sensitive to the induction of stochastic effects that
In radiobiology it is well established and accepted that have absorbed an equivalent dose HT.
densely ionizing radiation (e.g., α particles and neutrons) Table 13.3 reports the tissue weighting factors wT
will cause greater damage to a tissue/organ than γ rays and reported in ICRP 103 [2]. Note that the sum ΣwT = 1 (as
electrons when the target absorbs the same average absorbed expected, the sum of the probability of damage over the
dose, DT. In fact, dense ionization events mean higher prob- whole body must be 1).
ability of inducing irreversible damage to the irradiated The use of effective dose is useful in radiation protection.
organ/tissue. In fact, different exposure situations (e.g., internal versus
Thus, the average organ absorbed dose is multiplied by a external exposure to different types of radiation) can be com-
radiation weighting factor in order to obtain a quantity that bined and result in a single value, the effective dose.
Table 13.2 wR factors as reported in ICRP 103

Radiation type wR
Table 13.3 wT factors as reported in ICRP 103
Photons 1
Electrons, muons 1 Tissue wT ΣwT
Protons 2 Red bone marrow, colon, lung, stomach, breast, 0.12 0.72
Alpha particles, fission fragments, 20 remainder tissues
heavy nuclei Gonads 0.08 0.08
Neutrons Energy continuous dependent Bladder, esophagus, liver, thyroid 0.04 0.16
function Bone surface, brain, salivary glands, skin 0.01 0.04
276 A. C. Traino
13.2.4 Committed Dose officer, RPO. Radiation protection in a nuclear medicine

facility must be strictly linked to the medical activity in the
When radionuclides are incorporated in the body, the dose facility. This means that radiation protection in general must
absorbed by the body is related to the decay time of the consider patients and workers taking into account also eco-
radionuclide. Therefore, another quantity, defined as “com- nomic and social factors.
mitted dose,” has been introduced. From a radiation protec- Almost all countries have their own legislation in radia-
tion point of view, the committed equivalent dose can be tion protection. The regulatory systems require that any
expressed by: facility has a formal authorization to perform nuclear medi-
cine “practices” from the radiation protection regulatory
t0 +t
. institutions and that the persons working in nuclear medicine
H T (t ) = ò H T ( t ) dt (13.4) facilities have a specific training in radiation protection.
t0
The requirements to be fulfilled in order to grant such an
where t0 is the time of intake and τ is the time of integration. authorization vary from country to country, but in general
For adult workers τ = 50 years. A committed effective dose they comply with the requirements of the IAEA Basic
can be defined as: Safety Standards (BSS). The IAEA BSS system is a set of
published practical general recommendations taking into
E (t ) = åwT H T (t ) . (13.5) account the ICRP (and UNSCEAR) publications. This sys-

T tem requires that a set of responsibilities and duties be for-
mally assigned to the persons involved in working with
radionuclides.
Key Learning Points
• Knowledge of the absorbed dose and its SI unit.
• Knowledge of the radiation protection units (equiv- Key Learning Points
alent dose, effective dose, committed equivalent • Nuclear medicine procedures are subjected to radi-
dose, and committed effective dose) and their SI ation protection.
units. • Radiation protection in nuclear medicine requires a
specific authorization.
• Such authorization must be granted by the local
regulatory body; usually the regulatory rules follow
13.3 R
adiation Protection in Nuclear
the IAEA BSS publications.
Medicine
• The person responsible of the activity is the licensee;
he/she can delegate some specific activities con-
13.3.1 Introduction
cerning radiation protection to other persons.
• These persons (medical physicist, radiation protec-
Virtually all the possible procedures in the nuclear medicine
tion officer, nuclear medicine specialist, technolo-
facility must be subjected to radiation protection. These pro-
gist, radiopharmacist) must be specifically educated
cedures are:
and trained in radiation protection.
• Ordering radionuclides
• Unpacking and checking the shipment
• Storage of radionuclides 13.3.2 Formal Authorizations
• General rules for work in controlled and supervised areas and Responsibilities
• Preparation of radiopharmaceuticals
• Personal and workplace monitoring Before starting the activity, a nuclear medicine facility needs
• Internal transport of radionuclides to be formally authorized by the radiation protection regula-
• Management of radioactive waste tory body. The licensee is responsible of the application of
• Administration of radiopharmaceuticals to patients what the authorization requires. He/she can decide to assign
• Protection issues in patient examinations and treatments to other persons (nuclear medicine specialists, medical phys-
• Routine cleaning of facilities icists, technologists, radiopharmacists, radiation protection
• Decontamination procedures officers) some specific responsibilities in implementing radi-
• Care of radioactive patients ation protection in the nuclear medicine facility.
All the persons having responsibilities must possess the
Rules must be established and upgraded, under the required education level, must be trained, and must have the
responsibility of the licensee, by the radiation protection competence to perform their duties.
A radiation protection program must be adopted; the The opportunity to install ventilation systems (fume hood,
necessary resources must be given also for the education laminar-flow cabinet, or glove box) must be considered in
and training of the personnel working in the facility. the rooms where unsealed liquid or gaseous sources are
An example of radiation protection program can be found manipulated.
in [5]. In some cases and in some countries, legislation does not
allow the immediate release to the sewer system of aqueous
waste produced by the patients treated in the nuclear medi-
cine facility. In this case a dedicated tank system must be
13.4 Facility Design designed and used, following the authorizations for the dis-
charge. A separate bathroom for exclusive use by the patients
This is a duty/responsibility of the medical physicist, who to whom radiopharmaceuticals have been administered is
must take into account the following factors: recommended; the waste from this bathroom must be con-
nected to the dedicated tank system—if so indicated in the
• Safety of sources authorization requirements.
• Optimization of protection for staff and the general Additional details regarding floor planning and additional
public topics can be found in the IAEA Nuclear Medicine Resources
• Preventing uncontrolled spread of contamination Manual [6].
• Maintaining low background where most needed
• Fulfillment of national requirements regarding pharma-
ceutical work 13.4.2 Storage of Radioactive Sources
The licensee must establish a security system to prevent the

13.4.1 Location and General Layout loss, unauthorized use, or damage of sources. Radioactive
materials can be ordered only by the authorized persons. The
In the choice and design of the location of a nuclear medicine sources must be recorded in a document and stocked in dedi-
facility, the following factors must be addressed: (1) it must cated rooms, under the responsibility of the authorized per-
be readily accessible, especially for outpatients (who gener- sonnel. The regulatory body should promptly be informed in
ally constitute the majority of the patients), and (2) it should cases of lost or stolen sources.
also be located away from radiotherapy sources and other Possible accidents to the facility must be considered (fire,
sources of ionizing radiation (cyclotron), which can interfere water flood, earthquake), and the risks due do possible dis-
with the measuring equipment. persion of radionuclides in the environment must be consid-
The work areas must be separated from the room where ered and addressed.
patients who have already been administered with the radio-
pharmaceutical are waiting for the scan. The uncontrolled
spread of contamination must be avoided; therefore, the 13.4.3 Structural Shielding
rooms for preparation of radiopharmaceuticals must be far
away as possible from rooms for measurements and from the A possible shielding of walls, floors, and ceilings must be
patients’ waiting areas. considered, depending on the type and the quantity of radio-
The transfer of unsealed sources across the facility must nuclides used. In a PET/CT facility, structural shielding is
be avoided or reduced as much as possible. always necessary, due to the high energy of the annihilation
In general, the type of work to be performed and the char- radiation. Calculation of the degree of shielding required
acteristics and quantities of the sources employed must be must be performed by a qualified medical physicist ade-
considered. Based on this, the need of air changes, shielding quately trained in radiation protection. Radiation surveys
of walls, materials used for the walls, floors, and work should always be performed to ensure correctness of the
benches must be examined. calculations.
The rooms of the facility must be classified based on the
hazard as low, medium, and high hazard areas. The risk of
contamination must be reduced as soon as possible also by 13.4.4 Workplace Classification
containing the contamination and cleaning it, using floors
and work benches that are impermeable to water and chemi- With regard to occupational exposure, the BSS require the
cals, easily washable, and resistant. The floor cover should classification of workplaces as controlled areas or as super-
be curved to the wall. The walls should also be easily clean- vised areas.
able. Chairs and beds used in high hazard areas should be In a controlled area, individuals follow specific protective
easy to decontaminate. measures to control radiation exposures. The controlled area
278 A. C. Traino
must be clearly identified, and it is convenient to use existing information on the type and quantity of radionuclides con-
structural boundaries, which should already be considered at tained and on the external dose rate at a chosen distance from
the planning stage of a facility. the container must be supplied.
A supervised area is any area for which occupational If prescribed by the regulatory system, the liquid waste
exposure conditions are predictable and stable. They are kept must be retained in a recoil tank system until the radioactiv-
under continuous scrutiny, even though specific additional ity concentration reaches the values authorized by the regu-
protective measures and safety provisions are not normally latory body. After this period, they can be released into the
needed. sewage system of the city.
In a nuclear medicine facility, the rooms for preparation, For some radionuclides (e.g., 99Mo/99mTc generators),
storage (including radioactive waste), and injection of the return of the decayed source to the vendor is a good option,
radiopharmaceuticals are usually classified as controlled especially for long half-life radionuclides.
areas. Owing to the potential risk of contamination, the There must be separate bathrooms for the workers and for
imaging rooms and waiting areas for injected patients might the patients.
also be classified as controlled areas. The area housing a
patient to whom therapeutic amounts of activity have been
given will also be a controlled area.
The workplace must be monitored to check possible con- Key Learning Points
tamination of the surfaces. There are two kinds of possible • Knowledge of the main requirements of a nuclear
monitoring of the working places: monitoring of exposure medicine facility project, depending on the work
and monitoring of contamination. The first one can be per- done, on the radionuclides employed, and on their
formed by using instruments that measure external exposure characteristics.
(dosimeters, ionization chambers, Geiger counters) and the • Knowledge of the main characteristics of the site of
second one by recoiling the contamination on filters (e.g., by waste storage.
smear tests) and then counting them with a spectrometric • Shielding of the site according to energy (and quan-
probe. tity) of radionuclides employed.
Frequency of these measurements must be decided by the • Classification of the workplaces as controlled areas
medical physicist or the radiation protection officer, based on or supervised areas.
the work carried out in the facility and on the type and quan- • Solid and liquid waste management.
tity of radionuclides employed.
13.4.5 Waste Management Operations 13.5 Exposure of Workers
The management of waste is one of the main issues in the Detailed prescriptions on the workers’ exposure in a nuclear
radiation protection of workers and public in a nuclear medi- medicine facility can be found in the IAEA safety guides
cine facility. Usually the radionuclides used in nuclear medi- [7–9], which also include recommendations on how to meet
cine are short half-life radionuclides; thus, they can be the radiation protection standards. A summary of the relevant
retained in the facility until they decay at the background issues follows here below.
level, after which they can be released in the environment
without hazards for the public.
The contaminated waste can be either solid, liquid, or gas- 13.5.1 Sources of Exposure
eous. The first principle that a radiation protection program
must follow requires that the contaminated waste be reduced Occupational exposure to ionizing radiation in nuclear medi-
as much as possible, in order to avoid the release of radioac- cine is due to the external irradiation from the unsealed
tivity in the environment. The authorization of a nuclear sources and also to the possible introduction into the body
medicine facility must contain a section concerning the treat- (internal exposure) of radioactivity for ingestion or inhala-
ment of waste. tion inside the body.
Solid waste must be retained in specific containers (avail- The precautions against external irradiation depend on the
able near areas where the waste are generated), segregated in characteristics of the radionuclides employed. In fact, a lead-
specific locked areas far from the rooms of treatment until shielded apron similar to that used in radiology can be a good
their decay under the limits imposed from the regulation precaution against external irradiation from lower-energy
authorities. These containers must be clearly identified and radionuclides (i.e., 99mTc), but it is not equally effective as the
labeled, e.g., by listing the radionuclides contained. Specific energy of the photons increases (i.e., 131I).
Similarly, the hazards due to the internal incorporation of the responsibility to follow strictly the radiation protection
radionuclides depend on the decay type (α, β, or γ), on the rules given and to use properly the instruments, tools, and
decay constant, and on the energy of the source incorporated. devices that are supplied to them for radiation protection/
Furthermore, the biodistribution kinetics of the radionuclide radiation measuring purposes.
is very important to assess its effect from a radiation protec-
tion point of view.
The different activities (from unpacking the sources to 13.5.3 Pregnant Workers
administration of the radiopharmaceutical to the patient—as
also positioning the patient on the imaging table) contribute The dose limits for unborn children are usually the same as
to the radiation exposure of the worker. Usually the annual the public; thus, this limit must be applied also for the preg-
permitted doses are below 6 mSv. They are due mainly to the nant workers. Therefore, the possibility to remove the preg-
imaging and preparation-administration procedures. For this nant worker from her workplace in nuclear medicine must be
reason, it is very important to use adequate shielding for the seriously taken into consideration, especially due to the pos-
administration syringes and to work under a fume hood sibility, in these facilities, of internal contamination.
(whenever possible) when preparing the radiopharmaceuti-
cal for administration.
Spillage of radioactivity during the procedures must be 13.5.4 Protective Clothing
carefully avoided. If it occurs, the medical physicist and/or
the RPO must be immediately alerted. Adequate care must Protective clothing (gloves, lead-shielded aprons, glasses,
be taken when performing surgical procedures on patients shoes or overshoes, caps, masks for aseptic work) must be
who have been recently treated with radionuclides or when available to the workers in the facility. The usefulness and
performing autopsy of their corpse (especially after nuclear opportunity to use the protective clothing depend on the
medicine therapy procedures). works being carried out and on the characteristics of the
Adequate precautions must be used in cleaning all the radionuclides employed. This is especially true for the lead-
surfaces (doors, floor, walls) of a nuclear medicine facility. shielded aprons that can be very effective for non-high-
energy radionuclides.
13.5.2 Justification, Optimization, and Dose

Limitation 13.5.5 Personal Monitoring
All the procedures must be justified, taking into account the A personal and environmental monitoring program must be
principle that workers have not benefitted from their own assured under the supervision of the medical physicist and/or
radiation exposure. Justification of the procedures is thus the RPO. The responsibility to assure the monitoring program
important for both the patients and the workers. The risks in is of the licensee and the employer.
radiation work must be contained; however, they should not The RPO will decide which workers will be involved in
be greater than for any other works in a hospital setting. the program (which includes monitoring of external and
When justified, a procedure must be optimized, trying to internal exposure) and what is the most adequate monitoring
administer to patients the minimum possible activity com- frequency. Usually the workers involved in the program are
patible with the diagnostic/therapeutic value of the proce- persons who routinely work in nuclear medicine, including
dure performed. Such optimization requires the use of nurses.
shielding (when possible), but also a certain limitation of the Monitoring of external exposure usually implies the use
time and distance from the patient. of dosimeters capable to measure and retain the external
The doses reported above must be considered as maxi- radiation exposure. Thermoluminescent (TL) dosimeters are
mum limits of exposure. The effective dose absorbed must currently used to this purpose.
be further reduced under these limits as much as possible. Monitoring of the possible intake of radionuclides
Although the facility design, the use of adequate shielding (internal contamination) is rarely performed; such
barriers and coats, and good functioning of the fume hoods programs are usually implemented in those facilities where
are very important in the radiation protection of workers, a large amounts of 131I-iodide for the treatment of thyroid
training program and the education of workers can strongly disease are used. Monitoring of internal contamination due
help in implementing an effective radiation protection of the to 131I implies the use of an external probe to evaluate the
facility. activity present in thyroid. Based on this measurement, the
While the licensee and employer have the main responsi- committed effective dose can be calculated as described in
bility regarding limitation of the exposures, the workers have ICRP 78 [10].
280 A. C. Traino
13.5.6 Local Rules and Supervision References
After reaching a specific agreement with the workers (usu- 1. UNSCEAR. Effects of ionizing radiation. UNSCEAR 2006
Report to the General Assembly with scientific annexes. Volume 1.
ally through designated representatives), the employer and New York, NY: United Nations Scientific Committee on the Effects
licensee must: of Atomic Radiation; 2008.
2. ICRP. The 2007 recommendations of the International Commission
• Establish written local rules and procedures necessary to on Radiological Protection. ICRP publication 103. Ann ICRP.
2007;37(2-4):1.
ensure adequate levels of protection and safety for work- 3. ICRP. ICRP statement on tissue reactions/early and late effects of
ers and other persons. radiation in normal tissues and organs – threshold doses for tissue
• Include in the local rules and procedures the values of any reactions in a radiation protection context. ICRP Publication 118.
relevant investigation level or authorized level and the Ann ICRP. 2012;41(1/2):1.
4. IAEA. Implications for occupational radiation protection of the
procedure to be followed in the event that any such value new dose limit for the lens of the eye. IAEA-TECDOC No. 1731.
is exceeded. Vienna: International Atomic Energy Agency; 2013.
• Establish local rules and procedures, the protective mea- 5. International Atomic Energy Agency. Applying radiation safety
sures, and safety provisions known to those workers to standards in nuclear medicine. Safety reports series no. 40. Vienna:
IAEA; 2005.
whom they apply and to other persons who may be 6. International Atomic Energy Agency. Nuclear medicine resources
affected by them. manual. Vienna: IAEA; 2006.
• Ensure that work involving occupational exposure is ade- 7. International Atomic Energy Agency. Occupational radiation pro-
quately supervised and take all reasonable steps to ensure tection. IAEA safety standards series no. RS-G-1.1. Vienna: IAEA;
1999.
that the rules, procedures, protective measures, and safety 8. International Atomic Energy Agency. Assessment of occupational
provisions are observed. exposure due to intakes of radionuclides. IAEA safety standards
series no. RS-G-1.2. Vienna: IAEA; 1999.
All the procedures possible in the nuclear medicine facil- 9. International Atomic Energy Agency, Assessment of occupational
exposure due to external sources of radiation, IAEA safety stan-
ity (as listed above under the “Introduction” section) must be dards series no. RS-G-1.3, IAEA, Vienna 1999.
subjected to radiation protection. The rules must be estab- 10. ICRP. Individual monitoring for internal exposure of workers. CRP
lished and upgraded, under the responsibility of the licensee, publication 78. Ann ICRP. 1997;27(3-4):1.
by the RPO.
Key Learning Points

• Knowledge of the main sources of radiation expo-
sure in nuclear medicine.
• Relevance of the physical and chemical characteris-
tics of the radionuclides employed.
• Observation of the principles of radiation protec-
tion: justification, optimization, and limitation of
doses for workers employed in nuclear medicine.
• What to do with pregnant workers.
• Monitoring program of doses for workers: external
and internal irradiation.
• Duties and responsibilities of the employer and
licensee.
Essentials of CT Image Interpretation
14
Contents
14.1 Introduction.................................................................................................................................. 282
14.2 Visualization and Elaboration of CT Images............................................................................ 282
14.2.1 Multiplanar Reconstruction (MPR) .............................................................................................. 282
14.2.2 Maximum Intensity Projection (MIP) and Minimum Intensity Projection (MinIP) ..................... 282
14.2.3 3D Surface Rendering ................................................................................................................... 283
14.3 Image Analysis and Interpretation ............................................................................................ 283
14.3.1 Density Values and Hounsfield Units (HU) .................................................................................. 283
14.3.2 Morphological Evaluation of CT Images ..................................................................................... 284
14.3.3 Patterns of Contrast Enhancement................................................................................................. 285
14.3.4 Artifacts and Pitfalls in CT Imaging.............................................................................................. 286
14.4 Neck .............................................................................................................................................. 286
14.4.1 Cystic Lesions ............................................................................................................................... 287
14.4.2 Neoplastic Lesions ........................................................................................................................ 287
14.4.3 Infectious Diseases ........................................................................................................................ 288
14.4.4 Thyroid and Parathyroid Gland Pathologies.................................................................................. 288
14.5 Lungs and Airways...................................................................................................................... 288
14.5.1 Solitary Pulmonary Nodule: CT Image Interpretation .................................................................. 289
14.5.2 Infectious Diseases ........................................................................................................................ 290
14.5.3 Diffuse Lung Diseases: Patterns of Interstitial Involvement ......................................................... 290
14.6 Mediastinum and Pleura............................................................................................................. 292
14.6.1 Differential Diagnosis of Mediastinal Masses............................................................................... 292
14.6.2 Pleural and Chest Wall Tumors ..................................................................................................... 294
14.6.3 Nonneoplastic Pleural Diseases..................................................................................................... 294
14.7 Abdomen and Pelvis .................................................................................................................... 295
14.8 Liver.............................................................................................................................................. 295
14.8.1 Benign Lesions .............................................................................................................................. 295
14.8.2 Malignant Lesions ......................................................................................................................... 298
14.9 Biliary Tree................................................................................................................................... 301
D. Caramella (*)
Diagnostic and Interventional Radiology,
Department of Translational Research and Advanced Technologies
in Medicine and Surgery, University of Pisa, Pisa, Italy
e-mail: davide.caramella@unipi.it
M. Revelli
Department of Diagnostic Imaging and Laboratory Medicine,
AUSL di Reggio Emilia – IRCCS, Reggio Emilia, Italy
A. Villa
Unit of Radiology 2, ASL 5 “Spezzino”, Sarzana, La Spezia, Italy

282 D. Caramella et al.
14.10 Pancreas...................................................................................................................................... 301
14.10.1 Pancreatic Cystic Tumors ........................................................................................................... 301
14.10.2 Pancreas Solid Neoplasms........................................................................................................... 303
14.11 Spleen.......................................................................................................................................... 304
14.12 Adrenal Glands .......................................................................................................................... 304
14.13 Kidney......................................................................................................................................... 306
14.13.1 Renal Cystic Lesions ................................................................................................................... 306
14.13.2 Renal Non-Cystic Lesions........................................................................................................... 307
14.14 Small Bowel Disease .................................................................................................................. 310
14.15 Colorectal Cancer ...................................................................................................................... 311
14.15.1 Morphologic Patterns of Presentation ......................................................................................... 311
14.16 Pelvic Lesions.............................................................................................................................. 312
14.17 Bone Lesions............................................................................................................................... 313
14.17.1 Primary Bone Lesions ................................................................................................................. 313
14.17.2 Secondary Bone Lesions ............................................................................................................. 314
14.17.3 Miscellaneous Bone Lesions ....................................................................................................... 315
14.18 Abdominal Lymphadenopathies .............................................................................................. 315
Further Reading 316
Learning Objectives allow better assessment of anatomic details. The most fre-
• Discuss the non-enhanced and contrast-enhanced acquisi- quently used reconstructions are multiplanar reconstruction
tion techniques required to correctly interpret a CT exam. (MPR), volume rendering (3D) reconstruction, and also
• Describe the principal thoracoabdominal lesions with an maximum intensity projection (MIP) or minimum intensity
organ-oriented approach. projection (MinIP) reconstructions.
• Formulate a differential diagnosis on the basis of the
attenuation and contrast enhancement features of thora-
coabdominal lesions. 14.2.1 Multiplanar Reconstruction (MPR)
Multiplanar reconstructions are bidimensional images gen-

14.1 Introduction erated using data from acquired in the helical dataset.
Coronal, sagittal, oblique, and curved reconstructions are
The combination of information derived from positron- obtained interpolating data derived from adjacent voxels
emission tomography (PET) and anatomic information from (Fig. 14.1a). They can be useful to evaluate the relationships
computed tomography (CT) provides an important tool in between a neoplastic mass and the surrounding structures,
the management of patients with cancer and cardiovascular and they have a role in advanced CT applications, such as
disease. To effectively utilize CT images, knowledge of nor- vessel evaluation in CT angiography.
mal anatomic features and common pathological alterations
that can be assessed with CT is required. This chapter will
focus on the principles of performing and interpreting CT 14.2.2 Maximum Intensity Projection (MIP)
images. and Minimum Intensity Projection
(MinIP)
14.2 isualization and Elaboration of

V These are volume rendering techniques that allow definitions
CT Images of volumes of interest (VOI): images are obtained by project-
ing the VOI on a visualization plane selecting high Hounsfield
Multidetector helical CT scanners recording >64 slices are numbers for MIPs and low Hounsfield numbers for MinIPs,
the standard for image acquisition. In common clinical prac- providing an optimal contrast. MIPs are mainly used for CT
tice, a reconstructed slice thickness of 5 mm is employed for angiography applications and for the evaluation of pulmonary
image visualization on the axial plane. Thinner slices of parenchyma (Fig. 14.1b), while MinIPs play a role in the
0.625 mm can be acquired to improve image quality and study of air-filled structures such as the tracheobronchial tree.
14 Essentials of CT Image Interpretation 283
a b c
Fig. 14.1 Multiplanar reconstruction on the coronal plane of the chest (a); maximum intensity projection reconstruction on the axial plane for the
evaluation of pulmonary parenchyma and vessels (b); 3D volume rendering of the chest for the evaluation of the bony structures of the chest (c)
14.2.3 3D Surface Rendering surrounding tissues. CT images can be displayed using dif-
ferent gray level settings, the so-called windows, in order to
It is represented by a surface image providing a realistic improve contrast between structures.
three-dimensional depiction of a structure of interest inside On non-enhanced CT (NECT, CT images acquired with-
the acquired volume (Fig. 14.1c). 3D surface rendering out the use of iodinated contrast material), specific tissue
has the advantage of immediate and easy interpretation for components can be characterized based on native attenuation
surgeons during surgical planning and plays a role in virtual values.
endoscopy, providing reconstructions that allow interactive Air collections can be easily identified using a pulmonary
navigation. window that is routinely used for the assessment of lung
parenchyma. This can prove useful also in the evaluation of
extrathoracic areas, for example, to detect the presence of
Key Learning Points free air in the peritoneal cavity. Windows settings commonly
• Helical CT represents the standard for image acqui- used for the evaluation of parenchymas and soft tissues may
sition, and modern multidetector scanners can fail to identify gaseous areas, or low attenuation fat areas
improve image quality by using various image may be mistakenly interpreted as air (Fig. 14.2): a CT num-
algorithms. ber < −150 HU in soft tissues is pathognomonic for the pres-
• Multiplanar reconstruction (MPR), volume render- ence of gas. The search for air must always be carried out in
ing (3D) reconstruction, and maximum intensity the study of trauma patients or in cases of suspected perfora-
projection (MIP) or minimum intensity projection tion, and it should be remembered that the air tends to settle
(MinIP) reconstructions are widely used in diag- in the upper parts of the body. The presence of air inside a
nostic imaging to obtain additional information and fluid mass should raise the suspicion of an abscess.
for advanced applications, such as CT angiography Fat presents a density value of about −100 HU, and adi-
and virtual endoscopy. pose infiltration of various tissues (liver, pancreas, muscles)
has to be suspected when a reduction of expected density
values is noticed. The presence of fat in a lesion is important
to establish the diagnosis in many diseases, and its percent-
14.3 Image Analysis and Interpretation age may vary according to the type of lesion: a mass entirely
composed of adipose tissue is typically a lipoma, while the
14.3.1 Density Values and Hounsfield presence of other solid components may suggest the diagno-
Units (HU) sis of a more aggressive lesion.
Regarding fluid, pure water has a density of 0 HU,
CT numbers are normalized values of the calculated X-ray although – due to technical factors and variable composition
absorption coefficient of a voxel in a CT image, expressed in of bodily fluids – values ranging from −10 to 10 HU are
Hounsfield units (HU); they play a central role in the charac- consistent with watery fluid. Denser fluids can demonstrate
terization of different tissues, since air, fat, fluids, and blood higher attenuation values possibly overlapping with those of
have different attenuation values. Depending on the CT soft tissues; for these reasons, often a cystic lesion can be
number of voxels, their corresponding pixels are
safely distinguished from a solid lesion only after the admin-
defined as hyperdense, isodense, or hypodense with regard to istration of contrast material.
a b
Fig. 14.2 CT of a 55-year-old man with abdominal pain. On the soft tissue window (a), the free air collection may be difficult to depict, but the
use of the pulmonary window (b) makes it easy to identify the air collection adjacent to the stomach (arrow)
Blood attenuation is higher than water and depends on the ters of the lesion, possibly using MPR to obtain the most
concentration of hemoglobin; in case of anemia, the blood suitable planes for the measurements (Fig. 14.3); further-
may appear less hyperdense. Hemorrhage density is influ- more, a volumetric assessment can be made, by using multi-
enced by the time passed since the onset, the presence of plane measurements and automated software.
clots, and the site of bleeding: a recent hemorrhage is typi- When considering the margins of a focal lesion, the differ-
cally seen as a hyperdense collection (>80 HU). Subacute ence in tissue attenuation can help identify tissue borders;
hemorrhage tends to show a reduction of blood attenuation well defined margins are often associated with benign lesions,
due to lysis of red cells and to hemoglobin reabsorption. whereas ill-defined margins may be a sign of infiltration.
MRI is more sensitive in estimating the age of a hemorrhage, Sometimes it is difficult to attribute a lesion uniquely to a
since blood has specific T1 and T2 signal characteristics cor- specific organ, especially in the absence of adipose cleavage
relating with the stage of evolution. plans that allow a clear separation from the adjacent organs:
A density higher than 100 HU suggests the presence of in such cases, the angle between the lesion and the organ can
calcific deposits. Calcifications may be the consequence of point to the diagnosis, since an obtuse angle indicates the
many pathological processes, and their presence may be origin of the lesion from the organ in question, while an
indicative of the benign or malignant nature of various acute angle suggests an extrinsic lesion.
lesions, with specific features in the various body districts. A key feature to be evaluated is the differentiation between
Metallic foreign materials show the highest attenuation val- expansive lesions that tend to displace and compress the
ues, and they are often associated with peculiar imaging adjacent organs and infiltrative lesions that have ill-defined
artifacts. contours and extend to adjacent structures: in particular, vas-
cular infiltration represents a key element in the staging of
many cancers, often affecting surgical management, since a
14.3.2 M
orphological Evaluation of tumor that surrounds a vessel for more than 50% of its cir-
CT Images cumference is usually consistent with vascular infiltration.
In some cases it may not be easy to distinguish between a
The accuracy of image evaluation is related to technical tumor and an inflammatory process, and the evaluation of
aspects (i.e. pixel size, reconstruction algorithm), and is clinical parameters and laboratory tests should be the basis
influenced by acquisition parameters: as a consequence of for the differential diagnosis. In particular, the most chal-
incorrect acquisition protocols, some lesions can be difficult lenging situations occur in the lungs (differential diagnosis
to measure, and a comprehensive morphological evaluation of pneumonia and bronchial-alveolar carcinoma), in the pan-
cannot always be achieved. creas (chronic pancreatitis vs adenocarcinoma), and in gen-
Size evaluation is one of the most reproducible measure- eral in the follow-up of oncologic patients when it is
ments in CT images: focal lesions are usually measured by indispensable to distinguish between recurrence and surgery-
tracing two perpendicular lines along the two largest diame- related fibrosis.
a b
54,05 mm
43,72 mm 89,32 mm
45,77 mm
Fig. 14.3 Use of MPR reconstructions to obtain the largest diameters (b), we have a better and more reliable depiction of anteroposterior and
of an inguinal enlarged lymph node. On the axial plane (a), we can craniocaudal diameters
measure the largest latero-lateral diameter, while on the sagittal plane
14.3.3 Patterns of Contrast Enhancement vals of attenuation values in a region of interest (ROI) usually
placed at different level of the aorta depending on clinical
Contrast enhanced CT (CECT) is useful to detect and char- indications. Image acquisition automatically starts when a
acterize lesions in different anatomic sites. Multiphase CT predefined HU value is reached in the ROI.
protocols are routinely used and tailored on the clinical indi- A multiphase contrast enhancement CT protocol consists
cation; they may include a non-enhanced CT scan followed of various acquisitions that may include NECT (if needed),
by contrast-enhanced CT acquisitions at various intervals early and late arterial phase, portal phase, equilibrium phase,
after contrast administration. and delayed phases.
NECT is useful to detect stones in the urinary tract and to The early arterial phase is obtained at the time of maxi-
differentiate cystic lesions from solid lesions (i.e., renal cyst mum vessel enhancement (usually between 25 and 30 s after
from solid neoplasm), hematomas and hemorrhagic cysts injection of contrast), and it is used to evaluate the arterial
from clear fluid collections and simple cysts (between 0 and vessels in CT angiography. Late arterial phase is acquired
15 HU is fluid collection; more than 60 HU can be consid- about 10–15 s later and is commonly used for the detection
ered as hemorrhagic), and fat-containing lesions (density is of hypervascular tumors and for the assessment of organs
<0 UH) such as lipomas, renal angiomyolipoma, and lipid- such as the pancreas. Venous portal phase starts about
rich adrenal adenoma (density <10 HU is sufficient for diag- 30–35 s after the end of the late arterial phase and is very
nosis) from other solid lesions. NECT is also important to useful in the evaluation of parenchymal organs. The assess-
evaluate the amount of contrast enhancement when comment of contrast enhancement in late arterial and venous por-
pared with CECT acquisitions. tal phase is the mainstay to characterize solid lesions based
CECT scans are usually acquired at standardized times on the recognition of specific enhancing patterns. Equilibrium
after contrast administration to evaluate dynamic patterns of phase is obtained at 3–5 min after administration of contrast,
enhancement that may help in establishing a definite diagno- and it is commonly acquired in liver evaluation protocols to
sis. Since contrast enhancement kinetics may differ aid in lesion characterization.
among different patients, owing to both physiological and The iodinated contrast material is normally eliminated by
disease-related factors, a bolus-tracking technique is the kidneys, and it progressively opacifies the urinary col-
often used to ensure optimal enhancement. lecting system. Post-contrast acquisition in the delayed
Bolus tracking is performed with the application of dedi- phases, between 7 and 10 min after contrast administration,
cated software that allows a measurement at short time inter- allows the evaluation of the urinary system filled by
highly concentrated contrast material, which can be useful to

depict anatomic details, demonstrate intraluminal patholo-
gies, or other abnormalities of the urinary tract.
Parenchymal focal lesions may have a basal density at
NECT that makes them difficult to distinguish from the adja-
cent parenchyma; therefore, their vascular kinetics on
CECT is often a key element for guiding differential
diagnosis.
Based on contrast enhancement, focal lesion can be
divided as follows:
• Cystic lesions have variable attenuation values on NECT

based on their fluid composition, and do not exhibit signifi-
cant contrast enhancement on CECT. Small round cystic Fig. 14.4 Metal artifact, due to the presence of bilateral hip
lesions can present an apparent increase in attenuation val- prostheses
ues in the central portion after contrast administration
called pseudoenhancement, an artifactual phenomenon due (bones); motion or respiratory artifacts determine image
to the particular pattern of X-ray attenuation by surround- blurring, and detector size may cause linear and nonlinear
ing enhancing parenchyma. For this reason, a limited partial volume artifacts. Contrast media can also be a source
increase in attenuation after contrast (up to 15 HU) can still of pitfalls; as an example, an incomplete opacification of
be consistent with a cystic lesion. The presence of enhanc- venous vessels that can be observed in the arterial phase may
ing solid parts within a cyst should raise the suspicion of a occasionally mimic thrombosis.
complex cystic lesion that could be neoplastic in nature.
• Hypervascular lesions exhibit a stronger enhancement
compared to the surrounding parenchyma during the late Key Learning Points
arterial phase: typical examples of hypervascular tumors • Density characteristics, expressed in Hounsfield
are hepatocellular carcinoma, renal carcinoma, and hyper- units (HU), play a central role in the identification
vascular metastases. Lesions with a hypervascular behav- of different tissues, since air, fat, watery fluids, and
ior may show various patterns of contrast enhancement fresh blood have specific attenuation values.
that can range from homogeneously enhancing lesion to • Several morphological criteria, such as size and
marked heterogeneous masses. A particular pattern of margins, can be used in the assessment of CT
contrast enhancement is represented by “rim enhance- images.
ment” which consists of a continuous peripheral vascular- • Different image windowing (lung, abdomen, medi-
ization with a relative hypovascular central portion. Rim astinum) can lead to a better evaluation of different
enhancement is a typical finding in liver metastases. In lesions in various organs.
portal venous and equilibrium phases, hypervascular • Tumors may have specific patterns of contrast
lesions can demonstrate variable enhancement compared enhancement, and the knowledge of such character-
to late arterial phase; the washout phenomenon is a fre- istics during the different phases (arterial, portal,
quent occurrence in malignant hypervascular lesions, and venous) is essential to reach the correct diagnosis.
it is seen as a reduced attenuation relative to surrounding • Based on contrast enhancement, focal lesions can
parenchyma following the maximum enhancement. be divided into cystic, hypovascular, and
• Hypovascular lesions exhibit less vascularity than the adja- hypervascular.
cent parenchyma resulting in hypo-attenuation areas typi- • Recognition of CT artifacts is fundamental to dis-
cally seen in portal venous and equilibrium phases. tinguish them from pathological entities.
14.3.4 Artifacts and Pitfalls in CT Imaging
CT images may suffer from several kinds of artifacts. The 14.4 Neck
best known is the metal artifact, due to the fact that metal
causes high Xray scatter that limits the evaluation of sur- CT is a widely used imaging technique for the evaluation of
rounding tissues (Fig. 14.4). In addition, beam-hardening the neck; however, in some cases, the better contrast resolu-
artifacts can cause dark streaks between dense objects tion between soft tissues provided by MRI can help delineate
anatomical structures and characterize pathological condi- cranial extension, and perineural spread (along the
tions. CT is the modality of choice for the study of the lower mandibular nerve).
part of the neck, especially in debilitated and fragile patients, • Oropharyngeal carcinoma
such as those in intensive care units. Furthermore, CT may They are usually more aggressive and less differentiated
be preferable in patients with suspicion of osseous involve- than nasopharyngeal carcinomas, with an insidious onset,
ment, due to its ability to evaluate the cortical bone. A short often presenting with symptoms only in advanced and
overview of principal cystic and solid lesion of the neck will metastatic stage. They can originate from the basis of the
be provided in this paragraph. tongue, from the tonsillar region (Fig. 14.5) or from the
posterior wall. CT can delineate the relationship with the
surrounding structures as well as the invasion of neuro-
14.4.1 Cystic Lesions vascular bundle and soft tissue layers.
• Oral cavity carcinoma
A wide variety of cystic lesions may be found in the neck, They are often asymptomatic and may involve the lips,
and the knowledge of spatial and fascial neck anatomy is the floor of the mouth, the oral segment of the tongue, the
essential in order to formulate a correct differential diagno- oral mucosa or gums, and the palate. CT plays a role in
sis. Thyroglossal duct cysts appear as uniloculated fluid the staging of these tumors, depicting mucosal, bone,
masses in the anterior neck, with thin walls and slight perineural, and deep invasion.
peripheral enhancement. Wall focal thickenings or protrud- • Hypopharyngeal carcinoma
ing nodules may indicate malignant evolution. Branchial They usually develop in the submucosal layer, and CT
cleft cysts are displayed as unilateral well-defined hypodense may identify tumors that are not visible at endoscopy,
masses; walls may thicken in case of infection, and their providing information such as tumor volume (important
enlargement may lead to a compression of surrounding for treatment planning), contralateral extension, cartilagi-
structures. Thymic cysts are rare and may be seen anywhere nous and neurovascular invasion, and extralaryngeal
between the mandibular angle and the mediastinum: CT spread.
shows a fluid uni- or multiloculated mass with elongated • Laryngeal carcinoma
shape next to the carotid space; their size may reach 20 cm. They are usually squamocellular tumors related to alcohol
Cystic lymphangioma presents as a mass characterized by and tobacco abuse; they often develop in the submucosa,
multiple ill-defined cystic spaces, usually located in the pos- so endoscopy has a limited role. They can be surrounded
terior cervical space; it may show increased density when by a perilesional inflammation, so CT may overestimate
infected. Other rare cystic lesions of the neck include der- their size. They can be classified as supraglottic, glottic,
moid and epidermoid cysts, laryngocele, and Tornwaldt
cysts.
14.4.2 Neoplastic Lesions
A wide variety of benign tumors may affect the neck, includ-

ing neurogenic tumors, such as schwannomas and neurofi-
bromas, and mesenchymal tissue tumors, such as lipomas,
hemangiomas, and desmoid or fibrous tumors.
Regarding malignant lesions, the presence of lymph node
metastasis is very important for prognostic stratification. CT
can evaluate nodal involvement in staging head and neck
tumors, but more accurate techniques are currently
being developed for a more precise presurgical evaluation.
Malignant tumors of the neck are:
• Nasopharyngeal carcinoma
Mainly represented by squamocellular carcinomas with
preferential involvement of the lateral walls (fossa of
Rosenmuller). CT plays an important role in the evalua- Fig. 14.5 Oropharyngeal carcinoma originating from the left tonsillar
tion of deep-layer invasion, skull base erosions with intra- region (arrow)
and subglottic tumors, each one with specific TNM stag-

ing, spread pattern, and lymphatic drainage. CT may indi-
cate their volume, the invasion of different planes and the
extralaryngeal spread.
• Lymphomas
• Mesenchymal tissue and neurogenic malignant tumors
14.4.3 Infectious Diseases
CT plays a role in the evaluation of cases in which there is no

response to antibiotic therapy and it can help in differentia-
tion of complicated forms. Cellulitis may present as a soft
tissue mass showing marked enhancement after the adminis-
tration of contrast material, with edematous thickening of the
skin and obliteration of fascial and fat layers. Air collection
indicates the presence of gas-producing bacteria. Suppurative Fig. 14.6 Anaplastic thyroid carcinoma appearing as an ill-defined
lymphadenitis is characterized by node enlargement showing mass showing calcification and necrotic areas, with invasion of the sur-
signs of colliquative necrosis and surrounding edema. rounding structures and dislocation of the trachea and of the other upper
Abscesses are hypodense masses with capsular enhance- mediastinal structures
ment. Complications may include septic thrombosis of the
jugular veins and mediastinitis. Other granulomatous dis- between benign and malignant tumors are frequent
eases, such as tuberculosis, sarcoidosis, and cat scratch dis- and other investigations are often necessary to achieve a
ease, or conditions like infectious mononucleosis or correct diagnosis.
Castleman disease may all cause nodal involvement.
14.4.4 T
hyroid and Parathyroid Gland Key Learning Points
Pathologies • CT is a widely used imaging technique for the eval-
uation of the lower part of the neck, especially in
CT is not the modality of choice for the study of the thyroid; fragile patients and for the evaluation of osseous
however, during a CT examination for other indications, involvement.
important information about the thyroid gland can be col- • A wide variety of cystic lesions may be found in the
lected. Furthermore, an accurate evaluation of the thyroid neck, and the knowledge of neck anatomy is essen-
function should be performed before the administration of tial to reach a correct diagnosis.
iodinated contrast media. • CT can evaluate nodal involvement in staging head
A goiter is a benign enlargement of the thyroid gland, and neck tumors.
often asymmetric, characterized by mixed solid and cystic • CT plays a role in the evaluation of infectious neck
nodules; calcifications and hemorrhage may be present, and disease and can help in diagnosing complicated
contrast enhancement is usually patchy. forms.
Benign thyroid nodules are frequent, mainly represented
by colloid cysts. Follicular adenomas appear as hyperdense
well-defined masses, and they are virtually indistinguish-
able from a nodular goiter. Papillary carcinoma is the most 14.5 Lungs and Airways
common thyroid tumor, and it can be multifocal or bilat-
eral, with cystic, hemorrhagic, or calcific components. If plain radiographs still represent the first diagnostic modal-
Follicular carcinomas can be both capsulated and diffusely ity for the evaluation of the chest, CT is the most sensitive
infiltrating, with frequent invasion of vascular structures. technique for the assessment of pulmonary parenchyma,
Anaplastic carcinoma is more common in the elderly, and playing a central role in diagnosis, staging, and follow-up of
it can show calcification and necrotic areas, with invasion pulmonary and mediastinal cancers. High-resolution CT can
of surrounding structures (Fig. 14.6). Medullary carcinoma be useful to evaluate the presence of diffuse lung disease,
is rare and may be part of a MEN syndrome. CT findings of allowing an accurate analysis of pulmonary interstitium.
thyroid carcinomas are unspecific, and overlapping features Furthermore, CT can play a role in the evaluation of airways,
and it can be used to guide interventional procedures, such as sidered a benignity criterion. Peripheral subpleural loca-
biopsies or drainages. In this paragraph, we will provide an tion is also considered indicative of benignity.
easy and immediate guide to the interpretation of frequent • Shape and margins
entities such as the solitary pulmonary nodule, infections and Polygonal shape and high three-dimensional ratio (indi-
interstitial diseases. cating that the lesion is relatively flat) are considered
benign features. Smooth margins are usually associated
with benign nodules, while scalloped or lobulated mar-
14.5.1 S
olitary Pulmonary Nodule: CT Image gins may indicate a suspicious lesion; the corona radiata
Interpretation sign, a thin, irregular, and spiculated edge extending to
the periphery of the nodule, is frequently associated with
A solitary pulmonary nodule is defined as a rounded, well- malignant lesions.
defined opacity with a diameter less or equal to 3 cm, with- • Calcifications
out hilar or mediastinal interface and presenting without There are several patterns of benign calcifications, includ-
concomitant lymphadenopathies, atelectasis, or pleural ing diffuse, central, laminated, and popcorn calcifications,
effusion (Fig. 14.7). Differential diagnosis and manage- often seen in granulomatous diseases; calcifications pat-
ment of this entity is a matter of debate and depends on terns not belonging to these groups should not be consid-
whether the lesion is malignant, so it is crucial to establish ered as a sign of benignity.
imaging features to distinguish between benign and malig- • Nodule composition
nant nodules. We will review the main features that need to Nodules containing a ground-glass component are more
be considered in the diagnostic approach to the solitary pul- likely to be malignant, and the odds rise with mixed solid
monary nodule. and ground-glass components, while pure solid or pure
ground-glass nodules have lower rates of malignancy. The
• Size, location, and volume air bronchogram sign can also be associated with malig-
Lesions with a diameter larger than 3 cm are defined nant pulmonary nodules (most commonly bronchoalveo-
as masses and must be considered as malignant until lar cell carcinoma and adenocarcinoma).
proven otherwise. Comparison with previous imaging • Contrast enhancement
studies is mandatory to establish the size evolution of the Increase of density equal or less than 15 HU is considered
lesion: dimensional stability over a 2-year period is con- index of benignity. Higher enhancement, along with other
a b
10.03 mm
Fig. 14.7 Solitary pulmonary nodule (10 mm) in the middle lobe (pulmonary window in a). This nodule was an incidental finding. Soft tissue
window shows a central calcification (b). According to guidelines, a 12-month follow-up was performed
Table 14.1 Differential diagnosis of pulmonary nodules

Granulomatous diseases Tuberculosis
Histoplasmosis
Sarcoidosis
Benign tumors Hamartomas
Pulmonary pseudotumors
Malignant tumors Peripheral bronchial carcinoma
Bronchoalveolar carcinoma
Metastases (usually multiples)
Carcinoids
Kaposi sarcoma
Lymphoma
Other conditions Septic embolism
Intrapulmonary lymph nodes
Arteriovenous malformations
Round atelectasis
suspicious findings, requires further diagnostic work-up

and histological assessment.
The Fleischner Society periodically updates guidelines Fig. 14.8 Focal area of parenchymal consolidation with “air broncho-
gram” sign in the middle lobe, due to bacterial pneumonia
for the management of pulmonary nodules, considering
characteristics such as size, density, and location, stratified
for high- and low-risk patients. The 2017 updated guidelines chyma due to miliary spread; excavations with thin or thick
can be consulted online at https://fleischnersociety.org. walls, and enlarged mediastinal lymph nodes with central
About 50% of pulmonary nodules that undergo surgery are necrosis and intense peripheral rim enhancement.
benign. A schematic differential diagnosis of pulmonary
nodules is reported in Table 14.1.
14.5.3 D
iffuse Lung Diseases: Patterns
of Interstitial Involvement
14.5.2 Infectious Diseases
To understand the pathological alterations of interstitial lung
Pulmonary infectious diseases usually do not require CT diseases, it is important to know the anatomic organization
evaluation. However, CT can help identify complications or of the broncho-vascular structures (large bronchi, pulmonary
exclude predisposing conditions and monitor the progression vessels, and central interstitium) and of the secondary lobule,
of disease and the therapeutic response. Clinical features can the most peripheral parenchymal unit in which peripheral
point to an etiologic diagnosis, for example in debilitated or interstitium supports the fusion between airways and vascu-
immunocompromised patients. lar structures. Interpretation of interstitial diseases is mainly
Typical CT appearance of bacterial pneumonia includes based on the type and degree of involvement of the second-
areas of parenchymal consolidation with air bronchogram ary lobule. Vessels appear on CT images as linear opacities
(Fig. 14.8): these opacities can be focal or multiple, merging to (or circular one when viewed in section), that
segmentary or lobar involvement of air spaces. Central exca- branch dichotomously from the center to the periphery, side
vations and abscesses may develop. Atypical nonviral pneu- by side with the bronchial structures; on the other side, pul-
monias (i.e., caused by M. pneumoniae) are characterized by monary veins have a distinct course and show monopodic
centrilobular nodules (“tree in bud,” see the chapter Diffuse confluence of the minor branches into the major ones. The
Lung Disease: Patterns of Interstitial Involvement), air trap- secondary lobule measures about 1–2 cm and contains 5–15
ping phenomena, and mosaic attenuation, in association with pulmonary acini; it is supplied by a terminal bronchiole
ground-glass opacities and lobular parenchymal consolida- paired with a centrilobular artery (centrilobular area), while
tions. Viral infections show areas of patchy parenchymal pulmonary veins and lymphatics are located in the peripheral
attenuation and diffuse “tree in bud” aspects. CT features of interlobular septa (perilymphatic area).
tuberculosis may include acinar opacities with peribronchial CT interpretation of interstitial involvement requires a
distribution; nodular involvement of all pulmonary paren- structured approach considering the dominant pattern
Table 14.2 Differential diagnosis of diffuse lung disease according to • Random distribution
distribution within the lung Nodules not showing a prevalent distribution pattern,
Upper Sarcoidosis, silicosis, pneumoconiosis, such in cases of metastases or miliary spread of granulo-
distribution centrilobular emphysema, chronic hypersensitivity
matous diseases.
pneumonia, Langerhans cell histiocytosis
Lower Pulmonary edema, panlobular emphysema,
distribution aspiration-related conditions, usual interstitial The high-attenuation pattern can be represented by
pneumonia ground-glass opacity or consolidation on the basis of the par-
Central Sarcoidosis, bronchitis, cardiogenic pulmonary tial or complete increase of lung density.
distribution edema
Peripheral Chronic eosinophilic pneumonia, hematogenous
distribution metastasis, usual interstitial pneumonia,
• Ground-glass opacity
organizing pneumonia Ground-glass opacity can be the result of the alveolar fill-
ing by fluid, fibrosis, or cellular components, as well as of
marked thickening of alveolar walls or surrounding inter-
(reticular, nodular, high attenuation, low attenuation), its stitium: for these reasons, a ground-glass opacity is a
location within the secondary lobule (i.e., centrilobular or rather unspecific finding, and its distribution can be a
perilymphatic), its distribution in the lung (upper vs lower helpful feature to guide the differential diagnosis. Ground-
lobe, central vs peripheral location, Table 14.2), and the pres- glass opacity can be associated with a reticular pattern of
ence of additional findings such as pleural effusion or septal thickening, forming the so-called crazy paving
enlarged lymph nodes. (i.e., in alveolar proteinosis). Areas with increased densi-
The reticular pattern is characterized by septal thickening ties with lobular distribution alternated with non-affected
or can be related to advanced fibrosis: areas are referred as mosaic attenuation: this finding is
usually the expression of an obstruction of small airways,
• Thickening of the interlobular septa causing a compensatory oligemia with associated air trap-
Fluid, fibrous tissue or cell infiltration can lead to septal ping phenomena.
thickening and subsequent reticular opacity, which can be • Consolidation
smooth (i.e., in interstitial pulmonary edema) or nodular This condition is typically related to airspace diseases: acute
(i.e., in sarcoidosis). consolidation can be seen in pneumonias, hemorrhage, or
• Honeycombing cardiogenic pulmonary edema, while chronic consolidation
It is represented by microcystic spaces surrounded by can be observed in chronic and organizing pneumonia or in
thickened bronchiolar structures due to fibrous evolution, conditions characterized by fibrotic evolution.
and it is the typical pattern of usual interstitial pneumonia
(UIP). The low-attenuation pattern is the expression of fluid- or
air-filled pulmonary lesions, including:
The nodular pattern can be divided into three main classes
on the basis of the distribution of the nodules (Table 14.2), • Cystic lung disease
which includes: Lung cysts are defined as radiolucent areas with
<4 mm wall thickness (>4 mm are called cavities):
• Centrilobular distribution they can be observed in lymphangioleiomyomatosis,
Nodules confined to the centrilobular area are commonly Langerhans cell histiocytosis, and lymphocytic inter-
seen in infectious airway disease, hypersensitivity pneu- stitial pneumonia.
monitis, and respiratory bronchiolitis. They can appear • Emphysema
as ground-glass opacity or show the tree-in-bud sign, It presents as areas of parenchymal destruction with dif-
visualized as an irregular branching structure expression fuse low attenuation. It can be divided into centrilobular
of dilated and mucus-filled centrilobular bronchioles, emphysema, the most common, strongly associated with
usually indicating an infectious process involving the cigarette smoke and with prevalence in upper lobes;
airways. paraseptal emphysema, next to fissures and pleural sur-
• Perilymphatic distribution faces, frequently associated with bullae and spontaneous
Nodules are located in proximity of pleural surface, espe- pneumothorax in young patients; and panlobular emphy-
cially near fissures, and interlobular septa. Perilymphatic sema, affecting the whole secondary lobule, with preva-
nodules are usually seen in sarcoidosis, several pneumo- lence in lower lobes, usually associated with deficiency of
conioses, and lymphangitic spread. alpha-1-antitrypsin.
• Bronchiectasis Table 14.3 Role of NECT and contrast enhancement in the classifica-

They are represented by localized bronchial enlargement, tion of mediastinal masses
with associated bronchial wall thickening and mucous Role of NECT and contrast enhancement in the classification of
mediastinal masses
filling of the bronchial lumen. The most frequent causes
Fluid-containing (cystic) Cystic thymoma
are recurring infections, chronic bronchitis, chronic mediastinal masses Cystic teratoma
obstructive pulmonary disease, and cystic fibrosis. Pericardial cysts
Bronchogenic/esophageal cysts
Lymphangioma
Key Learning Points Cystic degeneration of several tumors
• CT is the most sensitive technique for the assess- Fat-containing mediastinal Lipoma
masses Liposarcoma
ment of pulmonary parenchyma: it has a pivotal
Thymolipoma
role in diagnosis, staging, and follow-up of pulmo-
Benign teratoma
nary cancers. Diaphragmatic hernia
• High-resolution CT can help in the evaluation dif- Calcium-containing Thyroid masses
fuse lung disease, allowing an accurate analysis of mediastinal masses Thymic masses
pulmonary interstitium. Germ cell tumors
• CT can play a role in the evaluation of airways, and Treated lymphomas
it can be used to guide interventional procedures, Lymph nodes (i.e., in tuberculosis,
such as biopsies or drainages. silicosis, sarcoidosis, etc.)
Vascular pseudomasses
• Differential diagnosis and management of the soli-
Enhancing mediastinal Hyperenhancing lymph nodes
tary pulmonary nodule depend on several features, masses Thyroid tissue
such as size, shape, composition, and contrast Paragangliomas
enhancement. Thus, it is crucial to establish imag- Hemangiomas
ing features that help the physicians to distinguish Vascular etiologies
between benign and malignant nodule. The
Fleischner Society periodically updates guidelines
for the management of pulmonary nodules. screening programs); however, they can cause clinical symp-
• CT interpretation of interstitial involvement requires toms (i.e., cough, dyspnea, dysphagia, etc.) requiring detec-
a structured approach considering the dominant pat- tion and characterization. Standard CT protocols include one
tern (reticular, nodular, high attenuation, low attenu- or two phases of contrast-enhanced images, while recent
ation), its location within the secondary lobule (i.e., advances in technology feature low-dose protocols (espe-
centrilobular or perilymphatic), its distribution in cially for lung cancer screening), gated acquisitions for more
the lung (upper vs lower lobe, central vs peripheral reliable vascular evaluation, and dual-energy CT for better
location), and the presence of additional findings tissue characterization. The role of CT densitometry on non-
such as pleural effusion or enlarged lymph nodes. enhanced images in the classification of mediastinal masses
can be summarized in Table 14.3.
The crucial point in differential diagnosis of mediastinal
masses is to distinguish between primary and secondary
14.6 Mediastinum and Pleura tumors, since the latter are more common and they are most
frequently represented by lymphatic masses or metastases.
CT has a pivotal role in the evaluation of the mediastinum, Discrimination of pseudolesions such as congenital vascular
and it is generally used for the evaluation of mediastinal variants may be challenging. In general, the evaluation and dif-
masses and for lymph node staging of lung cancer. The ferential diagnosis of mediastinal masses is based on their loca-
pleura and thoracic wall can be assessed as well, even if the tion (anterior, middle, posterior, and superior compartment),
emerging role of MRI in entities such as malignant pleural but a more accurate classification based on CT axial images has
mesothelioma is slowly changing the diagnostic work-up. been proposed by Fujimoto et al. (JART general rules).
The anterior mediastinum is most frequently involved
(54%), and most of the anterior mediastinal masses are
14.6.1 D
ifferential Diagnosis of Mediastinal included in the “5Ts” (Table 14.4):
Masses
• Thymic masses
Mediastinal masses are relatively rare (about 3% of chest Thymic hyperplasia can be detected in several pathologic
tumors), and they often represent an incidental finding on CT conditions, and it is a frequent occurrence as a “rebound”
performed for different indications (especially lung cancer phenomenon after stress or chemotherapy. Benign and
malignant thymomas represent a neoplastic proliferation in other mediastinal stations. These lesions are usually
of thymic epithelium: there are three main variants: benign isodense compared to soft tissues, and they show only mild
thymoma, invasive thymoma, and thymic carcinoma. In contrast enhancement. Calcifications are frequent only
consideration of the wide individual variability, it is diffi- after treatment. NHL patients often present with advanced
cult to estimate a “normal” size in patients <20 years old. disease: tumor spread can involve extranodal structures and
In grown-up subjects, maximum transverse diameter lung or pleural involvement can be seen.
should not exceed 15 mm. Benign thymomas present as • Extragonadal germ cell tumors
well-defined round masses with soft tissue density located They are frequently located within the thymic loggia and
in the anterosuperior mediastinum. Features indicative of they frequently affect young men. Dermoid cysts are
invasion of adjacent structures should raise the suspect of round capsulated masses with hydric or fatty density that
more aggressive lesions. Thymic cysts frequently develop may contain calcifications of ossification’s foci. Teratomas
following chemo- or radiotherapy, and they can show vari- may show variable density and malignant variants present
able density values (from −15 to 80 HU) due to the com- ill-defined margins and wide necrotic areas; they can also
mon occurrence of intracystic bleeding. dislocate mediastinal vascular structures. Other types of
• Thyroid or parathyroid masses malignant germ cell tumors, such as seminomas, are rare
They are mainly represented by intrathoracic goiter and and usually present as lobulated solid masses.
thyroid tumors. See Head and neck section for a more • Thoracic aneurysm and vascular anomalies
detailed discussion on these topics. Aortic aneurysms are represented by a permanent dilata-
• Lymphomas tion of the aortic wall. Ascending aorta is considered aneu-
They are primary neoplastic lesions of the reticuloendothe- rysmatic when showing a cross-sectional diameter >4 cm
lial system, and they can be classified into Hodgkin disease (while the threshold is >3 cm in descending aorta).
(HD) and non-Hodgkin lymphomas (NHL). CT can be use- Abnormal wall dilatation can lead to rupture. The detection
ful to determine disease extent, to assist therapeutic plan- of aortic wall anomalies including intraluminal calcifica-
ning and to evaluate the response to treatment. HD has an tion can be a sign of acute aortic syndrome, which requires
insidious onset, spreading within contiguous lymph node urgent diagnostic assessment and clinical follow-up.
groups. At CT examination HD may appear as a single
enlarged lymph node or large conglomerated masses. Middle mediastinal masses account for 20% of overall
Predominant thymic involvement is observed in 30% up to mediastinal masses and they include:
70% of patients, usually associated with nodal enlargement
• Lymphadenopathies
Table 14.4 The “5Ts” of anterior mediastinal masses Lymph nodes are the most common site of metastatic
Thymic masses (thymomas, thymolipomas, thymic carcinoma,
spread, and their evaluation has both a prognostic and a
thymic hyperplasia) therapeutic significance (i.e., operable vs non-operable
Thyroid or parathyroid masses (goiter, tumors) disease, adjuvant/neoadjuvant radio-chemotherapy).
(Terrible) Lymphomas (Hodgkin, non-Hodgkin) Imaging findings include size, typically assessed with
Teratoma (especially in children) long- and short-axis diameters (Fig. 14.9), morphology,
Thoracic aneurysm and vascular anomalies homogeneity, conspicuity, and clustering. Regarding size,
Fig. 14.9 Schematic
summary of size and Normal node Reactive node Metastatic node
morphology changes in
reactive lymph nodes, with
increase of long-axis
diameter, and metastatic
lymph nodes, with increase of
short-axis diameter and
rounding of nodal shape
normal mediastinal lymph nodes should not exceed extending caudally along distal thoracic spine but without
10 mm in short transverse diameter, with the exception of bone erosion phenomena. This condition is frequently
subcarinal nodes, considered normal up to 15 mm; how- associated with splenomegaly.
ever, though accepted and routinely used in RECIST cri-
teria, this threshold is debated and far from perfect in
terms of sensitivity and specificity. Other findings that 14.6.2 Pleural and Chest Wall Tumors
should alert the reader when evaluating lymph nodes are
inhomogeneity (often due to necrosis), strong enhance- Chest wall tumors are mainly metastatic, usually from ade-
ment (often associated with hypervascular primary nocarcinomas, or can be caused by the spread of a peripheral
tumors), and changes in size of morphology, such as the bronchogenic carcinoma. They appear as masses located in
presence of spiculated borders. Location (ipsi- vs contra- the muscular layers obliterating fat planes, and they are usu-
lateral) and clustering have to be considered as well in the ally evident only in post-contrast acquisition. The implant
diagnostic evaluation, since clustered nodes are suspi- area and the width of connection angles may indicate the
cious besides their effective enlargement. origin of the lesion (Fig. 14.10).
• Congenital mediastinal cysts Malignant pleural mesothelioma is highly correlated with
They include bronchogenic and enteric cysts, foregut dupli- asbestos exposure, with a long latency period (up to 30 years).
cation cysts, and pericardial cysts; they are often an inci- CT accurately depicts advanced tumors, while is less reliable
dental finding in adult asymptomatic patients. Bronchogenic in early lesions. Mesothelioma may develop as large plaques
cysts are located close to the tracheobronchial tree, and or as irregular nodular thickenings of the pleura, with hetero-
their CT features include the presence of a well-defined geneous enhancement after contrast media. Pleural effusion
non-enhancing oval mass, with a thin wall that can present and pulmonary entrapment may be seen, while some lesions
contrast enhancement. They are characterized by a homo- present with pericardial effusion, erosion of ribs, contralateral
geneous fluid density, and they typically show a paratra- pleural involvement, and intrapulmonary metastasis.
cheal or carinal location. Esophageal duplication cysts are
difficult to distinguish from bronchogenic cysts due to their
similar embryonic origin. They are more frequently located 14.6.3 Nonneoplastic Pleural Diseases
in posterior mediastinum, next to or inside the distal tho-
racic esophagus, although they usually do not communi- Inflammatory and post-traumatic pleural involvement is
cate with the esophageal lumen. Pericardial cysts originate quite frequent in routine clinical practice. The most com-
from the pericardium, but they rarely communicate with mon entity is pleural effusion (Fig. 14.11): it may consist of
the pericardial cavity. They are most frequently located in serous, lymphatic, or hemorrhagic (with higher density val-
the right cardiophrenic angle, and they usually present a ues) components. CT can reveal even small amounts of
water density with well-defined margins and spherical or effusion (>15 mL), and the fluid usually layers in postero-
triangular shape. basal regions, while organized effusions may be found any-
• Vascular anomalies and malformations where in the pleural cavity. Pleural empyema may be the
consequence of a purulent effusion: CT shows marked con-
Differential diagnoses of posterior mediastinal masses
(about 26% of total) include: 1
5
• Neurogenic tumors
They can be divided into nerve sheath tumors, such as 2
schwannomas or neurofibromas, and neuroblastic tumors,
such as malignant neuroblastoma and benign ganglioneu-
roma. The first ones are elliptical paravertebral masses
with mild contrast enhancement that may enlarge inter-
vertebral foramens. Ganglion cell tumors are rare, show
intense contrast enhancement, and are frequently located
next to the aortic arch or posterior pericardium.
• Foregut duplication cysts (see middle mediastinal masses)
• Lymphomas (see anterior mediastinal masses) 3 4
• Lymphadenopathies (see middle mediastinal masses)
• Extramedullary hematopoiesis Fig. 14.10 Schematic summary of pleural, extrapleural, and pulmo-
In patients with thalassemia or sickle cell anemia, hyper- nary lesions and differential diagnosis based on the implant area and
width of connection angles. Pleural lesion (1), pulmonary lesion (2),
plasia of hematopoietic tissue can produce large paraver- pleural effusion (3), extrapleural lesion (4), intrapulmonary lesion with
tebral masses with soft tissue density and convex margins, pleural and chest wall involvement (5)
14.7 Abdomen and Pelvis
CECT is a fundamental tool for the characterization of

lesions initially detected with other diagnostic exams (e.g.,
ultrasound), for diagnosis in emergency situations, and for
staging and follow-up of oncologic patients. CECT allows
for detection of pathological findings that provide a proba-
bilistic diagnosis based on the contrast enhancement pattern
of the lesions in the dynamic phase. The lesions most com-
monly encountered during a body CECT exam are discussed
below.
14.8 Liver
Liver lesions can be categorized into benign and malignant

lesions (primary and secondary). Flow charts to guide in the
Fig. 14.11 An 83-year-old patient with persistent left pleural effusion CT differential diagnosis of liver lesions is provided in
and atelectasis of adjacent lung parenchyma Tables 14.5 and 14.6.
trast enhancement and thickening of pleural sheets sur- 14.8.1 Benign Lesions
rounding the effusion (split pleura sign). Pneumothorax
can be outlined at CT by direct visualization of the pleural Liver cysts are the most common liver lesions, and they can
sheets separated by a gas collection. be solitary or multiple. On NECT, cysts have a watery den-
sity (0–15 HU). Simple cysts do not show contrast enhance-
ment after contact media injection. When multiple cysts with
different size involve the kidney as well, autosomal domi-
Key Learning Points nant polycystic kidney disease (ADPKD) should be consid-
• CT has a pivotal role in the evaluation of the medi- ered in the differential diagnosis.
astinum, and it is generally used for the evaluation Liver hemangiomas are benign vascular malformations,
of mediastinal masses and for lymph node staging and they are the second most common incidental finding that
of lung cancer. can be detected in asymptomatic patients. Considering their
• Mediastinal masses are relatively rare (3% of chest frequent occurrence during radiological examinations (espe-
tumors), and they often represent an incidental find- cially with ultrasound), the differential diagnosis with malig-
ing on CT performed for different reasons; how- nant hepatic lesions is of crucial relevance, especially in
ever, they can cause clinical symptoms (i.e., cough, those patients who have a history of neoplasia. Liver heman-
dyspnea, dysphagia, etc.) requiring diagnostic giomas can be mainly divided into cavernous hemangiomas
work-up and characterization. and in flash-filling hemangiomas that usually cannot be
• The crucial point in differential diagnosis of medi- detected on NECT. Large cavernous hemangiomas typically
astinal masses is to distinguish between primary show early peripheral contrast enhancement in arterial phase
and secondary tumors, since the latter are more with a globular pattern and a progressive centripetal filling in
common and they are most frequently represented portal and equilibrium phases (Fig. 14.12). The globular pat-
by lymphatic masses or metastases. tern implies the presence of a discontinuous and nodular
• A conventional approach to mediastinal mass eval- peripheral contrast enhancement that progressively fills the
uation and differential diagnosis is based on their lesion. Some malignant lesions like intrahepatic cholangio-
location. carcinoma can have peripheral pattern of contrast enhance-
• Chest wall tumors are mainly metastatic, usually ment in arterial phase with progressive fill, but without the
from adenocarcinomas, or can be caused by the nodular or globular pattern.
spread of a peripheral bronchogenic carcinoma. Lesions with complete arterial contrast enhancement and
• Inflammatory and post-traumatic pleural involve- blood pool pattern on portal and equilibrium phase could be
ment, such as in case of pleural effusion or pneumo- characterized as capillary hemangiomas. The blood pool pat-
thorax, is quite frequent in routine clinical tern can be observed when a liver lesion has the same density
practice. of the portal veins on all dynamic phases. When present,
blood pool is an important finding that suggests the diagnosis
Table 14.5 Flow chart of the differential diagnosis of hypervascular liver lesions
Differential diagnosis of Hypervascular liver lesions
Cirrhotic liver Non cirrhotic liver
Benign lesions Malignant lesions
HCC (wash out in portal phase) Cavernous Haemangioma: globular Mts from breast carcinoma neuroendocrine
pattern of c.e. from arterial to portal tumors, melanoma etc.: Hypervascular lesions
enhancement. without blood pool behavior in other phases
Capillary haemangioma: homogenous Fibrolamellar carcinoma: heterogenous arterial

arterial c.e. with blood pool behavior in and portal c.e.
Dysplastic nodule (no demonstrable other phases
wash out in portal phase)
Adenoma: homogenous arterial c.e.
isodense in portal phase to liver
parenchyma (hemorrhage changes can
be present)
FNH homogenous arterial contrast

enhancement isodense in portal phase
(central scar may be observed)
HCC hepatocellular carcinoma, c.e. contrast enhancement, FNH focal nodular hyperplasia, Mts metastases
Table 14.6 Flow chart of the differential diagnosis of hypovascular liver lesions
Differential diagnosis of Hypovascular liver lesions
Unknown primary neoplasm Known primary neoplasm
Cystic appearance Solid appearance
Liver Cyst: density similar to water Liver cystic mts from mucinous colon Liver mts from colon, pancreas, stomach,
and no enhancement is appreciable cancer, ovarian cancer etc. lung, renal, adrenal, etc.
on CECT.
Liver abscess (hypervascular rim can Liver involvement from Hodgkin/Non

be seen) Hodgkin Lymphoma.
Mts metastases
of hemangioma. Capillary hemangiomas differ from transient become isodense compared to the surrounding liver paren-
hepatic attenuating difference (THAD) and from other lesions chyma in the other CT phases.
that exhibit hyper-attenuation only in the arterial phase due to Focal nodular hyperplasia (FNH) is a solitary regenera-
the presence of blood pool behavior. THAD is secondary to tive hepatic lesion characterized by a localized anomaly of
perfusion alterations in the liver parenchyma that may occur the biliary draining system. It typically presents as a mass
isolated or in association with other liver pathologies. These with sharp margins, and it can be supplied by a hepatic artery
pseudolesions are appreciated only on arterial phase, and they with a dedicated hepatic venous drainage system. Focal nod-
a b
c d
Fig. 14.12 Liver hemangioma. NECT (a) shows a hypo-attenuating lesion in the right liver lobe. A globular pattern of contrast enhancement with
centripetal filling can be observed on arterial (b) and portal phase (c). Equilibrium phase (d) shows a complete enhancement of the lesion
ular hyperplasia is the second most common benign solid have been on oral contraceptives for a long time. These
hepatic lesion, and it is found especially in young and lesions are usually solitary and may frequently show hemor-
middle-aged women. The typical behavior of FNH on CECT rhagic phenomena. Hepatic adenomas have a right lobe pre-
is bright homogenous complete arterial contrast enhance- dilection in subcapsular location, and they usually present
ment without washout phenomena (Fig. 14.13). During the with well-defined margins. The attenuation of hepatic adeno-
portal and equilibrium phases, the lesion becomes isodense mas on NECT depends on the presence of fat tissue that
to the liver parenchyma. In 30% of cases, FNH has a central makes the lesion hypo-attenuating. Hemorrhagic phenomena
scar that is hypo-attenuating on NECT and arterial and portal can give the lesion a hyper-attenuating appearance, com-
phase and shows contrast enhancement in the equilibrium pared to liver parenchyma. On CECT, hepatic adenomas
phase. FNH is considered atypical when central scar and/or a present a strong arterial contrast enhancement, and a pseudo-
central artery can’t be observed. When the features of a capsule can be frequently seen. The enhancement in adeno-
hypervascular liver lesion are not completely suggestive of a mas typically does not persist due to the presence of
FNH, especially in woman with a prior history of breast can- arteriovenous shunting.
cer (breast metastasis may also appear hyper-attenuating on Liver abscesses are localized fluid collections containing
arterial phase), further examinations such as magnetic reso- necrotic and purulent material. Liver abscesses may have dif-
nance imaging must be considered. ferent causes: in developing countries, the parasitic cause is the
Hepatic adenomas are uncommon benign liver lesions most common, mainly represented by protozoan and helmin-
usually encountered in young and middle-aged women that thic diseases such as echinococcosis and amebiasis, while in
a b
c d
Fig. 14.13 Liver FNH. Arterial phase (a) shows a round lesion with the administrarion of hepatospecific contrast material (d) showing
rapid contrast enhancement, and a central scar can be observed. The hyperintensity of the lesion compared to liver parenchyma that is highly
lesion is hyper-attenuating compared to the liver parenchyma on portal suggestive of FNH
(b) and equilibrium phase (c). MR image acquired after 120 min after
western countries, the most frequent etiology is bacterial, and it 14.8.2 Malignant Lesions
occurs as comorbidity in immunocompromised patients, in
abdominal sepsis, and in necrotizing enterocolitis. Amebic • Primary liver lesions
abscess is usually solitary, whereas fungal and bacterial liver
abscesses are multiple. Liver abscesses do not have a typical Hepatocellular carcinoma (HCC) is the most common
pattern on NECT and CECT. Lesions are usually round with primary hepatic tumor, usually associated with hepatic cir-
central hypo-attenuation with respect to the liver parenchyma rhosis. The incidence of this lesion is strictly correlated to
in portal phase. A slightly hyper-attenuating peripheral rim can chronic HBV or HCV infection, especially in developing
be observed on NECT, showing contrast enhancement in arte- countries, and to alcohol abuse that lead to hepatic cirrhosis
rial and portal phase. Sometimes a central hypo-attenuating in western countries. Biliary atresia, biliary cirrhosis, food
area can be observed (fluid collection due to necrotic changes), toxins (such as aflatoxin), and inborn metabolism failures
surrounded by an inner rim of hyper-attenuation that shows (hemochromatosis, Wilson disease, and alpha-1-antitrypsin
strong contrast enhancement on the arterial phase and an outer deficiency) are also important risk factors. HCC usually
rim (due to liver parenchymal edema) that enhances on portal develops in a cirrhotic liver and can present as a solitary,
and equilibrium phases (the so-called double target sign). large mass with well-defined margins or multiple, with
a b
c d
Fig. 14.14 Hepatocellular carcinoma (HCC). NECT phase (a): right liver lobe hypo-attenuating lesion. On arterial phase (b), the lesion shows a
strong contrast enhancement that fills the lesion. On portal (c) and equilibrium phase (d), a washout phenomenon is demonstrable
variable attenuation. Another possible presentation is the temic venous shunts, THAD, focal fatty changes, or skip
infiltrative or diffuse type. In the latter case, the presence of areas in fatty liver may be observed.
nodular alterations of the parenchyma as regenerative nod- Dysplastic nodules enter in the differential diagnosis with
ules or dysplastic nodules may make the diagnosis of HCC HCC nodules. On CECT they often exhibit early contrast
more difficult. The typical HCC pattern of contrast enhance- enhancement on the arterial phase, but, unlike HCC nodules,
ment consists in a hyper-attenuation in arterial phase due to they do not show washout phenomena.
branch of hepatic artery vascularization supply. The pathog- Regenerative nodules are liver lesions observed in liver
nomonic sign of HCC is the early contrast enhancement cirrhosis and are classified according to their size as
with washout phenomena in portal and equilibrium phases micronodules, macronodules, and giant nodules. Nodules
(Fig. 14.14). When an early hyper-attenuation on arterial are characterized by the presence of a regular parenchymal
phase is seen into a regenerative nodule, a small HCC with architecture in which a certain amount of fibrotic tissue or
“nodule in nodule” appearance must be suspected. Nodular iron deposits may be included. These findings are responsi-
HCC presents in more than half of the cases (50–80%) with ble for the modest hyper-attenuation on NECT. There is no
a peripheral pseudocapsule. Furthermore, HCCs may infil- difference in attenuation of regenerative nodules with respect
trate the portal vein causing thrombosis and portal hyper- to liver parenchyma on CECT study.
tension. In this case a cavernous transformation of the Fibrolamellar carcinoma is a variant of HCC that occurs
portal vein can be observed (tortuous venous channel that in younger patients with no correlation with common HCC
replace the portal vein). Other liver lesions like portosys- risk factors. This type of tumor is a well-circumscribed
a b c
Fig. 14.15 CECT on portal phase (a) shows many round hypovascular history of mucinous colonic tumor requires further investigation with
lesions that are suggestive for metastases (the patient has a prior history MRI. The cystic nature of the lesion is confirmed by the T2 fat satura-
of colonic neoplasm). A small hypovascular lesion on VI liver segment tion MRI (c), and the thin septa can be observed on 3D T1 fat saturation
can be identified (b). Thin intra-lesion septa are present. The patient’s MRI (d)
lesion with sharp margins, and it contains a portion of fibrotic

tissue organized in lamellar bands usually forming a central Key Learning Points
scar. On NECT a central area of low density due to necrosis • Liver lesions can be categorized in benign and
or fibrosis can be seen. Calcifications can be observed in cen- malignant lesions, and malignant lesions can be
tral scar in up to 30% of all cases. Fibrolamellar carcinomas divided into primary and secondary.
commonly are heterogeneously hypervascular lesions in • Liver cysts are the most common liver lesions and
arterial and portal phases. they can be solitary or multiple.
• Liver hemangiomas are benign vascular malforma-
• Secondary liver lesions tions, and they are the second most common inci-
dental finding that can be detected in asymptomatic
Liver metastases are the most common malignant liver patients.
lesions in an non-cirrhotic liver, and in most cases they are • Focal nodular hyperplasia and hepatic adenomas
not appreciable on NECT. Metastases on CECT study can are uncommon benign liver lesions usually encoun-
be hypervascular (most commonly from renal cell carci- tered in young and middle-aged women.
noma, breast carcinoma, islet cell tumor, melanoma and sar- • Hepatocellular carcinoma (HCC) is the most com-
coma, pheochromocytoma, carcinoid and thyroid carcinoma) mon primary hepatic tumor, usually associated
or hypovascular (most commonly from the GI tract, breast, with hepatic cirrhosis. The typical HCC pattern of
pancreas, lung, and cervix) (Fig. 14.15). In the detection of contrast enhancement consists in a hyper-attenua-
small liver metastases, arterial phase is useful in both types tion in arterial phase due to branch of hepatic
of lesions: in hypovascular metastases, it can accurately artery vascularization supply. The pathognomonic
demonstrate the peripheral rim, and in hypervascular metas- sign of HCC is the early contrast enhancement
tases, it shows heterogeneous hyper-attenuation (this type of with washout phenomena in portal and equilib-
metastasis are likely isodense to liver parenchyma in rium phases.
the portal phase). The portal and equilibrium phases are the • Liver metastases are the most common malignant
most important phases in detection of hypovascular metas- lesions in non-cirrhotic liver; on CECT study they
tases that typically show hypo-attenuation compared to nor- can present hypervascular (most commonly from
mal liver parenchyma. NECT is helpful to distinguish renal cell carcinoma, breast carcinoma, islet cell
calcified metastases from the colon mucinous tumor, stom- tumor, melanoma and sarcoma, pheochromocy-
ach, adnexa, breast, thyroid, lung or kidney, carcinoid, and toma, carcinoid, and thyroid carcinoma) or hypo-
melanoma. When the patient has concomitant fatty liver dis- vascular behavior (most commonly from the GI
ease, liver metastases may appear iso- or slight hyper-atten- tract, breast, pancreas, lung, and cervix).
uating with respect to the liver parenchyma in portal phase.
14.9 Biliary Tree with common bile duct and intrahepatic bile ducts dilatation.
A mass-forming cholangiocarcinoma on CECT typically
Biliary lesions affect the gallbladder and both intrahepatic shows early complete rim enhancement with progressive,
and extrahepatic bile ducts. They can be benign or centripetal non-globular pattern of enhancement in portal
malignant. and equilibrium phase (Fig. 14.16). Invasion of portal veins
Gallbladder carcinoma is usually asymptomatic at the is not a rare occurrence, especially in hilar type. Infiltrative
time of the diagnosis, and it is often an incidental finding periductal cholangiocarcinoma is most common at hepatic
during an abdominal ultrasound or a CT exam performed for hilum, and it is often associated with narrowing or dilatation
other reasons. Long-standing gall bladder stones are the of peripheral bile ducts due to pathological thickening of
major risk factor for the development of this condition. periductal hepatic parenchyma. The intraductal infiltrative
Others risk factors include chronic cholecystitis, familial cholangiocarcinoma may show the presence of polypoid
adenomatous polyposis syndrome, gall bladder polyps (when lesion into a duct that demonstrates contrast enhancement on
>10 mm), porcelain gallbladder, and inflammatory bowel portal phase compared to NECT.
diseases. Gallbladder carcinoma is most frequent in elderly
women. Clinical presentation can be different depending on
the location of the tumor: if the lesion affects the fundus of
the gallbladder, the tumor can infiltrate the adjacent liver Key Learning Points
parenchyma or the colon, and the first symptom could be the • Biliary lesions affect the gallbladder and both intra-
upper abdominal quadrant pain; if the lesion affects the gall- hepatic and extrahepatic bile ducts. They can be
bladder infundibulum, the initial symptom can be jaundice. benign or malignant.
On NECT, calcifications of gallbladder’s walls could be • Cholangiocarcinoma is the second most frequent
present; when the wall is completely calcified and retracted, malignant hepatic tumor, and can arise from intra-
the gallbladder is called porcelain gallbladder. On CECT, hepatic or extrahepatic bile duct epithelium.
gallbladder carcinoma can be seen as a solid polypoid mass • Cholangiocarcinomas may be classified depending
with a diameter >10 mm, with diffuse wall thickening show- on the shape and biological behavior. It may be
ing hyper-attenuating pattern of contrast enhancement, or as mass forming, infiltrating, periductal, or
solid tissue that completely replaces the gallbladder. The lat- intraductal.
ter is the most frequent presentation: it appears as an infiltrat-
ing heterogeneous mass with slightly and patchy contrast
enhancement in arterial and portal phases. Dilatation of
intrahepatic biliary ducts, lymphadenopathies, and ascites 14.10 Pancreas
may be present as features of advanced disease.
Cholangiocarcinoma is the second most frequent malig- Lesions in the pancreas can be divided into solid and cystic
nant hepatic tumor and it originates from cholangiocytes. as summarized in Table 14.7).
The main risk factors for the development of cholangiocarci-
noma are primary sclerosing cholangitis (PSC, especially in
western countries), recurrent pyogenic cholangitis, choledo- 14.10.1 Pancreatic Cystic Tumors
cholithiasis, and Clonorchis sinensis infection in Asian
countries. Cholangiocarcinomas can arise from intrahepatic Pancreatic cystic tumors show fluid density on NECT. The
bile ducts or extrahepatic bile duct epithelium. The intrahe- lesions can be divided into intraductal papillary mucinous
patic form is divided into a peripheral type that is usually a neoplasm (IPMN), mucinous pancreatic cystadenomas, and
hypodense mass on NECT with associated intrahepatic bile pancreatic serous cystadenoma.
duct dilatation (capsular retraction is often seen) and a cen- There are three types of IPMN: secondary branch type,
tral (hilar) type that is a mass at the level of hepatic duct main duct type, and mixed type. The secondary branch type
confluence. Cholangiocarcinoma is considered peripheral if with a multicystic appearance is most likely due to a dilata-
it originates peripherally from a secondary branch of a bili- tion of secondary pancreatic ducts (usually at the level of
ary hepatic duct and hilar when it originates from a biliary pancreatic head). MR cholangiography is able to demon-
hepatic duct or common hepatic duct (Klatskin tumor). strate a communication between the multicystic lesions and
Cholangiocarcinomas may be classified also depending on the main pancreatic duct. The isolated main duct type of
the shape and biological behavior. It may be mass forming, IPMN consists of a dilatation of the main pancreatic duct of
infiltrating, periductal, or intraductal. The extrahepatic type <5 mm. Mucous material with fluid-density values may fill
of cholangiocarcinoma is typically infiltrative in association the duct even at the level of the Vater papilla, which bulges
a b
c d
Fig. 14.16 Intrahepatic liver CCC. NECT image (a): the biliary drain- The lesion is filled by contrast without a globular pattern on portal
age system of the left liver lobe (II segment mainly) and of IVa segment phase (c). MRI cholangiography (d) easily demonstrates the biliary
is dilated (white arrowhead). CECT on arterial phase (b) shows a obstruction at the level of left biliary duct bifurcation (white arrow)
peripheral contrast enhancement that fills the lesion (black arrowhead).
Table 14.7 Flow chart of the

differential diagnosis of pancreatic Differential diagnosis of pancreatic lesions
lesions
Cystic lesion Solid lesion
Hypervascular Hypovascular
PSC: multiple small cyst in Neuroendocrine tumor: body Pancreatic adenocarcinoma:

head of pancreas or tail with arterial contrast more likely in pancreatic
enhancement head. Double duct sign
MPC: Large unilocular or Pancreatic metastasis from

multilocular in body of GI cancer, etc.
pancreas
IPMN primary duct, secondary

branch or mixed. When
papillary projection present
high suspicious of malignancy
PSC pancreatic serous cystadenoma, MCP mucinous pancreatic cystadenoma, IPMN intraductal pap-
illary mucinous neoplasm, GI gastrointestinal
a b
Fig. 14.17 Pancreatic adenocarcinoma. CECT on arterial phase (a): a hypovascular lesion is demonstrable on the head of the pancreas (arrow-
head). The hypo-attenuating lesion persists on portal phase (b)
into the duodenum. The surrounding pancreatic paren- lesion with heterogeneous poor contrast enhancement in par-
chyma may be thinned due to atrophy. In the combined ticular in arterial phase on CECT, probably due to the pres-
type, both primary ducts and the main duct are enlarged. ence of fibrous tissue. A desmoplastic reaction can be usually
Any papillary projection into a pancreatic duct with con- observed around the lesion (Fig. 14.17). When the head of
trast enhancement on CECT is suspect for malignant the pancreas is involved, a “double duct sign” (dilatation of
transformation. main pancreatic duct and common biliary duct) can be
Pancreatic serous cystadenoma has a strong female predi- observed, and particular attention should be paid to the pres-
lection in elderly age. On CECT multiple (>6) small cysts ence of eventual vascular infiltration.
(<2 cm) with thin wall are observed, usually in the pancre- Endocrine tumors of the pancreas or pancreatic islet cell
atic head. Calcification with central scar can be seen. tumors are usually solitary, but they can be associated with
Pancreatic serous cystadenoma is considered a benign MEN I, von Hippel-Lindau, or tuberous sclerosis. Pancreatic
lesion. islet cell tumors can be classified in functional and nonfunc-
Mucinous pancreatic cystadenoma has a strong predilec- tional or syndromic and non-syndromic, depending on
tion for middle-aged women: the typical presentation is a whether or not they produce hormones. Another classifica-
large unilocular or multilocular cystic lesion (2–12 cm) most tion is based on the type of cells that constitute the tumor:
frequently observed in the body or tail of the pancreas. In glucagonoma, insulinoma, somatostatinoma, gastrinoma,
this type of lesion, a communication with the main pancre- and VIPoma. Pancreatic islet cell tumors can appear small or
atic duct should not be visible at MR cholangiography. large in size. When small in size, they usually show sharp
Mucinous pancreatic cystadenoma is considered a premalig- margins, sometimes with calcifications and homogenous
nant lesion, and any papillary projection with contrast contrast enhancement in arterial phase (Fig. 14.18); when
enhancement is suspicious for malignant transformation larger in size, the lesion can present central necrotic changes
(mucinous pancreatic cystadenocarcinoma). with fluid density. The margins are well circumscribed due to
the presence of a capsule, and a peripheral rim of c.e. in arte-
rial phase may be observed. This type of contrast enhance-
14.10.2 Pancreas Solid Neoplasms ment allows differential diagnosis with cystic lesions of the
pancreas. Typically, these tumors have a non-infiltrative
Pancreatic ductal adenocarcinoma is a malignant neoplasm growth pattern, and they cause displacement of adjacent
commonly affecting the elderly. The typical locations of pan- structures. This type of growth allows differentiation from
creatic ductal adenocarcinoma are the head of the pancreas pancreatic ductal adenocarcinoma that has an infiltrative
or the uncinate process. Pancreatic ductal adenocarcinoma behavior. The arterial phase on CECT is mandatory to detect
commonly appears like an isodense mass on NECT and as a pancreatic islet cell tumors because in portal phase, they can
a b
Fig. 14.18 Pancreatic carcinoid neoplasm. CECT on arterial phase (a): a hypervascular pancreatic lesion is seen (arrowhead). The lesion is not
easily appreciable on portal phase (b)
be isointense compared to the rest of pancreatic parenchyma Splenic infarction (usually triangular in shape), splenic cyst
and small lesions may be missed. and splenic lymphangioma (with absent c.e. on CECT), and
splenic abscess (low attenuation or absent c.e. lesions that may
demonstrate peripheral rim) are the most common benign
splenic lesions with near to fluid attenuation. Spleen hamar-
Key Learning Points toma and spleen hemangioma are benign solid tumors. Spleen
• Pancreatic lesions can be divided into solid and cys- hemangioma has a contrast enhancement on CECT similar to
tic lesions. that of liver hemangioma. The most common primary malig-
• Regarding pancreatic cystic tumors, they can be nant tumors are lymphomas, which usually occur as hypovas-
divided into intraductal papillary mucinous neo- cular mass lesions in the portal phase, and angiosarcoma.
plasms (IPMN), mucinous pancreatic cystadeno- Spleen metastases usually appear as multiple solid lesions
mas, and pancreatic serous cystoadenomas. with hypovascular appearance in the portal phase. Splenic
• Pancreatic ductal adenocarcinoma is a malignant cystic metastases can be observed as secondary involvement
neoplasm that commonly occurs in the elderly. The from ovary, breast, or mucinous GI tumors.
typical locations of pancreatic ductal adenocarci-
noma are the pancreatic head or the uncinate pro-
cess. Pancreatic ductal adenocarcinoma commonly
Key Learning Points
appears like an isodense mass on NECT, with het-
• Spleen lesions can be categorized as benign (usu-
erogeneous poor contrast enhancement in particular
ally with fluid attenuation pattern) and malignant
in arterial phase on CECT.
(most commonly solid).
• Endocrine tumors of the pancreas or pancreatic islet
• Splenomegaly is defined when the spleen is longer
cell tumors are usually solitary, but they can be
than 13 cm: it may result from hematologic condi-
associated with MEN I, von Hippel-Lindau, or
tions, lymphoproliferative disorder, infiltrative
tuberous sclerosis. They can be classified as syn-
pathologies, vascular congestion, or infectious-
dromic and non-syndromic, depending on whether
inflammatory causes.
or not they produce hormones.
14.11 Spleen 14.12 Adrenal Glands
Lesions in the spleen can be categorized as benign (usually Adrenal lesions are divided into benign and malignant. These
with fluid attenuation pattern) and malignant (most com- lesions are recognized more often due to the increase of CT
monly solid). Splenomegaly is defined when the spleen is examinations. Adrenal lesions are often an incidental finding
>13 cm in length: it may result from hematologic and lym- during a CT examination performed for other reasons. The
phoproliferative disorders, infiltrative pathologies, vascular radiologist plays a crucial role in the differential diagnosis of
congestion, or infectious-inflammatory conditions. these lesions (Table 14.8).
Table 14.8 Flow chart of the dif-

ferential diagnosis of adrenal lesions Differential diagnosis of adrenal lesions
< 10 HU on NECT > 10 HU on NECT. Need portal

phase and delayed phase 15 min
on CECT
>60% AW <60% AW
Adrenal lipid rich adenoma. Adrenal adenoma lipid poor Further diagnostic
No further diagnostic examinations are mandatory
exams (MRI in primis)
NECT non-contrast-enhanced CT, AW absolute washout
Adrenal adenomas are the most common incidental adre-

nal lesions. As a consequence, adrenal adenomas enter in dif-
ferential diagnosis with adrenal metastases, particularly
when a history of malignancy is present. Adrenal adenomas
can be classified as functioning and non-functioning. Non-
functioning adenomas constitute the majority of adrenal
adenomas. Another possible classification divides them into
typical and atypical adenomas: typical adenoma is <3 cm in
diameter with a high content of fat tissue. The atypical ade-
noma is defined by the presence of hemorrhage, lack of fat
tissue, necrosis, calcification, and larger size. When an adre-
nal lesion is larger than 4 cm, it has a high probability to be
malignant. Typical adrenal adenomas are well-circumscribed
lesions with a density inferior to 10 HU on NECT due to
high lipid content (“lipid-rich” adenoma) (Fig. 14.19). When
an adrenal lesion has a density >10 HU on NECT, contrast
media administration is mandatory to differentiate them Fig. 14.19 Adrenal adenoma. NECT showing left adrenal gland with
a round lesion demonstrating a density of 8 HU. The lesion’s density is
from metastases or other solid lesions. Adrenal washout is less than 10 HU. No further diagnostic work-up is needed, since the
diagnostic for “lipid-poor” adenoma. lesion can be classified as a lipid-rich adenoma
NECT ( HU ) − delayed CECT after 15 min ( HU ) / CECT ( HU ) in portal phase × 100
( The result is the relative washout )

CECT ( HU ) in portal phase − delayed CECT after 15 min ( HU ) / CECT ( HU ) in portal phase − NECT ( HU ) × 100
( The result is the absolute washout )

Relative washout > 40%and absolute washout ( AW ) > 60% are diagnostic for adenoma.

Adrenal carcinoma is rare and can be functioning or non- ferentiate if the lesion originates from the liver or from the
functioning depending on hormone secretion. When adrenal adrenal, especially if large in size.
carcinomas are non-functioning, they can appear as large Adrenal metastases are the most common adrenal malig-
lesions. On CECT, the adrenal carcinoma appears as a solid nant lesions and can be bilateral. The most frequent tumors
mass with possible areas of necrosis (cystic appearance), that metastasise to adrenal glands are lung cancer, colorectal
hemorrhage, and calcifications. These lesions tend to infil- cancer, breast cancer, and pancreatic cancer. HCC and renal
trate the liver, the inferior vena cava, and the kidney, based cancer may involve adrenal glands as well. It may be difficult
on their localization. In some cases, it is not possible to dif- to distinguish metastasis from “lipid-poor” adrenal adeno-
mas on CECT if a NECT and delayed phase are not acquired. form a differential diagnosis for cystic renal findings
Adrenal metastases are often round in shape, uni- or bilat- (Table 14.9).
eral, with NECT value > of 10 UH and prolonged washout
pattern (absolute washout value <60%).
14.13.1 Renal Cystic Lesions
Key Learning Points

Renal cysts are the most common benign findings in the
• Adrenal lesions are often an incidental finding dur- kidneys. Renal cysts can be divided into cortical and sinus
ing a CT examination performed for other reasons. cysts. Renal sinus cysts are benign. They can be subdi-
• Adrenal adenomas can be classified as functioning vided into parapelvic cysts that originate from the adja-
and non-functioning. Non-functioning adenomas cent parenchyma and protrude into the renal sinus and
constitute the majority of adrenal adenomas. peripelvic cyst that originates within the sinus itself. If a
Another possible classification divides them into renal cyst demonstrates no contrast enhancement or a dif-
typical and atypical adenomas: typical adenoma is ference in attenuation <10 UH from NECT to CECT (on
<3 cm in diameter with a high content of fat tissue. portal phase), it is considered as non-enhancing simple
• When an adrenal lesion has a density >10 HU on cysts. If there is an increase in density from 10 to 20 HU,
NECT, contrast media injection is mandatory to dif- it is indeterminate, and pseudoenhancement artifact
ferentiate them from metastases or other solid lesions. should be considered. A difference in density >20 HU is a
Adrenal washout is diagnostic for “lipid- poor” major criterion to classify it as a complex cyst. Bosniak
adenoma. classification is helpful in predicting the risk of malig-
• Adrenal metastases are the most common adrenal nancy of a complex renal cyst. Simple cysts are well-
malignant lesions and can be bilateral. The most defined, with thin or imperceptible wall, and they show
frequent tumors that metastasise to adrenal glands water attenuation (<20 HU) on NECT. These cysts are
are lung cancer, colorectal cancer, breast cancer, classified as Bosniak I. When a cyst presents with mini-
and pancreatic cancer. mally increased number of septa, but nonmeasurable c.e.
of a hairline-thin septa, they are classified as Bosniak
II. Hyper-attenuating cysts on NECT are well-capsulated
cysts with thin or imperceptible hyper-attenuating (60–90
14.13 Kidney HU) walls and are classified as Bosniak II F. Strict follow-
up is mandatory in II F cysts. Indeterminate cysts show
Renal lesions can be divided into cystic and solid lesions. thick margins, nodular multiple septa, or wall with mea-
A diagnostic road map based on morphology, HU, and surable enhancement, hyperdense on CT (Bosniak III).
contrast enhancement behavior will be provided to per- These lesions have a high probability to be malignant
Table 14.9 Flow chart of the differential diagnosis of cystic renal lesions
Differential diagnosis cystic renal lesions
Cystic lesion < 20 HU on NECT Cystic lesion > 20 HU on NECT

or c.e < 15HU on CECT portal
phase. Simple cyst.
No further examaminations.
50-60 HU on NECT and c.e < 15HU > 20HU on NECT and c.e > 15HU
on CECT portal phase on CECT portal phase
Single, no septa or hair line Hemorrhagic cyst. Bosniak IIF Solid component of the cyst are
septa are simple cyst (Bosniak present. Bosniak III or IV. Further
I or II). No further examinations are required
examinations
C.e. contrast enhancement

(50%). Malignant renal cysts are solid masses with a large cell carcinoma and metastasis. Multiple renal angiomyolipo-
cystic or a necrotic component (Bosniak IV) (Fig. 14.20). mas can be observed in tuberous sclerosis.
Renal cell carcinoma: it is the most common renal tumor.
It is derived from tubular epithelium. Renal cell carcinoma
14.13.2 Renal Non-Cystic Lesions most frequently occurs in middle-aged and elderly patients.
Clinically it may manifest with flank pain or hematuria.
Renal lesions that do not exhibit a cystic density such as Renal cell carcinomas, mostly when asymptomatic, are usu-
angiomyolipoma with negative densitometric values on ally found as small incidental lesions. The recognized risk
NECT and solid lesions that frequently exhibit high UH val- factors for kidney cancer are cigarette smoking, obesity, and
ues on NECT and contrast enhancement on CECT are also dialysis-related cystic disease. Renal cell carcinoma shows
considered as solid findings. The most common renal non- soft tissue attenuation on NECT. Larger lesions can fre-
cystic lesions are renal angiomyolipoma, renal cell carci- quently have areas of necrosis or fat, and some calcification
noma, and transitional cell carcinoma. can be present. On CECT they demonstrate variable enhance-
Renal angiomyolipoma: most lesions involve the cortex ment, usually less than the normal cortex. Large lesions have
and demonstrate macroscopic fat content (less than −20 irregular enhancement due to areas of necrosis (Fig. 14.21).
HU). When small in size, it can be difficult to differentiate When the renal cell carcinoma is well capsulated and a pseu-
them from a small simple cyst. Fat-containing lesions are not docapsule is present, the differential diagnosis with renal
pathognomonic of renal angiomyolipoma because lipid oncocytoma cannot be made. The clear cell subtype may
necrosis or osseous metaplasia can be present also in renal show much stronger enhancement.
a b
c d
Fig. 14.20 Bosniak IV cystic lesion. NECT (a) showing a lower pole mural nodular lesions with strong contrast enhancement. Wall thicken-
cystic lesion of the left kidney. The lesion’s density is 18 HU, similar to ing is better appreciable on delayed phase (d)
fluid. CECT on arterial phase (b) and portal phase (c) showing small
a b
c d
Fig. 14.21 RCC. NECT (a) shows a round lesion at the lower pole of rim of contrast enhancement on arterial phase (b). The delineation of
the right kidney. The lesion has a large central necrotic component that viable tumor and necrosis is better seen on portal phase (c) and delayed
is hypo-attenuating compared to the renal parenchyma with a peripheral phase (d)
Transitional cell carcinoma: it affects renal excretory and 10 min in which the contrast medium is excreted by the
pathways and may occur at the level of the calyx, pelvis, ure- renal excretory system and fills the excretory tracts allowing
ter, and bladder. Transitional cell carcinomas usually appear the visualization of any filling defects. If a filling defect is
as papillary lesions that partly or completely occupy the present, it is necessary to evaluate whether this finding dem-
excretory pathway. Another morphologic pattern of presenta- onstrates contrast enhancement between the NECT phase and
tion is non-papillary type with nodular or sessile tumors. the arterial and portal phases. To increase sensitivity to detect
When transitional cell carcinoma grows into the renal pelvis, transitional cell lesions, it is necessary to use a proper win-
it can completely occupy the lumen and obliterate the renal dowing by reducing the contrast and lowering the grayscale
sinus fat. Clinically, it presents frequently with micro- or once delayed scans are obtained.
macro-hematuria and pain at the side that mimics a renal colic If the lesion to be studied is in the bladder, it is necessary
when hydronephrosis is manifested as a result of obstruction. to wait longer than 10 min after contrast administration to
CT plays a fundamental role in the differential diagnosis acquire the delayed phases. In this way we try to fill the blad-
between urinary stones and a neoplastic lesion. The diagnosis der as much as possible with the contrast material especially
of obstructive renal calculus is already possible on the NECT for suspected anterior wall lesions. This is because the con-
phase where the presence of a hyper-attenuating calculus trast material accumulates from the posterior bladder wall to
in the urinary tracts can be observed. In addition to the usual the anterior wall due to gravity (Fig. 14.22). An anterior wall
image CT protocol, it is crucial to perform scans between 5 lesion can be studied also by positioning the patient prone on
a b
c d
Fig. 14.22 Bladder polyps. Wall bladder thickness is measurable on cannot be clearly visualized on delayed phase (c) due to their anterior
NECT (a). On CECT portal phase, small pedunculated lesions on wall localization. CECT on delayed phase in prone position (d): wall
the bladder anterior wall are demonstrable (b). Mural nodular lesions nodular lesions appear as endoluminal filling defects
the bed during the delayed phase. When performing a study

for the evaluation of a bladder lesion (observed by ultrasound • Renal cell carcinoma is the most common renal
or with cystoscopy), it is necessary to pay great attention to tumor deriving from tubular epithelium. It most fre-
the entire urinary tract because transitional lesions are often quently affects middle-aged or advance-aged
multicentric (Table 14.9). patients. It shows soft tissue attenuation on
NECT. On CECT they demonstrate variable
enhancement, usually less than the normal cortex.
Large lesions have irregular enhancement due to
Key Learning Points areas of necrosis.
• Renal lesions can be divided into cystic and non- • Transitional cell carcinoma may usually appear as
cystic (solid) lesions. papillary lesions that partly or completely occupy
• Renal cysts are the most common benign finding in the excretory tract lumen. CT plays a fundamental
the kidneys. They can be divided into cortical and role in the differential diagnosis between urinary
sinus cysts. stones and neoplastic lesions.
14.14 Small Bowel Disease intestinal stromal tumors (GIST) of the small bowel tend to
be more aggressive than gastric cancers. GISTs tend to be
The duodenum and small bowel can be studied on CECT round in shape and show heterogeneous contrast enhance-
using arterial and portal phase scans even after the oral ment on CECT despite a central hypodense area due to
administration of neutral contrast solution (used to distend necrosis. They are often exophytic even if intramural vari-
the bowel lumen) that allows visualization of any defects of ants (or even polypoid variants) have been described
the intestinal mucosa. Multiplanar imaging also improved (Fig. 14.23).
intestinal evaluation especially for the diagnosis of suspected Carcinoid tumors or neuroendocrine tumors are the
obstruction. The most common pathologies of the small most common small bowel malignant neoplasm. They are
bowel including congenital anomalies, neoplasms, and derived from neuroendocrine cells in the submucosal
inflammatory diseases will be discussed below. wall, and are usually small in size and may not be local-
The most frequently encountered small bowel congenital ized during a CECT examination. When visible, neuroen-
anomalies on an abdominal CT are duodenal diverticula, docrine tumors show a hypervascular behavior. These
especially in the second and third portion of the duodenum, lesions often affect mesenteric lymph nodes and show a
and Meckel’s diverticulum. desmoplastic reaction that retracts the mesentery and the
Duodenal diverticula typically extend toward the pan- small bowel loops.
creas, and their content are characterized by the presence of Small bowel adenocarcinoma is rare, most likely occur-
air bubbles and food residues. Duodenal diverticula may ing in the duodenum. The typical features on CECT are cir-
resemble cystic pancreatic lesions when air bubbles are cumferential wall thickening that cause duodenum
absent, and the differential diagnosis between these two obstruction. When the small bowel adenocarcinoma involves
entities can be difficult: in these cases the oral positive con- the entire thickness of the bowel wall, mesenteric fat strand-
trast media may be helpful because if the contrast medium is ing can be observed. When a small bowel intussusception is
able to fill the lesion, the diagnosis of duodenal diverticulum present, an intestinal neoplasm should always be considered
can be made. Meckel’s diverticulum is the most common as the cause.
congenital abnormality of the gastrointestinal tract. The most Risk factors for small bowel lymphoma include Crohn's
frequent location of Meckel’s diverticulum is 30 cm proxi- and celiac disease. Lymphoma most frequently affects the
mal to the ileocecal valve, but it can rarely be seen on CT distal region of the ileum and appears as a circumferential
examination due to the low difference in contrast compared parietal thickening with intestinal pseudoaneurysm dilata-
to the small bowel. It can only be visualized when there is tion. Mesenteric lymphadenopathy and splenomegaly are
inflammation of the Meckel’s diverticulum. also associated.
Small bowel neoplasms are relatively uncommon, and CT Even some inflammatory diseases such as Crohn’s dis-
remains the modality of choice to identify them. The gastro- ease may affect the small bowel and in particular the distal
a b
Fig. 14.23 GIST. CECT on portal phase (a) exhibits a round lesion plasm that originates from the jejunum (arrowhead in b); a central area
arising from the wall of small bowel (arrowhead). The lesion shows a of necrosis can be observed
peripheral rim of contrast enhancement. Another example of GIST neo-
tract of the ileum. Although it is preferable to perform CECT Jeghers syndrome, and hereditary nonpolyposis colon cancer
examination with fluid distension of the intestinal loops syndrome (HNPCC) may lead to colorectal carcinoma
using an oral contrast medium, it is also possible to observe development.
the most common radiological signs of this pathology on Colorectal cancer may appear as a polypoid adenomatous
CECT. In addition to the concentric thickening of the intesti- lesion that progressively develops genetic mutations that
nal wall, CECT may demonstrate the fat stranding around lead to neoplastic transformation. Furthermore, colorectal
the pathological mesenteric bowel loop, the “comb sing” due cancer may appear as a protruded lesion (pedunculated, sub-
to increase of mesenteric vessels, and the skip lesions. pedunculated, and sessile) or as a flat lesion that can be
observed only on virtual or optical colonoscopy. When
colorectal cancer is found in advanced stage of illness, it
often shows circumferential wall involvement. Ulcerated
Key Learning Points lesions may also be observed.
• The duodenum and small bowel can be studied on The clinical presentation of this disease can be subtle
CECT using arterial and portal phase scans even because if there is no rectal bleeding, colorectal cancer may
after oral administration of neutral contrast solution not be found until it has reached a large size and intestinal
(used to perform bowel lumen distension) to allow obstruction or secondary involvement of other organs may
visualization of any defects of the intestinal mucosa. be the first manifestations of the disease.
• The most frequently encountered small bowel con-
genital anomalies on an abdominal CT are duodenal
diverticula, especially in the second and third por- 14.15.1 Morphologic Patterns of Presentation
tion of the duodenum, and Meckel's diverticulum.
• Small bowel neoplasms are relatively uncommon, Colorectal cancer may appear as a mass that narrows the
and CT remains the modality of choice to identify bowel lumen with soft tissue density on NECT and strong
them. contrast enhancement on CECT. Ulcerations and shoulder-
• Carcinoid tumors or neuroendocrine tumors are the ing of the lesions margins may be observed in a large mass.
most common small bowel malignant neoplasm. Mucinous type tumors more frequently appear more
They derive from neuroendocrine cells in the sub- hypodense due to mucus accumulations. Tumors with extra-
mucosal wall, and they usually are small in size and parietal extensions are commonly associated with perivis-
may not be localized during a CECT examination. ceral fat stranding and perivisceral enlarged lymph nodes.
When visible, neuroendocrine tumors show a Conventional CT has a low sensitivity to visualiz-
hypervascular behavior. ing colonic lesions without proper intestinal prepara-
• Some inflammatory diseases such as Crohn’s dis- tion including gas bowel insufflation and antispastic
ease may affect the small bowel and in particular medication. The stools inside the intestinal lumen and intes-
the distal tract of the ileum. CECT examination tinal peristalsis decrease the ability to diagnose these lesions
with fluid distension of the intestinal loops using an in their early stages. CT is therefore used for regional and
oral contrast media is preferable. systemic staging of intestinal cancer. When cancer affects
the rectum, MRI is recommended for local staging, due to its
greater sensitivity and specificity in the study of the rectal
wall and soft tissues (mesorectal space).
14.15 Colorectal Cancer Colorectal cancer in advanced stages is often associated
with the presence of pathological lymph nodes that show a
Colorectal cancer is the most common tumor of the gastroin- characteristic spread along the lymphatics associated with
testinal tract. It most frequently affects the rectum, then the the arterial supply of the affected intestinal tract. For exam-
cecum and ascending colon, while transverse and descend- ple, if cancer involves the rectum, pathological lymph nodes
ing colon tracts are less commonly involved. The most are more commonly those in the mesorectal space and belong
important acquired risk factors are related to a fat-rich diet, to the lower mesenteric artery territory.
chronic inflammatory bowel disease, obesity, asbestos expo- The most important CECT criteria to determine whether a
sure, pelvic irradiations, and family history of benign or lymph node is pathological is the size (in particular if the
malignant colonic lesions. Some hereditary syndromes such short axis of the lymph node exceeds 10 mm) and morpho-
as familial adenomatous polyposis syndrome (FAP), Peutz- logical criteria (the shape and the presence of a conserved
fatty sinus). Unfortunately, using these criteria, it is likely to tile women (derived from Graafian follicles), while cystic
underestimate the colorectal cancer due to low sensitivity formations (in particular if large in size, with septa or nod-
and specificity. The risk of an underestimation can lead ules) in a menopausal woman, should be considered with
to incorrect management and may influence a patient’s prog- suspicion. However, in postmenopausal women, small cystic
nosis. MR and PET/CT are more accurate modalities in lesion can be appreciated in the adnexa until 5 years after the
detecting the presence of colorectal cancer menopausal onset.
lymphadenopathies. Simple cystic lesion management is based on dimensional
criteria and stratified for risk in accordance with premeno-
pausal or postmenopausal status. In premenopausal women, a
simple cyst from 3 cm to 5 cm in diameter must be mentioned
Key Learning Points
in the CT report; a simple cyst with a diameter between 5 and
• Colorectal cancer is the most common tumor of the
7 cm must be mentioned and should undergo ultrasound fol-
gastrointestinal tract.
low-up until it disappears; MRI and surgery are mandatory
• Colorectal cancer may appear as a mass that nar-
for a simple cyst with a diameter of > 7 cm. In postmeno-
rows the bowel lumen with soft tissue density on
pausal women, simple cysts between 2 cm and 7 cm must be
NECT and strong contrast enhancement on
mentioned in the CT report, and a yearly follow-up with ultra-
CECT. Ulcerations and shouldering of the lesions
sound should be performed; when simple cysts have a diam-
margins may be observed in a large mass. Tumors
eter >7 cm, further examinations are mandatory.
with extra-parietal extensions are commonly asso-
Hemorrhagic cysts show density values near to blood
ciated with perivisceral fat stranding and enlarged
products (from 50 UH to 60 UH). If a hemorrhagic cyst in a
perivisceral lymph nodes.
premenopausal woman has a diameter of > 5 cm and remains
• Conventional CT has a low sensitivity for visualiz-
unchanged after 1 year of follow-up, MRI examination must
ing colonic lesions without proper intestinal prepa-
be performed. MRI or surgical evaluation is mandatory in
ration including gas bowel insufflation and
women in early menopause with a hemorrhagic cyst of
antispastic medication. When cancer affects the rec-
> 5 cm or for any hemorrhagic cyst in women in late meno-
tum, MRI is recommended for local staging, due to
pause. Clot in hemorrhagic cysts may mimic a solid nodule,
its greater sensitivity and specificity in the study of
and, if visualized, a MRI exam should be performed.
the rectal wall and soft tissues (mesorectal space).
The CT findings of cystic mature teratoma are a hypo-
attenuating unilocular mass near to fluid density in which a
mural nodule (Rokitansky nodule) or calcifications can be
observed. The presence of fat tissue (negative density values
14.16 Pelvic Lesions on ROI measurement) is pathognomonic for mature cystic
teratoma diagnosis.
Lesions of the male pelvis, such as prostate cancer, are not The other indeterminate cystic lesions of the adnexa are
well appreciated on CT due to the poor contrast difference cystadenoma, cystadenofibroma, mucinous cystadenocarci-
between tissues that decreases the sensitivity and specificity noma, and serous ovarian cystadenocarcinoma.
of the method. Therefore, only the most common female pel- Cystadenoma and cystadenofibroma are benign ovarian
vic lesions are discussed here. neoplasms that can be serous (uniloculated) or mucinous
Ovarian lesions can be classified as benign or malignant (multiloculated). When these lesions are bigger than 7 cm
and solid or cystic. The type of diagnostic and therapeutic in size, have multiple septa and demonstrable contrast
approach changes in relation to the menopausal status of the enhancement on CECT, mural nodules showing contrast
patient. Cystic lesions, as in other parenchymal organs, are enhancement, and vascularized thick walls, they must be
characterized by fluid attenuation (similar to water density considered as having a high likelihood of neoplastic trans-
from 0 HU to 10 UH). During menopause the ovaries gener- formation. Mucinous ovarian cystadenocarcinoma and
ally decrease in size and are predominantly replaced by serous ovarian cystadenocarcinoma are the malignant vari-
fibrotic tissue. In the premenopausal time, the ovaries have ant. Advanced-stage ovarian tumors are often associated
different morphology relating to the menstrual cycle. with the presence of ascites and peritoneal carcinomatosis,
The ovarian lesions have some characteristic CT patterns characterized by the thickened appearance of peritoneal
that can aid in categorizing the lesions as simple cysts, hem- sheaths or with small parietal nodules with demonstrable
orrhagic cysts, mature cystic teratomas (dermoid cyst), and contrast enhancement.
other indeterminate cystic lesions. Metastatic lesions of the ovaries are more often solid
Simple cystic appearance lesions, particularly when single (Krukenberg metastases) and bilateral, but cystic metastases
and from 3 to 7 cm in diameter, are commonly found in fer- can also occur.
14.17.1 Primary Bone Lesions

Key Learning Points
• Ovarian lesions have some characteristic CT patterns The differential diagnosis of a primary bone lesion is heavily
that can aid to categorize the lesions as a simple cyst, influenced by the age of the patient and the location.
hemorrhagic cyst, mature cystic teratomas (dermoid Most likely malignant findings at CT are periosteal reac-
cyst), and other indeterminate cystic lesions. tion (Codman’s triangle), sunburst or lamellated periosteal
• Simple cystic lesion management is based on dimen- reaction, cortical destruction, and soft tissue involvement. CT
sional criteria and stratified for risk in accordance benign findings of periosteal reaction are well-defined mar-
with premenopausal or postmenopausal status. gins and benign periosteal reaction (thick, continuous, or
• The CT findings of cystic mature teratoma are a wavy periosteal thickening). Most radiological characteristics
hypo-attenuating unilocular mass near to fluid den- could overlap in many cases, so osteolytic or sclerotic type,
sity in which a mural nodule (Rokitansky nodule) age of patient, and site of the lesions can reduce the differen-
or calcifications can be observed. tial diagnosis. Bone tumor may appear as predominantly lytic
• The presence of fat tissue (negative density values lesions if bone destruction is predominant or sclerotic if new
on ROI measurement) is pathognomonic for mature bone formation or tumor matrix calcifications are present.
cystic teratoma diagnosis. Osteolytic lesions include the following:
• Metastatic lesions of the ovaries are more often Fibrous dysplasia or fibrous cortical defect, osteoblastoma,
solid (Krukenberg metastases) and bilateral, but giant cell tumor, aneurysmal bone cyst, chondroblastoma or
cystic metastases can also occur. chondromyxoid fibroma, hyperparathyroidism (brown tumor),
non-ossifying fibroma (NOF), enchondroma or eosinophilic
granuloma, and simple (unicameral) bone cyst.
Sclerotic lesions include the following:
14.17 Bone Lesions When solitary, the differential diagnoses include enosto-
sis (bone island), osteosarcoma, chondrosarcoma, calcifying
Bone lesions can be divided into bone neoplasms, bone infec- enchondroma, osteoblastoma, osteoid osteoma, Paget’s dis-
tions, bone-related lymphoproliferative disorders, bone- ease, callus after fracture, and chronic osteomyelitis
related endocrinological disorders, and miscellaneous. Bone (Fig. 14.24). When sclerotic lesions appear as multifocal, the
neoplasms can be primary or secondary. Primary bone tumors differential diagnoses include the metastasis from prostate
can be classified in relationship to primary or secondary and and breast cancer, bone islands, hemangiomas (especially
with respect to the type of tissue: bone-, cartilage-, or fibrous when located in the spine), multiple infarct (drugs related),
forming tumors, bone marrow tumors, and miscellanea. and Paget’s disease.
Fig. 14.24 Sclerotic primary

bone tumor. NECT on coronal a b
plane (a) shows a large
sclerotic lesion in the
diaphysis of the femur with a
“sunburst” periosteal reaction
(white arrowhead): the lesion
resulted in osteosarcoma at
biopsy. NECT on coronal
plane (b) exhibit a large
lesion with multiple
calcifications and scalloped
margins (red arrowhead).
Low-grade chondrosarcoma
was diagnosed at biopsy
14.17.2 Secondary Bone Lesions usually as part of a disseminated process presenting as lytic
infiltrative lesions (with permeative pattern) or bone sclerosis
Skeletal metastases account for 70% of all malignant bone neo- that may mimic other primary and secondary bone neoplasms
plasms. Lung cancer, breast cancer, renal cell carcinoma, and on CT imaging. The pelvis and spine are commonly involved.
prostate cancer account for approximately 80% of all skeletal Skeletal metastases have three different patterns, depend-
metastases. Skeletal metastases tend to be multiple when pres- ing on the preponderant process of bone resorption (osteo-
ent. The lymphoproliferative disorder may involve the bone lytic) or bone formation (osteoblastic) (Fig. 14.25) or mixed:
a b
Fig. 14.25 Bone metastases. Sclerotic metastases from prostate carcinoma on dorsal and lumbar spine (a). Osteolytic metastases on dorsal spine
(b) and sacrum (c) from lung adenocarcinoma
Osteolytic metastases are most likely due to thyroid can-

cer, lung cancer, renal cell cancer, gastrointestinal carcino- Key Learning Points
mas, melanoma, hepatocellular carcinoma, and squamous • The differential diagnosis of a primary bone lesion
cell carcinoma of the skin. is heavily influenced by the age of the patient and
Sclerotic metastases are most likely due to prostate carci- the location. Most likely malignant findings on CT
noma, breast carcinoma (may be mixed), transitional cell are periosteal reaction (Codman’s triangle), sun-
carcinoma (TCC), carcinoid, medulloblastoma, neuroblas- burst or lamellated periosteal reaction, cortical
toma, mucinous adenocarcinoma of the gastrointestinal tract destruction, and soft tissue involvement.
(e.g., colon carcinoma), and lymphoma. • Most radiological characteristics could overlap in
Mixed lytic and sclerotic metastases are related to breast many cases, so osteolytic or sclerotic type, age of
carcinoma and lung carcinoma; metastases from cervical patient, and site of the lesions can reduce the dif-
cancer and testicular tumors are typically lytic but rarely can ferential diagnosis.
be mixed. Prostate carcinomas are typically sclerotic but can • Skeletal metastases account for 70% of all malignant
be mixed in a small fraction of patients. bone neoplasms. Lung cancer, breast cancer, renal
cell carcinoma, and prostate cancer account for
approximately 80% of all skeletal metastases. Skeletal
14.17.3 Miscellaneous Bone Lesions metastases tend to be multiple when present.
• Osteolytic metastases are most likely due to thyroid
Paget’s disease is a common, chronic bone disorder that cancer, lung cancer, renal cell cancer, gastrointesti-
leads to bone deformity. Paget’s disease usually presents nal carcinomas, melanoma, hepatocellular carci-
with all three types of bone alteration (sclerotic, lytic, and noma, and squamous cell carcinoma of the skin.
mixed) due to different amounts of osteoclastic or osteoblas- • Sclerotic metastases are most likely due to prostate
tic activity at different stages of disease: lytic form (initial carcinoma, breast carcinoma (may be mixed), tran-
active moment), mixed form (active form), and sclerotic sitional cell carcinoma (TCC), carcinoid, medullo-
form (late stage). When the spine is involved, the typical cor- blastoma, neuroblastoma, mucinous
tical thickening and sclerosis of subchondral and mural bone adenocarcinoma of the gastrointestinal tract (e.g.,
gives a “picture frame” appearance of the vertebral bodies. colon carcinoma), and lymphoma.
When the skull is involved, in the acute phase osteoporosis • Mixed lytic and sclerotic metastases are related to
circumscripta (lytic lesion) may be observed. In long bones, breast carcinoma, lung carcinoma (typically lytic but
the classic sign is the “candle flame sign” that is an area of a rarely can be mixed), carcinoma of the cervix, and
V-shaped lucency extending from the subchondral space testicular tumors. Prostate carcinomas are typically
toward the diaphysis. sclerotic but mixed in a small fraction of patients.
Infectious disorders: osteomyelitis is an infection of the • Paget’s disease is a common, chronic bone disorder
bone most likely due to bacterial pathogens (pyogenic) or that leads to bone deformity. Paget’s disease usually
other pathogens such as tuberculosis, syphilis, or fungus. presents with all three types of bone alteration
Osteomyelitis can be acute or chronic. In the acute phase (sclerotic, lytic, and mixed) due to different amounts
(after 2 weeks), the most important CT findings are bone of osteoclastic or osteoblastic activity at differ-
marrow lucency areas, cortical thickening with periosteal ent stages of disease.
reaction (periostitis), and endosteal scalloping. In the chronic
phase the most relevant sign is sequestrum. The cloaca sign
may also be observed.
Involvement of the spine (spondylitis) by tuberculosis is 14.18 Abdominal Lymphadenopathies
also called Pott disease. Usually, both the vertebral body and
the disk area are simultaneously affected by the disease. CT interpretation criteria of lymphadenopathies: CT imag-
The most relevant findings on CECT are irregularity of ing for the assessment and characterization of pathological
affected vertebrae at the level of the end plates (initial phase), lymph node disease takes into account the size, the number,
while extension into paravertebral soft tissue is more fre- the morphology, the margins, the attenuation characteristics,
quently observed in the late phase. Vertebra plana is a late and the enhancement after contrast.
finding in which the vertebra has completely lost its anterior Increase in size is not always pathologic, because often
and posterior height. reactive nodes are bigger than 10 mm of transverse diam-
eter (considered as benign limit in size). Hyper-attenuating Erasmus JJ, McAdams HP, Connolly JE. Solitary pulmonary nodules:
Part II. Evaluation of the indeterminate nodule. Radiographics.
lymphadenopathy on NECT can be observed in Kaposi 2000b;20(1):59–66.
sarcoma, Castleman disease, carcinoid, and angioimmuno- Grazioli L, Olivetti L, Fugazzola C, et al. The pseudocapsule in hepato-
blastic lymphadenopathy. Calcifications are present in cellular carcinoma: correlation between dynamic MR imaging and
tuberculosis, sarcoidosis, lymphoma treated with radio- pathology. Eur Radiol. 1999;9:62–7.
Greenspan A, Jundt G, Remagen W. Differential diagnosis in orthopae-
therapy, papillary thyroid carcinoma, and breast cancer. dic oncology. Philadelphia: Lippincott Williams & Wilkins, c2007.
Lymph node metastatic disease commonly occurs along (2006) ISBN:0781779308.
lymphatic pathways. For example, breast cancer tends to Helvie MA, Rebner M, Sickles EA, et al. Calcifications in metastatic
metastasize to axillary lymph nodes and rectal cancer to breast carcinoma in axillary lymph nodes. AJR Am J Roentgenol.
1988;151(5):921–2.
lymph nodes of the inferior mesenteric artery and meso- Herts BR, Megibow AJ, Birnbaum BA, et al. High-attenuation
rectal space. lymphadenopathy in AIDS patients: significance of findings at
CT. Radiology. 1992;185(3):777–81.
Hoang JK, Vanka J, Ludwig BJ, Glastonbury CM. Evaluation of cer-
vical lymph nodes in head and neck cancer with CT and MRI:
Key Learning Point tips, traps, and a systematic approach. AJR Am J Roentgenol.
• CT imaging for the assessment and characterization 2013;200(1):W17–25.
Ichikawa T, Federle MP, Grazioli L, et al. Fibrolamellar hepatocellu-
of pathological lymph node disease takes into lar carcinoma: imaging and pathologic findings in 31 recent cases.
account the size, the number, the morphology, the Radiology. 1999;213:352–61.
margins, the attenuation characteristics, and the Itai Y, Matsui O. Blood flow and liver imaging. Radiology.
enhancement after contrast. 1997;202:306–14.
Jeung MY, Gangi A, Gasser B, Vasilescu C, Massard G, Wihlm
JM, Roy C. Imaging of chest wall disorders. Radiographics.
1999;19(3):617–37.
del Pilar Fernandez M, Redvanly RD. Primary hepatic malignant neo-
Further Reading plasms. Radiol Clin N Am. 1988;36:333–48.
Lucey BC, Stuhlfaut JW, Soto JA. Mesenteric lymph nodes: detection and
Bernardino ME. Computed tomography of calcified liver metastases. J significance on MDCT. AJR Am J Roentgenol. 2005;184(1):41–4.
Comput Assist Tomogr. 1979;3:32–5. Nasseri F, Eftekhari F. Clinical and radiologic review of the nor-
Carter BW, Benveniste MF, Madan R, Godoy MC, de Groot PM, Truong mal and abnormal thymus: pearls and pitfalls. Radiographics.
MT, Rosado-de-Christenson ML, Marom EM. ITMIG classification 2010;30(2):413–28.
of mediastinal compartments and multidisciplinary approach to Nickell LT Jr, Lichtenberger JP 3rd, Khorashadi L, Abbott GF, Carter
mediastinal masses. Radiographics. 2017;37(2):413–36. BW. Multimodality imaging for characterization, classification,
Chen IY, Kats DS, Jeffrey RB, et al. Do arterial phase helical CT image and staging of malignant pleural mesothelioma. Radiographics
improve detection or characterization of colorectal liver metastases? 2014;34(6):1692-1706.
J Comput Assist Tomogr. 1997;21(3):391–7. Nishino M, Ashiku SK, Kocher ON, Thurer RL, Boiselle PM,
El-Sherief AH, Lau CT, Wu CC, Drake RL, Abbott GF, Rice Hatabu H. The thymus: a comprehensive review. Radiographics.
TW. International association for the study of lung cancer 2006;26(2):335–48.
(IASLC) lymph node map: radiologic review with CT illustration. Truong MT, Ko JP, Rossi SE, Rossi I, Viswanathan C, Bruzzi JF, Marom
Radiographics. 2014;34(6):1680–91. EM, Erasmus JJ. Update in the evaluation of the solitary pulmonary
Erasmus JJ, Connolly JE, McAdams HP, Roggli VL. Solitary pulmo- nodule. Radiographics. 2014;34(6):1658–79.
nary nodules: Part I. Morphologic evaluation for differentiation of Whitten CR, Khan S, Munneke GJ, Grubnic S. A diagnostic approach
benign and malignant lesions. Radiographics. 2000a;20(1):43–58. to mediastinal abnormalities. Radiographics. 2007;27(3):657–71.
Essentials of MR Image Interpretation
15
Contents
15.1 Introduction 317
15.2 Signal Characteristics of Tissue Components in MR Images 318
15.2.1 T1-Weighted Images 318
15.2.2 T2-Weighted Images 318
15.2.3 Water-Fat Signal Decomposition, Fat Suppression, and Dual-Phase Chemical Shift Imaging 319
15.2.4 Diffusion-Weighted Imaging 320
15.2.5 Contrast-Enhanced Images 320
15.3 Fundamentals of Imaging Interpretation: Common Patterns of Disease 321
15.3.1 Fluid-Containing Lesions 321
15.3.2 Signal of Blood-Containing Lesions 322
15.3.3 Fat-Containing Lesions 322
15.3.4 Solid Lesions 324
15.4 Essentials of Differential Diagnosis by Anatomical Region 325
15.4.1 Brain Lesions 325
15.4.2 Neck Masses 329
15.4.3 Evaluation of Mediastinum and Pleura 330
15.4.4 Evaluation of Abdomen and Pelvis 333
Further Reading 350
Learning Objectives • Correlate the most common signal alterations in brain

• Understand how to distinguish various tissue components lesions with pathology.
based on signal intensity on T1-weighted and T2-weighted • Correlate the most typical imaging features of cystic and
images. solid lesions in different anatomical regions with pathology.
• Describe the most common causes of signal alterations on • Describe the most common signal alterations in the mus-
diffusion-weighted images. culoskeletal system.
• Define an approach to lesion characterization based on
identification of four main internal components (fluid, fat,
blood, and solid tissues).
15.1 Introduction
The definition “hybrid imaging” refers to the combination of

D. Caramella (*) cross-sectional imaging techniques used in radiology (CT
Diagnostic and Interventional Radiology, Department of and MRI) with radionuclide imaging used in nuclear medi-
Translational Research and Advanced Technologies in Medicine
and Surgery, University of Pisa, Pisa, Italy cine. PET/MRI is a rapidly evolving hybrid imaging tech-
e-mail: davide.caramella@med.unipi.it nique being increasingly employed in clinical practice
F. Chiesa and may represent the most challenging modality for nuclear
Unit of Radiology, ASL 5 “Spezzino”, Sarzana, La Spezia, Italy medicine specialists. In the simple PET/CT examination,

318 D. Caramella and F. Chiesa
low-dose unenhanced CT images are acquired for both atten- Table 15.1 Signal intensities of different body components on
uation correction and anatomical localization, with a limited T1-weighted and T2-weighted images
contribution for characterizing lesions detected by tracer Signal
intensity T1-weighted T2-weighted
uptake in the PET scan. In contrast, even the most generic
High Fat, highly proteinaceous Clear fluids (e.g.,
whole body PET/MRI hybrid imaging, without specific fluids, gadolinium, CSF, bile, urine), fat
organ- targeted or contrast-enhanced sequences, requires methemoglobin, other (in TSE sequences)
interpretation of multiple images including T1-weighted, paramagnetic substances
T2-weighted, and diffusion-weighted images. In the first part Intermediate Most parenchymal organs, Most parenchymal
of this chapter, the appearance of various tissue components muscle tissue, proteinaceous organs, muscle
fluids tissue
on pulse sequences commonly used in MR examinations will Low Clear fluids, tendons, Tendons, ligaments,
be described, and the most common imaging patterns of dis- ligaments, fibrous tissue fibrous tissue
ease – classified on the basis of prevalent tissue composi- Signal void Calcium, cortical bone, air Calcium, cortical
tion – will be illustrated. In the second part, imaging findings bone, air, fast-
useful for differentiating pathological processes in various flowing blood
organs will be discussed, with particular emphasis on mass
lesions commonly evaluated on MRI. cellular and extracellular molecules and variable amounts
of free water. For this reason all parenchymal organs dem-
onstrate lower T1 signal intensity than fat, with organ-spe-
15.2 S
ignal Characteristics of Tissue cific differences determined by complex and often
Components in MR Images incompletely understood biochemical and structural
factors.
15.2.1 T1-Weighted Images Muscle tissue has intermediate signal on T1-weighted
images comparable to that of some parenchymal organs.
Pulse sequence parameters of T1-weighted acquisitions Dense fibrous tissue, the main constituent of tendons, liga-
(short TR and TE) create images with contrast between flu- ments, and mature fibrotic scars, exhibits very low signal
ids and tissues, both normal and pathological, to create intensity on T1-weighted images owing to the low content
images, based on differences in T1 relaxation characteristics of free water and strong internal dipolar interactions.
(Table 15.1). Shorter T1 relaxation times correspond to a Similarly, cortical bone and dense calcium deposits pres-
higher signal intensity on T1-weighted images. Normal tis- ent as signal voids due to the very low density of resonat-
sues that show the shortest T1 relaxation times are composed ing protons. Gaseous components show no appreciable
of molecular species capable of rapid exchange of absorbed signal.
radio-frequency energy with the surrounding medium. This Vascular structures usually appear hypointense on
is usually achieved through dipolar interactions in a thermo- T1-weighted images. In specific pulse sequences called
dynamically irreversible process generating heat. The most “time of flight” a phenomenon known as flow-related
important molecular entities with these properties are: enhancement can be exploited to generate angiographic
images in which flowing blood shows a markedly high signal
–– Fatty acids, the predominant biochemical constituents of intensity.
mature adipose tissue
–– Water solutions containing dispersed macromolecules,
more often represented by proteinaceous and/or muci- 15.2.2 T2-Weighted Images
nous substances that slow random movements of water
molecules whose tumbling rate matches the Larmor Pulse sequence parameters for acquisition of T2-weighted
frequency images (long TR and TE) are designed to create image
–– Substances with paramagnetic properties, including contrast between fluids and tissues, both normal and path-
gadolinium-based contrast media, specific blood degrada- ological, based on differences in T2 relaxation time
tion products, and some metallic atoms, such as copper (Table 15.1). Longer T2 relaxation times correspond to a
and manganese. higher signal on T2-weighted images. Molecular tumbling
rate and the consequent frequency of dipolar interactions
Water solutions with no appreciable proteinaceous or are the main factors influencing the T2 relaxation
paramagnetic content show very low signal intensity on mechanism.
T1-weighted images. Free water protons, as found in physiological bodily flu-
Parenchymal organs are composed of a mixture of dif- ids like bile, urine and cerebrospinal fluid, have the longest
ferent tissues, which in turn contain a wide range of intra- T2 relaxation time and thus show the highest signal. In
15 Essentials of MR Image Interpretation 319
pulse sequences currently used in clinical practice for 15.2.3 Water-Fat Signal Decomposition, Fat
obtaining T2-weighted images (such as HASTE, single- Suppression, and Dual-Phase Chemical
shot TSE, and FSE) adipose tissue, mainly composed of Shift Imaging
fatty acids, shows hyperintense signal, which is slightly
lower than water but higher than parenchymal organs. Fat is predominantly composed of medium and long chain
T2-weighted sequences with very long TE can be used to fatty acids containing numerous hydrogen atoms linked to a
image viscera containing nearly static fluid, with a common carbon framework which is responsible for the characteristic
example being magnetic resonance cholangiopancreatogra- spectral behavior (difference in precession frequency com-
phy (MRCP). pared to water) called “chemical shift.” This physical prop-
Parenchymal organs demonstrate low to intermediate sig- erty can be exploited by using particular pulse sequences to
nal intensity on T2-weighted images. Organ-specific differ- achieve specific effects in image contrast.
ences are due to variations in tissue composition, especially In Dixon technique, water and fat signal is decomposed to
free water content and concentration of mineral ele- generate three-dimensional images containing signal from a
ments such as iron. single proton species (“water-only” and “fat-only” images)
Muscle tissue shows lower signal intensity than most or a combination of the two. This method is commonly
parenchymal organs on T2-weighted images. Similar to T1- applied to three-dimensional gradient-echo sequences for
weighted images, fibrous tissue, cortical bone, and dense cal- generating attenuation maps used for correction of PET
cifications appear dark. Gaseous components exhibit no transmission scan in hybrid imaging.
signal. On fat-saturated sequences and STIR sequences, the sig-
Substances with superparamagnetic properties and highly nal from fat is suppressed by spectral selective saturation and
concentrated paramagnetic molecules cause local magnetic inversion recovery techniques respectively. On T1-weighted
field inhomogeneities that accelerate T2 relaxation processes images, fat saturation is commonly applied to three-dimen-
and cause a loss of signal on T2-weighted images. The most sional fast gradient-echo sequences for breath-hold studies
notable examples are hemosiderin, the main storage form of of the chest and abdomen before and after administration of
iron in tissues, deoxyhemoglobin packed in red blood cells, gadolinium-based contrast agents. On T2-weighted fat-satu-
concentrated gadolinium normally excreted in the urinary rated sequences, nulling of fat signal increases the contrast
tract, and by-products of melanin metabolism in melanocytic resolution of other tissue components and increases the con-
tumors. spicuity of fluid-containing structures. STIR sequences
Vascular structures can show variable signal on achieve similar results with more complex effects on tissue
T2-weighted images depending on flow characteristics. contrast due to the T1 dependency of inversion recovery
Rapid flowing blood in arterial vessels usually appears dark technique. The main practical difference between inversion
because of a flow-related phenomenon denominated “flow recovery and fat saturation techniques is the low susceptibil-
void”. However, arterial vessels may show high signal inten- ity of the former to magnetic field inhomogeneities. STIR
sity similar to static fluid because of random synchronization sequences are therefore advantageous in the study of ana-
of acquired data with the diastolic phase of cardiac cycle. tomically complex regions such as the neck and whole body
This finding can be typically seen in the aorta and other large imaging of the skeletal system.
arteries running perpendicular to the imaging plane on rapid- In MR imaging protocols for the evaluation of thoracic
shot sequences. and abdominal regions, the most commonly used pulse
In fluid-attenuated inversion recovery (FLAIR) sequences, sequence exploiting the chemical shift effect is dual-phase
signal of static fluids is nulled by precisely timed inversion T1-weighted gradient echo. Two different types of images
pulses. FLAIR sequences are used in brain MRI to suppress are obtained at specific TE, depending on magnetic field
cerebrospinal fluid signal in order to increase the conspicuity intensity. In the first, named “out-of-phase” image, water and
of parenchymal lesions. fat null each other if equally represented in the same image
Gradient-echo pulse sequences with relatively long TE voxel. In the second, named “in-phase” image, water and fat
are weighted for a parameter named T2* (“T2 star”) and contained in the same voxel combine their signal. Dual-
are very sensitive to magnetic field inhomogeneities. T2*- phase chemical shift imaging with in-phase and out-of-phase
weighted sequences are useful to identify small calcifica- acquisitions allows recognition of microscopic quantities of
tions and superparamagnetic deposits such as hemosiderin, fat that occupy less than a voxel, undetectable by selective fat
often not detectable on T2-weighted sequences. On gradient- suppression techniques that act on voxels containing pre-
echo T2*-weighted images, calcium and iron deposits appear dominantly fat. These sequences may also be used to detect
as very low signal intensity areas with exaggerated size com- substances typically associated with susceptibility artifacts
pared to their real volume (“blooming” effect). such as iron and calcium deposits; the resulting blooming
effect is more conspicuous on “in-phase” compared to “out- –– High cellularity that correlates microscopically with the
of-phase” images due to the longer TE. presence of many cellular barriers that hinder water diffu-
sion, as found in many malignant and some benign
tumors.
15.2.4 Diffusion-Weighted Imaging –– Intracellular edema with relative reduction of extracellu-
lar volume as seen in cerebral ischemia.
Diffusion-weighted sequences are obtained with the applica- –– Fluid collections with a highly compartmentalized struc-
tion of symmetrical gradient waveforms before and after a ture like purulent material in abscesses and clotted blood
spin-echo refocusing pulse with an echo planar data readout. in hematomas.
The extrinsic parameter “b-value” determines the sensitivity
to molecular diffusion phenomena: a b-value of 0 means no
diffusion weighting, while progressively increasing b-values 15.2.5 Contrast-Enhanced Images
correspond to higher diffusion sensitivity. In the typical
diffusion-weighted imaging protocol, two or three b-values Gadolinium-based contrast agents are water-soluble mole-
are set (from 0 up to a maximum b-value of 1000 for most cules with strong paramagnetic properties that shorten
applications), and apparent diffusion coefficient (ADC) T1 relaxation time of fluid compartments in which they dis-
maps are generated. tribute, resulting in signal enhancement. Immediately after
Tissues or pathological processes characterized by a intravenous administration, the plasma concentration of
restricted diffusion of water molecules show high signal gadolinium rapidly rises and reaches a peak in arterial
intensity on high b-value diffusion-weighted images. On the blood vessels followed by a gradual decline as contrast
other hand, free water (as found in simple fluid-containing passes in the intravascular compartment and then diffuses
structures) appears hyperintense on images with a b-value of in the extracellular space. Multiphase dynamic evaluation
zero, which are predominantly T2-weighted, and shows of contrast enhancement is commonly performed in MR
no signal on high b-value images. In some cases, lesions with imaging protocols using fat-saturated three-dimensional
high signal intensity on images with a b-value of zero may T1-weighted gradient-echo sequences that can be acquired
appear bright on moderate to high b-value images even if in less than 20 s during a single breath-hold. Turbo spin-
they do not significantly restrict diffusion, a phenomenon echo T1-weighted sequences are typically used in static
known as “T2 shine-through”. ADC maps synthetically and anatomical regions such as the head, neck, spine, pelvis,
quantitatively convey the information of tissue diffusion and limbs.
characteristics at different b-values and prove very useful to Images acquired during the maximum arterial concentra-
avoid the pitfall of T2 shine-through effect. tion of gadolinium are used in contrast-enhanced MR angi-
At low b-values (b < 100), bulk movement of free water in ography. In late arterial phase, starting about 10 s after
blood vessels and perfusive components linked to transcapil- arterial peak of contrast, parenchymal organs begin to
lary diffusion of water dominate over random Brownian enhance, with higher contrast enhancement seen in more
movements, determining complete signal loss in vascular vascularized tissues such as renal cortex, spleen, and pan-
structures, called “black blood” effect. Diffusion-weighted creas. In the dynamic contrast-enhanced study of the liver,
images obtained at low b-values are mostly used in evalua- late arterial phase is called hepatic arterial phase or “portal
tion of the liver because they allow detection of focal lesions, inflow” phase because portal veins and their branches start to
both benign and malignant, with a high sensitivity. Indeed, enhance, while hepatic veins remain hypointense. Late arte-
small liver lesions may be difficult to identify on T2-weighted rial phase images are essential in detecting hypervascular
images with or without fat suppression due to low contrast lesions in the liver, and for the evaluation of pancreatic
between lesion and parenchyma or may be mistaken for a pathology.
normal vascular structure. In images acquired during the venous phase (that starts
In normal conditions, peripheral lymphoid tissue such as approximately one minute after gadolinium administration)
lymph nodes and spleen, shows restricted diffusion and arterial enhancement decreases, venous vessels progres-
appears hyperintense on high b-value images compared to sively enhance, spleen homogeneously enhances, and con-
organs like the liver or kidneys. This property makes trast appears in renal medulla. In the liver, the venous phase
diffusion-weighted imaging very sensitive for detection of is also called portal venous phase because of balanced
both normal and pathological lymph nodes. enhancement between portal system and hepatic veins. In
Disease processes that demonstrate restricted diffusion these images the highest contrast between hypovascular
are mostly characterized by: lesions and parenchyma can be obtained.
During the equilibrium phase, occurring between three

and five min after contrast administration, gadolinium dis- • Simple fluids like bile, urine, and cerebrospinal
tributes from the intravascular compartment to the interstitial fluid show high signal intensity on T2-weighed
extracellular space. Images in equilibrium phase contribute images.
to the characterization of focal lesions, especially in the liver, • Substances with superparamagnetic properties like
based on the pattern and the temporal evolution of contrast hemosiderin show low signal intensity on gradient-
enhancement. Lesions demonstrating the “washout” phe- echo T2*-weighted images.
nomenon, often due to intralesional arteriovenous shunts and • Dixon technique allows separation of water and fat
altered vascular permeability, appear hypointense compared signal. Fat signal can be nulled on fat-saturated and
to parenchyma, while accumulation of gadolinium resulting STIR images.
in hyperintensity may correlate with an increased interstitial • Restricted diffusion appears hyperintense on high
space, as commonly found in fibrotic lesions. b-value diffusion-weighted images.
Hepatospecific contrast agents are a particular class of • Low b-value diffusion-weighted images result in a
gadolinium-based contrast agents that can accumulate in “black blood” effect that may increase the detection
normally functioning liver tissue by a carrier-mediated of focal lesions in liver studies.
uptake and be excreted in biliary canaliculi. Contrast- • Contrast-enhanced images after intravenous admin-
enhanced images acquired in delayed hepatobiliary phase, istration of gadolinium can be acquired in multiple
after a variable time from gadolinium administration (10 min phases.
to a few hours depending on the contrast agent), are useful in
characterizing liver lesions and in detecting liver metastases.
Most gadolinium-based contrast media are predominantly
eliminated via glomerular filtration, which can be exploited 15.3 Fundamentals of Imaging
to visualize the collecting system on images that are acquired Interpretation: Common Patterns
in the late urinary excretion phase (more than 5 min after of Disease
gadolinium administration).
15.3.1 Fluid-Containing Lesions
Fluid components have variable appearances in different MR

Key Learning Points pulse sequences. Water solutions with low concentration of
• Shorter T1 relaxation times correspond to a higher macromolecules show the signal typical of clear fluid:
signal intensity on T1-weighted images. marked T2 hyperintensity, T1 hypointensity, and elevated
• Fat and fluids with high proteinaceous and/or muci- ADC on diffusion-weighted images. Moderate concentra-
nous content and paramagnetic substances, includ- tions of macromolecules shorten T1 with minor effects on
ing gadolinium-based contrast media, show high T2, as typically found in lesions containing proteinaceous or
signal intensity on T1-weighted images. mucinous fluids (Fig. 15.1). Higher concentrations of macro-
• Longer T2 relaxation times correspond to a higher molecules cause T2 shortening that parallels a decrease in
signal intensity on T2-weighted images. signal intensity on T2-weighted images, at times seen only in
the dependent portion of a fluid collection due to sedimenta-
a b c
Fig. 15.1 Autosomal dominant polycystic kidney disease. the presence of blood degradation products determines variable hyperin-
T2-weighted image (a), T1-weighted image (b), and contrast- tensity on T1-weighted image and lower signal on T2-weighted image
enhanced image (c) are shown. Multiple cysts of variable size and signal compared with simple cysts
characteristics are seen bilaterally. Increased proteinaceous content or
tion phenomena. Restricted diffusion may be seen in corpus- described by five consecutive phases following hemorrhage:
culated fluid collections compared to clear fluid. hyperacute (less than 24 h), acute (up to 3 days), early sub-
The presence of blood degradation products may show acute (3 to 7 days), late subacute (up to a month), and
variable signa characteristics related to their concentration chronic.
and to the physico-chemical transformation of hemoglobin. Hyperacute hematoma shows signal features similar to
Cysts and cyst-like masses are the most commonly other fluids, with high signal intensity on T2-weighted
encountered lesions in various anatomical regions, charac- images and intermediate to low signal intensity on
terized by a predominant fluid content. They may share simi- T1-weighted images due to a higher protein content com-
lar imaging findings despite having different pathological pared to clear fluid.
features and clinical significance. Simple cystic lesions Acute-phase hematoma is characterized by deoxygen-
appear as clear fluid-containing formations with sharply ation of hemoglobin; since paramagnetic deoxyhemoglobin
defined margins, usually round or lobulated contours, thin packed inside red blood cells distorts local magnetic fields
walls, and no internal structure, with the possible exception resulting in signal loss, acute hematomas show decreased
of few hairline-thin non-enhancing septa. Benign epithe- signal intensity on T2-weighted images.
lium-lined true cysts and pseudocysts encapsulated by thin In subacute hematoma, the presence of methemoglobin,
fibrotic walls typically appear as simple cystic lesions. Cyst- an oxidation product of hemoglobin with paramagnetic
like fluid collections of inflammatory origin may be delim- properties that can shorten T1 by interacting with water mol-
ited by variably thick walls that enhance after gadolinium ecules, causes high signal intensity on T1-weighted images
depending on vascularity. typically starting from the periphery of hematoma and grad-
Complicated cysts are fluid-containing lesions whose sig- ually spreading in the central portion as hemoglobin catabo-
nal characteristics depend on the presence of proteinaceous lism proceeds and red blood cells lyse.
or bloody material, as seen in hemorrhagic cysts. Abscesses In the early subacute phase, methemoglobin is concen-
consist of circumscribed fluid collections of infectious etiol- trated in intact red blood cells causing susceptibility effects
ogy containing purulent material that usually appears fluid- similar to deoxyhemoglobin (low signal intensity on
like on T1- and T2-weighted images but typically shows T2-weighted images).
restricted diffusion compared with clear fluid. After gado- In the late subacute phase (Fig. 15.2), red blood cell
linium administration a peripheral rim of enhancement is membranes rupture and release free methemoglobin in the
frequently seen surrounding the abscess that may be associ- extracellular fluid with disappearance of local magnetic field
ated with edema of adjacent tissues, appearing as increased distortions (hematoma becomes hyperintense again on
signal on T2-weighted images. T2-weighted images). Notably, fat-suppressed T1-weighted
Complex cystic masses are characterized by a predomi- sequences are useful to distinguish T1 hyperintense fat that
nantly fluid content but show more complex imaging features characteristically loses signal from subacute hematoma that
including thick or irregular walls and internal septations that remains hyperintense.
may enhance after gadolinium, a multilocular appearance In the chronic stage of hematoma evolution there is pro-
with fluid loculations of variable signal intensity, and intral- gressive accumulation of hemosiderin, initially more con-
esional solid components originating from the walls and centrated at the periphery of the hematoma, with strong
septa. superparamagnetic effects that determine low signal inten-
Some complex masses with a higher proportion of solid sity on both T1- and T2-weighted images. Gradient-echo
enhancing tissue often present as lesions with mixed cystic- T2*-weighted images are very effective for detecting blood
solid structure. Solid masses with necrotic or cystic changes products in various phases of degradation owing to their high
show intralesional fluid components and must not be mis- sensitivity to local magnetic field distortions.
taken for complex cystic lesions, usually characterized by Hemorrhagic phenomena may complicate other focal
more regular-shaped loculations delineated by septa. lesions, both cystic and solid, whose signal characteristics
depend on the same processes of blood products degradation
as described for hematomas.
15.3.2 Signal of Blood-Containing Lesions
The most common blood-containing lesion is hematoma, 15.3.3 Fat-Containing Lesions

whether spontaneous or traumatic. Signal of blood outside
the vessel lumen undergoes time-specific changes that For a simplified approach to image analysis of fat-contain-
reflect the progressive biochemical transformations of ing lesions, it is useful to consider separately macroscopic
blood components. Hematoma signal evolution can be and microscopic fat (Fig. 15.3). Macroscopic fat, as found
a b c
Fig. 15.2 Right occipital lobe subacute hemorrhage. due to the presence of paramagnetic extracellular methemoglobin.
T1-weighted image (a), T2-weighted image (b), and ADC maps (c) are Hemorrhage may exhibit restricted diffusion in its early phases likely
shown. Hyperintensity on both T1-weighted and T2-weighted images is due to clot formation
a b
c d
Fig. 15.3 Macroscopic and microscopic fat in a presacral myeloli- T1-weighted out-of-phase image (white asterisk). An area of macro-
poma. T2-weighted image (a), fat-suppressed T2-weighted image (b), scopic fat in the left margin of the mass shows high signal intensity on
T1-weighted in-phase image (c), and T1-weighted out-of-phase image T1-weighted and T2-weighted images with no signal on the fat-
(d) are shown. An inhomogeneous mass in the presacral region contains suppressed image (white arrowhead)
large amounts of microscopic fat detectable by the loss of signal on
in mature adipose tissue and lipomatous lesions, is readily characteristics correlated with their pathological features.
identified by its signal behavior: markedly high signal Most entirely or partially solid lesions usually appear mild to
intensity on T1-weighted images, high signal intensity on moderate hypointense on T1-weighted images and variable
T2-weighted TSE sequences, and signal nulling on fat- hyperintense on T2-weighted images. Contrast enhancement
suppressed images. A mass entirely composed of fat tissue in solid lesions can show variable imaging patterns depend-
typically shows a thin rim of signal loss on dual-echo ing on tissue homogeneity, vascular density and permeabil-
chemical shift imaging localized at the interface with sur- ity, and extracellular space size.
rounding water-containing tissues, denominated “India Certain tissues may have more specific signal characteris-
ink” artifact, caused by similar amounts of lipids and watertics that reflect their distinct pathological features. Tissues
coexisting in the same voxel. In fact this imaging feature is
characterized by a high cellularity typically show restricted
recognizable at every lipid/water interface such as between diffusion and may occasionally exhibit low signal intensity
parenchymal organs and perivisceral adipose tissue. On on T2-weighted images, as can be found in neoplasms com-
T1-weighted sequences employing Dixon technique, fat posed of small tightly packed cells with high nuclear-
shows high signal in “fat-only” images and appears dark in cytoplasmic ratio.
“water-only” images, similar to other fat suppression Myxoid tissue, more frequently found in soft tissue tumors,
techniques. typically appears markedly hyperintense on T2-weighted
Microscopic fat, which may refer to fatty molecules images, at times approaching the signal of fluid due to the
stored in intracellular compartments or very small extracel-prevalent water-rich extracellular matrix, and in most cases
lular fat deposits intermixed with water-rich tissue compo- demonstrates a moderate to marked low signal intensity on
nents in the same voxel, is typically not recognizable on T1-weighted images. After gadolinium administration myxoid
fat-suppressed images but is readily identified by dual- tissue shows mild and progressive enhancement, often with a
phase chemical shift imaging. Lesions containing micro- septal or inhomogeneous lace-like appearance (Fig. 15.4). A
scopic fat show signal loss on dual-phase chemical shift minority of solid adenocarcinomas, mostly derived from the
images that is more conspicuous in areas where fat and gastrointestinal tract, may be predominantly composed of
water are equally represented. A significant predominance mucinous tissue with imaging features resembling myxoid tis-
of one molecular component over the other may cause a sue, characteristically showing unrestricted diffusion.
relatively small signal loss, for example, in conditions like Dense fibrous tissue that may constitute the main solid
mild liver steatosis and inflammatory edema of mesenteric component in some tumors typically shows hypointense sig-
adipose tissue, the latter corresponding to “misty mesen- nal on T2-weighted images and variable signal intensity on
tery” appearance on CT. T1-weighted images that ranges from nearly isointense to
moderately hypointense compared to muscle. Fibrous tissue
classically demonstrates a progressive and delayed enhance-
15.3.4 Solid Lesions ment due to gradual accumulation of gadolinium in the abun-
dant extracellular space. Immature fibrous tissue, as found in
The generic term “solid” refers to a wide spectrum of non- fresh scars, usually shows mildly to moderately elevated sig-
cystic tissue components that may have variable imaging nal intensity on T2-weighted images due to interstitial edema
a b c
Fig. 15.4 Intramuscular atypical lipomatous tumor of the left thigh. image and low signal intensity on fat-suppressed image). An area most
T1-weighted image (a), fat-suppressed T2-weighted image (b), and consistent with myxoid tissue shows high signal intensity on
contrast-enhanced image (c) are shown. The tumor presents as a mass T2-weighted image (white arrowhead) and some peripheral and patchy
mainly composed of adipose tissue (high signal intensity on T1-weighted internal enhancement, after contrast administration
and neovascularization often reflected by a more intense intensity compared to brain parenchyma on T1-weighted
contrast enhancement. images attributable to a relative increase in free water con-
Some solid masses contain calcium deposits whose identi- tent. Some pathological processes may show increased sig-
fication may be useful for lesion characterization. While calci- nal on T1-weighted images, mostly because of the presence
fications typically appear as signal voids in all MR images, of fat, highly concentrated proteinaceous fluids, mineral
gradient-echo sequences allow a better detection of small cal- deposits, blood components, and other substances with para-
cium deposits due to the associated blooming effect. In masses magnetic properties whose identification may assist
with complex structures, solid enhancing tissue can be vari- in diagnosis.
ably mixed with other constituents like fluid, fat, and blood Subacute hemorrhagic collections contain paramagnetic
degradation components. Differential diagnosis of solid methemoglobin that shows high signal intensity on
lesions relies on recognition of tissue composition together T1-weighted images (Fig. 15.2). Lipomas and dermoid cysts
with anatomical location and relevant clinical information. are fat-containing lesions typically located near the midline;
the former appear as a non-enhancing mass entirely com-
posed of fat and the latter as a cystic lesion with a T1 hyper-
intense content due to the presence of fat-rich sebaceous
Key Learning Points material with variable signal intensity on T2-weighted
• Simple cysts and cyst-like fluid collections show images and often calcified walls. Dermoid cysts may compli-
markedly high signal intensity on cate with rupture and the spread of their fatty content in the
T2-weighted images and low signal intensity on subarachnoid space.
T1-weighted images. Cystic lesions with a highly proteinaceous fluid content
• Complicated cysts with proteinaceous or hemor- that may show high signal intensity on T1-weighted images
rhagic content show high signal intensity on are more frequently represented by colloid cyst, a round-
T1-weighted images and variable signal intensity shaped cyst typically located at the foramen of Monro of
on T2-weighted images, often lower than simple third ventricle, and craniopharyngioma often presenting as a
fluids. complex partially solid-cystic mass in the suprasellar region.
• Purulent material in abscesses shows restricted In various toxic-metabolic disorders, mineral substances
diffusion. such as calcium, copper, and manganese can accumulate in
• Blood components exhibit variable signal features certain brain regions, typically in basal ganglia, appearing
according to their stage of evolution. hyperintense on T1-weighted images. Paramagnetic sub-
Deoxyhemoglobin in acute hematomas shows low stances other than gadolinium and methemoglobin that can
signal intensity on T2-weighted images. cause increased signal intensity on T1-weighted images are
Paramagnetic methemoglobin in subacute hemato- by-products of melanin metabolism, frequently seen in mela-
mas shows high signal intensity on T1-weighted noma metastases. The posterior lobe of hypophysis normally
images. Hemosiderin can be detected with a high shows high T1 signal intensity and can be easily identified in
sensitivity on gradient-echo T2*-weighted images. cases of ectopic position.
• Solid tissues demonstrate variable signal intensity
and contrast enhancement features. Myxoid tissue 15.4.1.2 S ignal Alterations on FLAIR
exhibits very high signal intensity on T2-weighted and T2-Weighted Images
images, whereas fibrous tissue appears markedly Most brain pathologies present as hyperintense signal altera-
hypointense. tions compared to brain parenchyma on T2-weighted and/or
• Calcifications appear as signal voids. FLAIR images mostly due to the increased free water con-
tent. Cerebrospinal fluid collections show markedly high sig-
nal intensity on T2-weighted images and signal suppression
in FLAIR sequences. The most common entities that present
15.4 E
ssentials of Differential Diagnosis by as circumscribed clear fluid-filled spaces are dilated perivas-
Anatomical Region cular spaces and chronic lacunar infarcts in parenchymal
regions supplied by deep penetrating arteries; the latter can
15.4.1 Brain Lesions be recognized by the presence of a hyperintense rim typi-
cally surrounding the fluid cavity on FLAIR images.
15.4.1.1 S ignal Alterations on T1-Weighted Arachnoid cysts are relatively common thin-walled cystic
Images lesions of developmental origin located in the subarachnoid
Most brain diseases including vascular, neoplastic, inflam- space, more commonly in middle fossa, that follow cerebro-
matory, and demyelinating frequently show reduced signal spinal fluid signal in all imaging sequences.
A wide variety of intra-axial and extra-axial lesions may Metastases, high-grade glial tumors, and abscesses may
demonstrate a predominantly cystic appearance with fluid share similar imaging features including the presence of
content of variable composition that is often not entirely sup- abundant T2 hyperintense fluid-like components that do not
pressed on FLAIR images. Epidermoid cysts are pure cystic suppress on FLAIR, walls of variable thickness, and mass
masses most frequently located in the cerebellopontine angle effect on surrounding structures. These lesions are frequently
or suprasellar cisterns that show some FLAIR hyperintensity associated with perilesional vasogenic edema that apperas as
due to corpusculated content. The most common extra-axial high signal intensity with a characteristic digitiform appear-
solid tumors, schwannoma and meningioma, may show cys- ance (Fig. 15.5). Malignant necrotic neoplasms tend to have
tic changes that can occasionally result in a predominantly thick irregular walls surrounding the central necrotic portion,
cystic appearance. Some low-grade primary brain neoplasms while abscesses, especially of bacterial etiology, are typically
can appear as predominantly cystic masses with marked high delineated by a thin outer hypointense rim on T2 weighted
signal intensity on T2-weighted images, typically associated images. In fact, it is often impossible to distinguish a necrotic
with a solid component that enhances after gadolinium neoplasm from a pyogenic abscess on the basis of signal char-
administration. A variety of neoplastic and infectious dis- acteristics on T2-weighted and FLAIR images alone.
eases can appear as predominantly fluid-containing lesions Differential diagnoses include other infectious diseases of
that may assume a cyst-like appearance. variable etiology and require evaluation of diffusion-weighted
a b c
d e f
Fig. 15.5 Left temporal lobe glioblastoma. Contrast-enhanced images solid component shows contrast enhancement, relative hypointensity
(a, d), T2-weighted images (b, e), and apparent diffusion coeffi- on T2-weighted image, and reduced diffusion compared to adjacent
cient images (c, f) are shown. The tumor appears as an inhomogeneous tumor (white arrow in c, d, and e), likely attributable to a higher
mass with a rim-enhancing necrotic portion that exhibits unrestricted cellularity
diffusion (white arrowhead in a, b, and c). In a more cranial slice, a
a b c
Fig. 15.6 Right frontal lobe abscess. Contrast-enhanced image (a), showing hyperintense signal on T2-weighted image and restricted dif-
T2-weighted image (b), and apparent diffusion coefficient image (c) are fusion due to the purulent content. Associated perilesional edema shows
shown. Abscess presents as a rim-enhancing mass with a central portion high signal on T2-weighted image
images to identify purulent material (Fig. 15.6). –– Areas of high cellularity in some malignant tumors
Neurocysticercosis often appears as multiple cystic lesion (Fig. 15.5) that may associate with a moderately restricted
disseminated in brain parenchyma and subarachnoid spaces diffusion, most typically seen in densely cellular tumors
with variable imaging features according to the disease stage. such as medulloblastoma and lymphoma.
This has to be distinguished from other multifocal cyst-like –– Some melanoma metastases due to the T2-shortening
masses of infectious and neoplastic origin. effects of intratumoral paramagnetic compounds.
Many other parenchymal diseases show prevalent high –– Cystic lesions with a highly proteinaceous or mucinous
signal intensity on T2-weighted and FLAIR images without content such as colloid cysts.
macroscopic accumulation of fluid. A large number of pri-
mary tumors with different clinicopathological features and The high sensitivity of T2*-weighted gradient-echo
variable location (intraparenchymal, intraventricular, and sequences for susceptibility artifacts generated by calcium
extra-axial) often appear as infiltrative masses composed of and hemosiderin deposits in brain allows detection of small
predominantly hyperintense tissue on T2-weighted images. calcifications and foci of hemorrhage that may help in the
Parenchymal territories involved in ischemic infarction differential diagnosis of various brain pathologies including
show increased signal intensity on T2-weighted and FLAIR tumors, vascular malformations, and vascular pathologies.
images starting in the acute phase and gradually increasing
as the necrotic process evolves. Intracranial hemorrhages 15.4.1.3 S ignal Alterations in Diffusion-
from various causes, both intraparenchymal and extra-axial, Weighted Images
appears predominantly hyperintense on T2-weighted images Diffusion-weighted imaging is an indispensable technique
in the subacute stage (Fig. 15.2). to reach the correct diagnosis in specific brain diseases. The
Disorders primarily affecting white matter, most notably presence of restricted diffusion in acute ischemic stroke is
small vessel disease, inflammatory and demyelinating disor- produced by intracellular cytotoxic edema and allows early
ders, and leukoencephalopathies, present as areas of high sig- recognition and estimation of the infarcted area; diffusion
nal intensity on T2-weighted and FLAIR images with variable then slowly normalizes after the first week. Bacterial pyo-
morphology and spatial distribution and potentially overlap- genic abscesses can be distinguished from other fluid-
ping features. The interpretation of imaging findings in the containing masses, such as necrotic tumors, because
clinical context is required to reach the correct diagnosis. purulent material shows significantly restricted diffusion
A rather limited number of diseases can show a prevalent (Fig. 15.6), while colliquative necrosis resembles clear
low signal intensity on T2-weighted images, including: fluid. Other infectious diseases such as viral encephalitis
mainly caused by herpes simplex virus can present with
–– Acute hematomas and masses complicated by restricted diffusion that typically involves the cerebral cor-
hemorrhage. tex of the temporal lobe and insular region. Epidermoid
cysts can be distinguished from more common arachnoid appear as ring-enhancing lesions due to blood-brain barrier
cysts because they show restricted diffusion due their con- disruption at the border between hematoma and normal
tent. Some brain tumors may show restricted diffusion parenchyma.
because of a high cellularity, notably lymphomas, that Tumefactive demyelination is a particular imaging pre-
are usually located supratentorially in the periventricular sentation of demyelinating diseases such as multiple sclero-
region, and medulloblastomas, that typically occur in poste- sis, appearing as a ring-enhancing lesion. Useful imaging
rior fossa near the fourth ventricle. Acute and subacute findings that can help in reaching the correct diagnosis are
hematomas may show a variable appearance on diffusion- the presence of an incomplete enhancing rim, little mass
weighted images; nevertheless a reduction of ADC values is effect on surrounding structures relative to its size, and a lim-
a common finding in early stages of hematoma evolution ited perilesional vasogenic edema.
(Fig. 15.2) and it has to be considered in the differential A common pattern found in metastatic disease besides
diagnosis of diffusion-restricting lesions. rim enhancement is nodular subcortical enhancement, typi-
cal of small lesions that appear as homogeneously enhancing
15.4.1.4 C ommon Patterns of Contrast nodules. The distribution of nodular enhancement depends
Enhancement on the preferential dissemination pathways of primary tumor;
An intact blood-brain barrier impedes significant diffusion of metastases that spread via the hematogenous arterial route
gadolinium in brain parenchyma with the exception of some tend to localize in supratentorial gray-white matter junction,
anatomical regions where this specialized structure is nor- while metastases spreading through paravertebral venous
mally absent, notably hypophysis, pineal gland, choroid plexus preferentially involve posterior fossa structures like
plexuses, hypothalamic infundibulum, and area postrema. cerebellum and brain stem.
Meningeal layers are extra-axial structures that do not sig- Primary brain tumors may show variable contrast
nificantly enhance in normal conditions. enhancement depending on pathological features. In most
Various diseases are characterized by the presence of solid tumors of glial origins contrast enhancement tends to
contrast enhancement that mainly results from a limited correlate with pathological grade, with some notable excep-
number of mechanisms including passage of gadolinium in tions such as oligodendroglioma. Typically enhancing
brain parenchyma due to blood-brain barrier disruption, tumors are represented by lymphoma, meningioma, schwan-
neovascularization as seen in tumors, and increased capil- noma, intraventricular neoplasms like ependymoma and
lary permeability in meningeal pathologies. Different dis- choroid plexus papilloma, and tumors arising in pituitary and
orders tend to be associated with particular patterns of pineal glands.
contrast enhancement that must be recognized to reach The presence of strong nodular enhancement in the wall
the correct diagnosis. of predominantly cystic masses is the typical pattern of
Gyriform enhancement appears as a “serpentine” hyper- low-grade fluid-secreting tumors including pilocytic astro-
intensity that involves gray matter in cerebral cortex. This cytoma and hemangioblastoma, most commonly located in
pattern is caused by a damaged blood-brain barrier and is cerebellar hemispheres, and pleomorphic xanthoastrocy-
typically found in vascular diseases, such as ischemic cere- toma and ganglioglioma that are commonly
bral infarction in the subacute phase and posterior reversible supratentorial.
encephalopathy syndrome, and in infectious processes like Enhancement of periventricular subependymal layer can
viral encephalitis and meningitis, the latter in association be found in primary central nervous system lymphoma and
with the classical leptomeningeal enhancement. high-grade glial tumors, the former typically showing intense
Ring enhancement, represented by a rim of contrast and often homogeneous enhancement.
enhancement surrounding a central region, is the typical Extra-axial diseases involving meningeal layers can show
pattern of cerebral abscesses (Fig. 15.6), some primary specific patterns of enhancement that can prove useful in
tumors such as high-grade gliomas (Fig. 15.5) and lympho- narrowing the differential diagnosis. Leptomeningeal
mas in the immunocompromised patient, and metastatic enhancement, consisting of hyperintensity of pia-arachnoid
neoplasms. Diffusion-weighted imaging allows to reliably complex that follows along the surface of the brain, is typi-
differentiate pyogenic infectious from neoplastic causes. cally seen in diseases spreading in the subarachnoid space,
Rim-enhancing lesions are often associated with perile- such as infectious and carcinomatous meningitis.
sional vasogenic edema and mass effect. Subacute basal Leptomeningeal enhancement is commonly associated with
ganglia infarctions and subacute intraparenchymal hemor- high signal intensity of subarachnoid space on FLAIR
rhagic collections, both traumatic and non-traumatic, can images.
Pachymeningeal or dural enhancement involves the 15.4.2 Neck Masses

dura, the outermost layer of meninges surrounding the
central nervous system that forms the tentorium cerebelli 15.4.2.1 Cystic Lesions
and falx cerebri. Tumors of meningeal origin like menin- Cystic masses are frequently encountered lesions in neck
giomas typically appear as intensely enhancing extra-axial imaging. Most cysts, both congenital and acquired, share
masses associated with circumscribed area of dural similar imaging features, and etiology can be suggested by
enhancement (“dural tail”), in continuity with the tumor. the specific anatomical location in deep neck spaces.
Dural involvement by metastatic carcinoma and primarily Thyroglossal duct cyst is the most common developmental
or secondarily disseminated lymphoma often produces a lesion located in anterior neck that can show variable signal
diffuse irregular enhancement that may coalesce to form intensity on T1-weighted images due to proteinaceous con-
large enhancing masses. Less common tumors of intra- tent. The presence of intracystic solid tissue in a thyroglossal
axial origin abutting pial surface, such as pleomorphic duct cyst should suggest the development of malignancy.
xantoastrocytoma, can be associated with focal dural Second branchial cleft cysts are the most frequent bran-
enhancement. chial cleft anomalies and are typically located in the lateral
Dural involvement may also be found in granulomatous aspect of the neck, often anterior or medial to sternocleido-
disease of both infectious (e.g., tuberculosis) and immune- mastoid muscle. Laryngoceles can appear as completely cys-
mediated (e.g., sarcoidosis) etiology, typically appearing tic lesions located in the supraglottic larynx.
as a pachymeningeal enhancement with a predominantly Lymphangiomas are developmental cystic masses often
basal location that may be diffused or localized, some- originating in the posterior triangle, but they can extend in
times forming dural masses with a plaque-like other adjacent neck spaces when of large size.
morphology. Cystic masses in salivary glands are mainly represented
by ranula or sublingual gland mucocele and, in the parotid
gland, by various congenital and acquired cysts, the latter
mainly represented by mucocele and cystic carcinomas.
Tumors that usually appear as solid masses may some-
Key Learning Points times appear as partially or predominantly cystic lesions
• T1 hyperintense brain alterations are mainly repre- because of intralesional necrosis, hemorrhage, or cystic
sented by fat-containing lesions (e.g., lipoma, der- changes; they are more frequently represented by neurogenic
moid), proteinaceous fluid-containing lesions (e.g., tumors (schwannoma and neurofibroma) and cystic lymph
colloid cyst, craniopharyngioma), and node metastases, typically from papillary thyroid cancer.
subacute hematomas. Necrotic lymphadenopathy with predominant fluid com-
• Most brain pathologies appear as hyperintense ponents can also be caused by infectious diseases, notably
alterations on T2-weighted and FLAIR images. T2 tuberculosis. Abscesses, more commonly located in subman-
hypointense brain alterations are mainly repre- dibular or retropharyngeal spaces, appear as fluid-containing
sented by acute hematomas, highly proteinaceous lesions with a peripheral rim of contrast enhancement, simi-
fluid-containing lesions (e.g., colloid cyst), and lar to those found in other anatomical sites.
densely cellular tumors (e.g., lymphoma,
medulloblastoma). 15.4.2.2 Solid Masses
• Hemorrhage can be readily detected on gradient- The most common solid masses occurring in the neck are
echo T2*-weighted images. represented by lymphadenopathies that may have variable
• Pyogenic abscesses can be distinguished from causes, including metastatic, lymphomatous, and infectious.
necrotic neoplasms due to the restricted diffusion of Both normal and pathologically enlarged non-necrotic
purulent material. lymph nodes have widely overlapping signal features: mod-
• Densely cellular tumors (e.g., lymphoma, medullo- erate hyperintensity on T2-weighted and mild hypointensity
blastoma) may show restricted diffusion. on T1-weighted images often with restricted diffusion.
• The specific pattern of contrast enhancement may Imaging findings typically used to differentiate a solid
help in the differential diagnosis. malignant lymphadenopathy from a reactive enlarged lymph
• Rim enhancement is mainly seen in necrotic neo- node are based on morphologic criteria with limited diag-
plasms, both primary and secondary, and infections nostic accuracy, including loss of normal architecture with
such as pyogenic abscesses. obliteration of the fatty hilum, round shape, irregular corti-
cal thickening, and inhomogeneous signal. Necrotic lymph
a b
Fig. 15.7 Paraganglioma of the left carotid space. Fat-suppressed T2-weighted image with small low signal foci likely representing flow
T2-weighted images (a) and contrast-enhanced images (b) are shown. voids and shows strong contrast enhancement
The mass (white arrowhead) appears predominantly hyperintense on
nodes are characterized by the presence of internal fluid-like masses, primary head and neck cancers, and soft tissue
non-enhancing components. tumors.
Uncommon solid masses that must be distinguished
from lymphadenopathies are neurogenic tumors such
as peripheral nerve sheath tumors, including schwannoma
Key Learning Points
and neurofibroma, and paraganglioma. Peripheral nerve
sheath tumors can have a variable appearance, from homo- • Cystic lesions of the neck can be readily distin-
geneous T2-hyperintense and moderately enhancing lesions guished from lymphadenopathies and other solid
to inhomogeneous masses with areas of myxoid degenera- masses on MR images.
tion, cystic changes, and hemorrhage, the latter more fre- • Imaging features correlated with the anatomical
quently seen in schwannomas. These tumors, typically location are the mainstay of diagnosis.
located in the carotid space where they may displace the • Schwannomas and neurofibromas usually appear as
carotid artery and the jugular vein, cannot be reliably dis- hyperintense on T2-weighted images, moderately
tinguished from each other on the basis of imaging enhancing after contrast administration. They may
characteristics. exhibit myxoid degeneration, cystic changes, and
Paraganglioma is classically located at the carotid bifur- hemorrhage.
cation (carotid body tumor) and usually causes splaying of • A typical imaging feature of paraganglioma on
internal and external carotid arteries (Fig. 15.7). T2-weighted images is the presence of intralesional
Paragangliomas can be recognized by their intense vascular- flow voids giving a “salt-and-pepper” appearance.
ity with higher contrast enhancement compared with schwan-
noma and neurofibroma. A typical imaging feature of
paraganglioma on T2-weighted images is the presence of
intralesional flow voids giving a “salt-and-pepper” 15.4.3 Evaluation of Mediastinum and Pleura
appearance.
MR imaging plays an important role in the evaluation of 15.4.3.1 Mediastinal Masses
many other types of mass lesions occurring in the neck for Mediastinal cystic lesions can be easily recognized by their
both characterization and staging purposes, including parotid smooth margins, fluid content, and thin walls. Pericardial
a b
c d
Fig. 15.8 Large B-cell lymphoma with mediastinal involvement. infiltration of the anterior chest wall on the right side (white arrowhead),
T2-weighted image (a), contrast-enhanced image (b), b-750 diffusion- and a right axillary lymphadenopathy can be seen. The lymphoprolifera-
weighted image (c), and ADC maps (d) are shown. Multiple coalescing tive tissue exhibits intermediate signal on T2-weighted image, slightly
lymphadenopathies that encase mediastinal vessels, lymphomatous inhomogeneous contrast enhancement, and restricted diffusion
cysts present as unilocular cystic lesions in contact with the appear as solid masses often with smooth or lobulated con-
pericardium, frequently located in right cardio-phrenic tours that may encase vascular structures, with appearance
angle. Bronchogenic cysts, usually located near the carina, that can range from homogeneous signal and enhancement
and duplication cysts, often in contiguity with the esopha- to inhomogeneous masses with areas of necrosis. Lymphomas
geal wall, present as simple cystic lesions unless complicated are frequently associated with mediastinal lymphadenopathy
by hemorrhage or infection. (Fig. 15.8).
The most common fat-containing anterior mediastinal Posterior mediastinal masses located in the paravertebral
mass is lipoma that shows characteristic imaging features. region are most commonly represented by neurogenic
Teratoma, a benign germ cell tumor, typically located in tumors, including peripheral nerve sheath tumors (schwan-
the anterior mediastinum, presents as complex solid-cystic noma and neurofibroma), tumors originating from sympa-
mass with intralesional fat, sometimes with fat-fluid levels thetic ganglions (ganglioneuroma and neuroblastoma), and
within loculations. Malignant germ cell tumors appear as paraganglioma. Mediastinal peripheral nerve sheath tumors
solid masses with lobulated or irregular margins that can show the same imaging features as their counterparts occur-
have relatively homogeneous (as in seminomas) or markedly ring in the neck. Ganglioneuromas typically appear as inho-
inhomogeneous signal and contrast enhancement (as in non- mogeneous hyperintense masses on T2-weighted images
seminomatous tumors). with hypointense curvilinear bands of fibrous tissue; neuro-
Anterior mediastinum is a common site of involvement blastomas appear as heterogeneously enhancing masses
for lymphoproliferative disorders. Lymphomas typically often with necrotic areas. Paragangliomas often present as
a b
Fig. 15.9 Thymic hyperplasia. T1-weighted in-phase image (a) and out-of-phase image (b) are shown. Hyperplastic thymus loses signal on
T1-weighted out-of-phase image (white arrowhead) owing to microscopic fat interspersed between normal thymic tissues
vertically oriented hypervascularized masses with the same 15.4.3.3 Pleural Disease
signal characteristics as described for carotid body tumors. Magnetic resonance can be a useful adjunct to other diagnostic
Mediastinal lymphadenopathies commonly appear as solid imaging modalities for evaluation of pleural diseases thanks to
enhancing masses with increased signal intensity on diffu- its superior contrast resolution. Pleural thickening, especially if
sion-weighted images (Fig. 15.8). Intranodal fluid-like areas it is mild and not circumferential, is an unspecific finding that
are characteristic of necrotic lymph nodes as can be found in can be seen in both benign and malignant diseases. Malignant
other anatomical sites. pleural mesothelioma is a relatively rare primary tumor of
the pleura associated with asbestos exposure. In patients with
15.4.3.2 Thymic Enlargement asbestos-related pleural disease, thickening of the pleura due to
Thymic hyperplasia, both true epithelial hyperplasia and fibrotic plaques or chronic pleuritis often associated with pleu-
lymphoid “rebound” hyperplasia, commonly presents as a ral effusion is a frequent occurrence. Distinguishing benign
diffuse and symmetric enlargement of thymus. In some pleural disease from a malignant tumor can be a difficult task in
cases, enlarged thymus may appear asymmetric or with lobu- the early stages of mesothelioma. On T2-weighted images,
lated contours, thus raising the suspicion of neoplastic dis- pleural plaques typically show low signal intensity due to abun-
ease such as thymic epithelial neoplasia (e.g., thymoma) or dant collagenous fibrous tissue, often associated with calcifica-
lymphoma. Normal and hyperplastic thymus typically con- tions, while mesothelioma usually appears slightly to
tains microscopic fat interspersed within thymic tissue that moderately hyperintense. Diffusion-weighted images may
causes signal loss in dual-phase chemical shift imaging allow a more accurate diagnosis of pleural mesothelioma in
(Fig. 15.9); this feature can be helpful to differentiate hyper- high-risk patients as thickened pleura in mesothelioma usually
plastic enlargement from neoplastic lesions. Congenital thy- shows areas of restricted diffusion (Fig. 15.10); this appearance
mic cysts commonly appear as simple cystic lesions and is not specific for malignant mesothelioma and can also be
sometimes may show T1-hyperintensity because of protein- found in metastatic pleural disease.
aceous or hemorrhagic content. Acquired thymic cysts may
have multilocular appearance with thin internal septations.
The presence of thick septa or solid components within a
cystic mass should suggest a cystic thymoma. The presence Key Learning Points
of abundant fat with minor solid components in a thymic • Mediastinal cystic lesions can be readily distin-
mass is characteristic of benign thymolipoma. Thymic epi- guished from lymphadenopathies and other solid
thelial tumors have variable appearance that can range from masses on MR images.
round or lobulated homogeneously solid enhancing masses • Imaging features correlated with the anatomical
to irregularly shaped infiltrative tumors with inhomogeneous location are the mainstay of diagnosis.
signal and contrast enhancement. Thymic lymphoma may • Thymic hyperplasia can be differentiated from
result from secondary involvement or more rarely as an iso- other causes of thymic enlargement due to the pres-
lated site.
a b
Fig. 15.10 Left malignant pleural mesothelioma. Contrast- edly elevated signal intensity on diffusion-weighted image indicating
enhanced image (a) and b-750 diffusion-weighted image (b) are shown. restricted diffusion
The irregular pleural thickening shows contrast enhancement and mark-
Acute inflammatory diseases of different etiology tend to

ence of microscopic fat detectable on dual-phase show similarly unspecific imaging findings and their diagno-
chemical shift images. sis requires interpretation in the clinical context. In contrast
• Malignant pleural mesothelioma typically appears enhanced images acquired during the hepatic arterial phase,
as pleural thickening with restricted diffusion. heterogeneous parenchymal enhancement can be seen with
areas of transitory enhancement that typically fade to isoin-
tensity compared to liver parenchyma in portal venous and
equilibrium phase images. On T2-weighted images, hyper-
15.4.4 Evaluation of Abdomen and Pelvis perfused areas may show a slightly increased signal intensity
that is attributable to parenchymal inflammatory infiltrates.
15.4.4.1 D iffuse Parenchymal Alterations In conditions characterized by diffuse liver inflammation as
of the Liver found in acute hepatitis, a periportal edema with high signal
Hepatic steatosis, also called fatty liver disease, is a very intensity of portal spaces on T2-weighted images is com-
common pathologic condition consisting of an abnormal mon. In more focal processes like cholangitis, wedge-shaped
increase in the amount of triglyceride molecules stored areas of parenchymal inflammation surrounding dilated
within intracellular vacuoles in hepatocytes. In most cir- intrahepatic biliary ducts are typical findings.
cumstances, steatosis diffusely involves the liver paren- In chronic inflammatory diseases evolving toward liver
chyma, although inhomogeneous fat accumulation with cirrhosis, imaging findings indicate the presence of paren-
areas of relatively spared parenchyma is a frequent occur- chymal fibrosis. Cirrhotic liver shows nodular contours,
rence, sometimes resulting in a patchy appearance. Less irregular shape with variable involvement of different liver
common variants of liver steatosis are represented by segments, often with relative hypertrophy of the caudate and
preferential fat deposition in specific liver territories, typ- left lobes, and an inhomogeneous parenchymal pattern due
ically in the right lobe and rarely in a subcapsular or peri- to fibrotic tissue deposition delineating regenerative nodules.
vascular distribution. Occasionally, fatty liver disease Fibrosis is better detected in delayed contrast-enhanced
may mimic a focal liver lesion either because steatosis images acquired in equilibrium phase (3–5 min after contrast
presents as a circumscribed signal alteration, or injection) when the gadolinium-based contrast agent diffuses
because the liver parenchyma is focally spared. Imaging in the interstitial space and accumulates in fibrous tissue
findings in fatty liver disease are consistent with the pres- resulting in high signal intensity compared to normal liver
ence of intracellular fat in affected areas, which is readily parenchyma. In cases of mild fibrosis, liver parenchyma
detectable on dual-phase chemical shift imaging assumes a granular texture attributable to micronodular
(Fig. 15.11). Focal fat deposition typically appears as a regenerative hyperplasia; with more severe involvement, a
sharp defined area, with no mass effect on adjacent nor- coarser nodular pattern with linear strands of fibrosis and
mal vascular and biliary structures. large regenerative nodules can be observed. In some circum-
a b
Fig. 15.11 Diffuse hepatic steatosis. T1-weighted in-phase image (a) and out-of-phase image (b) are shown. Liver parenchyma demonstrates
marked signal loss on dual-phase imaging due to increased intracellular fat deposition with some areas of fatty sparing (white arrowheads)
stances, fibrotic tissue completely substitutes areas of liver walls, and sometimes mural nodules that enhance after gad-
parenchyma and can appear as a focal lesion (so-called con- olinium administration; the fluid content can have variable
fluent fibrosis); useful imaging findings that can suggest a signal intensity on T1-weighted images reflecting the pres-
fibrous composition are wedge-shaped margins, preferential ence of proteinaceous, mucinous, or hemorrhagic compo-
subcapsular location, typical delayed contrast enhancement, nents. Solid papillary projections and extensive solid
and mildly to moderately increased signal intensity on components are more frequently observed in malignant
T2-weighted images compared to parenchyma probably due cystadenocarcinomas.
to vascular congestion or edema. Liver abscesses can be unilocular, as more frequently
seen in amebic etiology, or have a clustered multilocular
15.4.4.2 Focal Liver Lesions appearance generated by the coalescence of multiple fluid-
containing areas, more suggestive of pyogenic abscess.
Cystic Lesions Abscesses usually show thick enhancing walls and may be
The most common hepatic cystic lesion is represented by associated with perilesional edema appearing as high signal
simple cyst, a developmental lesion that does not commu- intensity on T2-weighted images; the purulent fluid content
nicate with the biliary system. Simple cysts can be single can show restricted diffusion. Hydatid disease in its initial
or multiple and are readily recognized because of their stage can present as a unilocular non-enhancing cystic
unilocular structure, clear fluid content, round or some- lesion indistinguishable from simple liver cysts. More typi-
times lobulated shape with well-delineated margins, and cally, hydatid cysts demonstrate a complex internal struc-
absence of enhancement after contrast administration. ture with a thickened pericyst and multiple daughter cysts
Intracystic hemorrhage is a very rare complication in sim- or intracystic floating membranes (classical “water lily”
ple liver cysts. Autosomal dominant polycystic liver dis- sign).
ease, often associated with cystic kidney disease, appears Other focal liver lesions can sometimes appear as pre-
as multiple cysts that more frequently show a complicated dominantly or partially cystic. Hepatic benign and malignant
appearance. Developmental lesions with similar appear- masses such as hemangioma and hepatocellular carcinoma
ance are biliary hamartomas, small nodules that do not can rarely undergo intralesional colliquation or cystic degen-
communicate with the biliary tree and typically present as eration. Similarly, liver metastases from certain primary
multiple cysts less than 1 cm in size (von Meyenburg com- tumors may assume a cystic appearance owing to central
plexes) that may occasionally show a thin rim of contrast necrosis or cystic degeneration, such as in melanoma and
enhancement. neuroendocrine tumors, or due to a mucinous histology,
A multilocular cystic appearance is typically found in more frequently seen in colorectal and ovarian cancer.
rare neoplasms of biliary origin (cystadenoma and cystade- Nevertheless, liver cystic metastases often show a thin rim of
nocarcinoma), pyogenic abscesses, and hydatid disease. peripheral enhancement or restricted diffusion that is not
Biliary cystadenomas usually show internal septa, thickened consistent with the diagnosis of simple hepatic cyst.
a b c
Fig. 15.12 Liver hemangioma. T2-weighted image (a), contrast- (white arrowhead) exhibits markedly high signal intensity on
enhanced late arterial image (b), and portal-venous image (c) are T2-weighted image and early peripheral nodular enhancement with
shown. A well-defined lobulated mass in the right lobe of the liver centripetal nearly complete fill-in, consistent with hemangioma
Hepatic Benign Tumors marked arterial enhancement, with the exception of the
Besides simple cysts that are the most frequently encoun- central scar, and prevalent isointensity to liver parenchyma
tered benign liver focal lesions, some common benign liver on portal venous and equilibrium phases with progressive
tumors can be readily recognized on MRI when they pres- enhancement of the central fibrous scar. Because of the
ent with typical imaging findings. Hemangiomas are benign presence of functioning hepatocytes and the lack of nor-
masses often incidentally discovered that show very high mal biliary canaliculi, FNH tends to accumulate hepato-
signal intensity on T2-weighted images, almost reaching the specific gadolinium-based contrast agents in delayed
signal of clear fluid due to the presence of dilated vessel hepatobiliary phase often resulting in hyperintensity com-
filled with slow-flowing blood (Fig. 15.12). On T2-weighted pared to liver parenchyma (Fig. 15.13).
images with long TE, hemangiomas maintain their signal Other focal liver lesions that should be included in the
hyperintensity, while more solid lesions such as metastases differential diagnosis of suspected FNH are benign and
tend to lose signal compared to fluid-containing structures. malignant masses with hypervascular appearance, such as
After contrast administration hemangiomas typically dem- flash-filling hemangioma, hepatic adenoma, hepatocellular
onstrate nodular peripheral enhancement with progressive carcinoma, and hypervascular liver metastases. Among
centripetal fill-in that may be incomplete due to central these pathological entities, hepatic adenoma and fibrolamel-
areas of scarring or cystic degeneration. The enhancing por- lar variant of hepatocellular carcinoma may show partially
tions of hemangiomas reach the same signal intensity as in overlapping imaging features with FNH. Adenoma may
liver vessels, a useful finding to distinguish hemangioma demonstrate a similar appearance on dynamic contrast-
from malignant lesions that can show a similar centripe- enhanced images with hypervascularity in arterial phase,
tal fill-in after contrast administration. Some small heman- although less intense and homogeneous than FNH and with-
giomas do not exhibit the typical nodular centripetal out a central scar, and most frequently appears isointense
enhancement but rather a complete homogeneous hyperin- compared to liver parenchyma in portal venous and equilib-
tensity (so-called “flash-filling” hemangioma). This variant rium phase. The most useful imaging feature to differentiate
of hemangioma may occasionally pose a differential diag- FNH from adenoma is the absence of hepatospecific con-
nostic problem, since it can appear similar to other focal trast accumulation on the delayed phase images due to the
liver lesions, such as homogeneously enhancing hypervas- lack of excretory function resulting in low signal intensity.
cular liver metastases. In addition, adenomas may show variable signal intensity on
Focal nodular hyperplasia (FNH) is the second most T1- and T2-weighted images because of excessive intracel-
common benign liver lesion. It consists of regenerative lular glycogen storage (mild to moderate T1 hyperinten-
hepatic tissue that in most cases manifests as a hypervas- sity), intracellular fat deposition (recognizable on dual-phase
cular mass with circumscribed margins. A typical imaging chemical shift images), and intralesional hemorrhage (sig-
feature is the presence of a prominent central scar with nal abnormalities consistent with the presence of blood
fibrous septations and a radiating pattern. FNH usually components).
appears slightly hypointense compared to liver paren-
chyma with a central scar of low signal intensity on Hepatic Malignant Tumors
T1-weighted images and nearly isointense with a hyperin- Liver metastases are the most frequently encountered malig-
tense central scar on T2-weighted images. After con- nant focal solid masses. Metastases often demonstrate a
trast administration, FNH classically demonstrates a mildly to moderately elevated signal intensity on T2-weighted
a b
c d
Fig. 15.13 Focal nodular hyperplasia. T2-weighted image (a), chyma on T2-weighted image, diffuse arterial enhancement with thin
contrast-enhanced late arterial image (b), portal-venous image (c), and hypointense septa, mild hyperintensity on portal-venous phase, and
delayed hepatobiliary image (d) are shown. A solid nodule in the right increased hyperintensity on delayed images
lobe (white arrowhead) exhibits nearly isointense signal to liver paren-
a b
Fig. 15.14 Liver hypovascular metastases from colorectal adenocarci- contrast enhancement compared to liver parenchyma with central
noma. T2-weighted image (a) and contrast-enhanced portal- necrotic changes appearing as high signal intensity areas on T2-weighted
venous image (b) are shown. Multiple liver masses exhibit lower image (white arrowhead)
images, which is a useful feature to distinguish metastases enhancing less than the surrounding hepatic parenchyma
from liver cysts and hemangiomas. Larger lesions can show (Fig. 15.14). Metastases from hypervascular neoplasms,
a heterogeneous appearance on T2-weighted images due to including neuroendocrine tumors, renal cell carcinoma, mel-
necrotic changes or cystic degeneration (Fig. 15.14). On anoma, thyroid cancer, and sometimes breast cancer, often
T1-weighted images, metastases are usually hypointense demonstrate intense enhancement in the arterial phase
compared to liver parenchyma with some exceptions, such as images. Small metastases may show a rather homogeneous
hemorrhagic metastases or melanoma metastases that arterial
enhancement and resemble other hypervascular
show the presence of intracellular paramagnetic substances. lesions both benign, especially flash-filling hemangioma,
Diffusion-weighted imaging has a high sensitivity for the and malignant such as hepatocellular carcinoma. A contrast
detection of liver metastases and can be useful for lesion enhancement pattern considered typical for both hypervas-
characterization despite considerable overlap in diffusion cular and hypovascular liver metastasis is represented by a
features between solid hepatic lesions, both benign and peripheral rim of enhancement that can persist or wash out in
malignant. In dynamic contrast-enhanced images, liver portal venous or equilibrium phases. The central portion of
metastases can have a variable appearance depending on metastases may show progressive filling with accumulation
lesion size and pathological features of the primary tumor. of contrast that occasionally reach isointensity compared to
Most liver metastases appear as hypovascular focal masses, liver parenchyma although a relative hypointensity is more
a b
c d
Fig. 15.15 Hepatocellular carcinoma. T2-weighted image (a), image, arterial enhancement with contrast washout and a thin pseudo-
contrast-enhanced later arterial image (b), and portal-venous image (c) capsule in the portal-venous phase, and hypointensity to liver paren-
images are shown. A round mass in the right lobe of the liver (white chyma in the delayed phase
arrowhead) shows moderately high signal intensity on T2-weighted
commonly found. Unlike hemangiomas, metastases demon- extrahepatic cholestasis. Mass-forming cholangiocarcinoma
strate a c ontinuous peripheral enhancement that may prog- typically presents as a large lesion with mildly to moderately
ress centrally with a lower signal intensity compared to elevated signal intensity on T2-weighted images and low sig-
blood pool. Larger metastases can show more complex nal intensity on T1-weighted images. After contrast adminis-
appearances: as an example, large hypervascular metastases tration, the most frequent finding is an irregular peripheral rim
can have central areas of necrosis or can show a heteroge- of enhancement with progressive centripetal fill-in that can
neous arterial enhancement that may resemble hepatocellu- result in hyperintensity compared to liver parenchyma in equi-
lar carcinoma, while hypovascular metastases, typically librium phase due to abundant fibrous tissue (Fig. 15.16), not
from colorectal cancer, may present with scalloped margins to be mistaken for the discontinuous nodular enhancement
and internal enhancing fibrous septations that progressively pattern typically seen in hemangiomas. In the delayed hepato-
enhance (so-called “cauliflower” appearance). biliary phase cholangiocarcinomas most commonly exhibit
The most frequent primary malignant tumor of the liver is low signal intensity; occasionally, the more fibrotic central
hepatocellular carcinoma (HCC) that is strongly associated portion of the tumor can reach isointensity relative to the liver
with chronic liver disease and cirrhosis from various causes due to diffusion and retention of gadolinium in the expanded
including viral, exotoxic, autoimmune, and genetic diseases. interstitial space.
The typical imaging findings of HCC include a
T2-hyperintense mass with arterial enhancement that can be
homogeneous in small HCC and markedly heterogeneous in
larger tumors, often showing a mosaic-like pattern. On post-
contrast images acquired in portal venous and equilibrium Key Learning Points
phases, HCC usually appears hypointense compared to liver • Fatty liver disease results in signal loss on dual-
parenchyma as a result of washout, often presenting an phase chemical shift images.
enhancing surrounding pseudocapsule constituted by com- • Liver acute inflammatory diseases may manifest
pressed liver parenchyma (Fig. 15.15). In less typical cases, with areas of hyperintensity on T2-weighted images
HCC does not show significant washout of contrast and can and transient parenchymal enhancement in the late
appear very similar to other benign and malignant hypervas- arterial phase.
cular lesions. In delayed hepatobiliary phase after adminis- • Liver fibrosis is better appreciated on contrast-
tration of hepatospecific contrast agents, HCC typically enhanced images acquired in the equilibrium phase.
exhibits a lower signal intensity compared to liver paren- • Simple liver cysts are frequently occurring focal
chyma. Nevertheless, a minority of well-differentiated lesions. Multilocular cysts are mainly represented
tumors can uptake and accumulate contrast with consequent by biliary cystadenoma/cystoadenocarcinomas,
hyperintensity on delayed phase images and may be mis- pyogenic abscess and hydatid disease.
taken for FNH. Fibrolamellar carcinoma has traditionally • Liver abscess usually shows restricted diffusion and
been considered a pathologic variant of hepatocellular carci- peripheral enhancement.
noma although it probably represents a distinct clinical • Liver hemangiomas often show typical imaging
entity. Fibrolamellar carcinoma tends to appear as a large features (marked T2 hyperintensity and globular
heterogeneous hypervascular mass with a central scar that centripetal enhancement).
shows low signal intensity on T2-weighted images, in part • Hypervascular benign liver lesions are mainly rep-
attributable to calcific deposits. This imaging feature may be resented by “flash-filling” hemangiomas, focal nod-
helpful to distinguish fibrolamellar carcinoma from FNH, ular hyperplasia (FNH), and hepatic adenomas.
the latter usually showing a T2 hyperintense scar. On post- FNH often shows typical imaging features with
contrast images acquired on portal venous and equilibrium hyperintensity on delayed images obtained with
phases, fibrolamellar carcinoma may appear nearly isoin- hepatospecific contrast agents.
tense compared to liver parenchyma or demonstrate washout • Hepatocellular carcinoma presents as a hypervascu-
resulting in hypointensity; the assessment of contrast lar solid lesion.
enhancement on the delayed phase images typically shows • Liver metastases often show a rim of contrast
lack of contrast uptake with predominant hypointensity, enhancement on arterial phase images with wash-
opposite to FNH that characteristically appears out in the portal venous and equilibrium phases.
iso-hyperintense. • Cholangiocarcinoma often shows a peripheral rim
Cholangiocarcinoma is the second more common tumor of of contrast enhancement with progressive centripe-
the liver. It can present as an intrahepatic mass-forming lesion, tal fill-in that must not be mistaken for the nodular
a papillary intraductal growth in the biliary tree, or a periduc- enhancement of hemangiomas.
tal-infiltrating tumor causing biliary strictures with consequent
a b
c d
Fig. 15.16 Mass-forming cholangiocarcinoma. T2-weighted (a) and contrast enhancement on late arterial phase, and progressive fill-in on
contrast-enhanced images acquired in late arterial phase (b), equilib- equilibrium phase image. In the delayed phase image, the mass appears
rium phase (c), and delayed hepatobiliary phase (d) are shown. A mass mainly isointense to liver parenchyma. A thin hypointense rim sur-
is seen in the right lobe of the liver (white arrowhead). The lesion exhib- rounding the mass and the retraction of liver margin are useful signs
its moderate hyperintensity on T2-weighted image, a continuous rim of to reach the correct diagnosis
15.4.4.3 Pancreatic Masses serous cystadenoma, a benign multicystic mass often located
in pancreatic head; another classical feature of serous cyst-
Cystic Lesions adenoma is the presence of a central fibrotic scar that may
The most common cystic lesions of the pancreas are pseu- calcify. Less frequently encountered variants are oligocystic
docysts occurring as a complication of pancreatitis. serous cystadenoma that may resemble mucinous cystade-
Pseudocysts often appear as unilocular well-circumscribed noma and the pseudosolid variant that can mimic a solid
cystic masses with fluid content and no solid enhancing neoplasm. Mucinous cystadenomas, typically located in the
components with the possible exception of a thin enhancing body and tail of the pancreas, frequently appear as unilocu-
pseudocapsule. Pseudocysts can be distinguished from lar cysts similar to inflammatory pseudocysts. Sometimes
walled-off necrosis, a complication of necrotizing pancre- intracystic septa may be present giving a multilocular
atitis, because the latter contains necrotic debris that can appearance with fewer large cystic spaces compared to cyst-
deposit in the dependent portion of the fluid cavity, appear- adenomas; signal intensity of fluid in loculations can appear
ing as low signal intensity material on T2-weighted images. similar to simple cysts or show high signal intensity on
The most common cystic pancreatic masses of neoplastic T1-weighted images due to the mucinous content. Mucinous
origin are serous cystadenoma, mucinous cystadenoma and cystadenoma is considered a borderline lesion with malig-
cystadenocarcinoma, and intraductal papillary mucinous
nant potential; its malignant counterpart, mucinous cystad-
neoplasm. A multilocular microcystic lesion that can show a enocarcinoma, can be distinguished by the presence of solid
“honeycomb” appearance is the typical presentation of enhancing tissue.
Intraductal papillary mucinous neoplasms (IPMN) are between tumor and pancreatic parenchyma on T1-weighted
mucinous neoplasms that grow inside the pancreatic ducts. precontrast and arterial phase postcontrast images often
IPMN can be classified into three types that carry a different allows the identification of small lesions that do not alter
prognostic significance: the main duct type, the branch duct the pancreatic contour and that may be undetectable with
type, and the mixed type (combined IPMN). Main duct other imaging modalities. For the same purpose, it is also
IPMN, frequently associated with malignant invasive carci- useful the evaluation of secondary signs that could suggest
noma, involves diffusely or in a segmental pattern the main the presence of an infiltrative process, such as abrupt
pancreatic duct that appears dilated often with lobulated con- obstruction of the main pancreatic duct with upstream duct
tours. On T2-weighted images, main duct IPMN typically has dilatation and the “double duct sign,” that refers to dilata-
a fluid-like appearance and may sometimes show intraductal tion of both biliary tree and main pancreatic duct caused by
dense mucinous material with low signal, while the presence tumors located in the pancreatic head (Fig. 15.17).
of solid enhancing nodules should suggest a malignant poten- Some benign diseases and pseudolesions may occasion-
tial. Branch duct IPMN commonly appears as single or often ally mimic a malignant mass on imaging, notably focal fatty
multiple round or lobulated cysts that directly communicate infiltration and mass-forming chronic pancreatitis; the for-
with the main pancreatic duct, better demonstrated on cholan- mer can be recognized by the loss of signal on dual-phase
giopancreatographic T2-weighted images. This is a useful chemical shift imaging, while the latter can be indistinguish-
feature in order to differentiate IPMN from other neoplastic able from tumor. It is reported that the presence of ductal
cystic lesions. Mixed type IPMN combines imaging features structures within a mass should suggest focal pancreatitis,
of both main duct and branch duct IPMN. Some solid tumors while abrupt duct termination at the border of the mass is
can uncommonly appear as predominantly or partially more consistent with adenocarcinoma; often only biopsy can
lesions, such as ductal adenocarcinoma with cystic degenera- make a definite diagnosis.
tion and cystic neuroendocrine neoplasms, the latter charac- Solid masses other than adenocarcinoma are infrequent
terized in the majority of cases by a thin peripheral rim of and are mainly represented by neuroendocrine tumors and
viable hypervascular tissue. metastases. Neuroendocrine tumors typically appear as sin-
gle or multiple hypervascular masses that enhance more than
Solid Masses pancreatic parenchyma in arterial phase, with variable
The most frequent solid pancreatic lesion is ductal adeno- appearance in portal and delayed phases. Rarely, neuroendo-
carcinoma. This tumor often appears as an infiltrating mass crine tumors show a predominantly cystic appearance usu-
that shows low signal compared to pancreatic parenchyma ally with a hypervascular enhancing rim. Pancreatic
on T1-weighted images and variable signal intensity on metastases can present as single or multiple masses with sig-
T2-weighted images, most often mild hyperintensity. After nal characteristics and vascularity reflecting primary tumor
contrast administration, ductal adenocarcinomas typically pathological features; as an example, melanoma and renal
exhibits hypovascular features on arterial phase with pro- cell carcinoma metastases more likely present as hypervas-
gressive enhancement in delayed phase related to the abun- cular lesions, while lung or colorectal cancer metastases are
dant fibrous component. The high contrast difference typically hypovascular.
a b
Fig. 15.17 Pancreatic ductal adenocarcinoma. T2-weighted image (a) hyperintensity on T2-weighted image and hypointensity on contrast-
and contrast-enhanced late arterial phase image (b) are shown. An ill- enhanced image. The common pancreatic duct distal to the mass
defined mass in the pancreatic body (white arrowhead) exhibits mild appears mildly dilated
commonly results in homogeneous enhancement in delayed

Key Learning Points phase. Hamartoma is a rare benign nonneoplastic tumor of
• Pancreatic cystic lesions are mainly represented by the spleen that classically appears as a solitary round-shaped
inflammatory pseudocysts, serous and mucinous solid mass that demonstrates low signal on T1-weighted
adenomas, and intraductal papillary mucinous neo- images and heterogeneous high signal on T2-weighted
plasms. Serous cystadenoma often appears as a images, less intense than typically found in hemangiomas.
multilocular microcystic mass, whereas mucinous Hamartoma may show progressive enhancement with accu-
adenoma usually presents as a unilocular or multi- mulation of contrast within dilated sinusoids in delayed
locular oligocystic mass. Mucinous adenocarci- phase. In contrast to hemangioma, a diffuse heterogeneous
noma usually shows intralesional solid enhancing pattern in arterial phase is usually observed.
components. Because of its rarity, it is often not possible to reliably distin-
• The most frequent malignant pancreatic tumor is guish hamartoma from other more prevalent solid masses,
ductal adenocarcinoma that often appears as a such as metastases.
hypovascular mass associated with biliary and/or The spleen is frequently involved in granulomatous dis-
pancreatic duct dilatation. eases with systemic dissemination, both of infectious and
• Pancreatic neuroendocrine tumors typically immune-mediated etiology. Splenic granulomas often
appear as hypervascular masses. appear as multiple nodules that are hypointense on
• Pancreatic metastases may be hypervascular or T2-weighted images due to a collagenous fibrous compo-
hypovascular depending on the primary tumor. nent, with mild and typically delayed enhancement after
contrast. This appearance should not be confused with sid-
erotic nodules (Gamna-Gandy bodies) that are iron-contain-
ing foci usually associated with diseases causing portal
15.4.4.4 Spleen Masses hypertension.
Splenic lymphoma is the most common splenic malig-
Cystic Lesions nancy that can present as a diffuse spleen enlargement or as
Splenic cysts are the most commonly encountered splenic mass lesions, both solitary and multiple. Imaging features
masses with fluid content; causes may vary from devel- that can rise the suspicion of lymphoma include isointen-
opmental, parasitic, and pseudocyst formation usually sity or mild hypointensity on T2-weighted images and
from prior hematoma. These lesions typically appear as reduced enhancement compared to spleen parenchyma
simple cysts and are often indistinguishable from each (Fig. 15.18).
other by imaging criteria alone. Parasitic echinococcal The spleen is an uncommon site of metastatic involve-
cysts can show a multilocular appearance owing to the ment. Splenic metastases usually present as solitary or mul-
presence of daughter cysts as often found in other ana- tiple nodules, either solid or partially necrotic in their central
tomical sites. Splenic abscesses commonly appear as portion, with variable intensity of enhancement after contrast
fluid-containing lesions, sometimes with irregular mar- administration depending on primary tumor. Metastases are
gins and a thin peripheral rim of contrast enhancement; commonly hyperintense on T2-weighted images, which can
they may be solitary or multiple with unilocular or multi- help in differentiating them from lymphoma.
locular appearance. Purulent material in abscesses often
shows restricted diffusion and occasionally may be asso-
ciated with the presence of gas that appears as signal
voids and susceptibility artifact better recognizable on Key Learning Points
dual-phase chemical shift imaging. Splenic hematomas, • Splenic cystic lesions share similar imaging
both subcapsular and intraparenchymal, show the typical features.
signal characteristics of blood collections as found in • Hydatid disease may appear as multilocular splenic
other organs. cysts.
• Spleen hemangioma is a frequently encountered
Solid Masses benign lesion with typical imaging features.
Hemangioma is the most common benign splenic tumor with • Granulomatous diseases and lymphomas may
imaging features similar to liver hemangiomas; it shows high appear as splenic nodules with low signal intensity
signal on T2-weighted images and peripheral nodular on T2-weighted images.
enhancement in arterial phase with progressive fill-in that
a b
Fig. 15.18 Splenic involvement in Hodgkin’s lymphoma. T2-weighted image (a) and contrast-enhanced image (b) are shown. A lymphomatous
nodule in the spleen (white arrowhead) exhibits low signal intensity on T2-weighted image and hypointensity after contrast administration
a b c
Fig. 15.19 Cystic renal cell carcinoma. T2-weighted image (a), internal septa that demonstrate contrast enhancement and some fluid
T1-weighted image (b), and contrast-enhanced image (c) are shown. In loculation with high signal on T1-weighted images, likely due to a high
the right kidney a complex cystic mass (white arrowhead) is seen with proteinaceous or hemorrhagic content
15.4.4.5 Renal Masses Complicated cysts show characteristic signal abnormali-

ties; besides hyperintensity on T1-weighted images, they can
Cystic Lesions exhibit low signal on T2-weighted images. Less frequently,
Renal cystic lesions are the most commonly encountered focal chronically complicated cysts can appear pseudosolid due to
lesions of the kidney; they include simple cysts containing slightly hyperintense signal on T1-weighted images and near
clear fluid, complicated cysts with proteinaceous or hemor- isointense signal compared to renal parenchyma on
rhagic content, and complex cysts with thick walls, intracystic T2-weighted images; contrast-enhanced imaging can dem-
septations, and sometimes mural nodules. Simple renal cysts onstrate the cystic nature of the lesion. Bosniak classification
appear similar to other pure cysts found in different anatomi- adapted to MR imaging is commonly used to define the risk
cal sites; they are often multiple and can be very numerous in of malignancy and allow correct management of renal cystic
polycystic renal disease (Fig. 15.1). A particular type of renal lesions. Simple cyst, cystic lesions with few hairline-thin
cysts develops in the renal sinus and can have a lymphatic septa and small complicated cyst without contrast enhancing
origin (peripelvic cysts) or results from a renal parenchymal mural or intracystic component are typically benign.
cyst extending into the renal sinus (parapelvic cyst). Peripelvic Complex cysts with thick enhancing walls and septations are
cysts are frequently bilateral and have elongated morphology considered to be at increased risk of malignancy, as they may
that can mimic dilated renal calyces and pelvis; contrast- represent a cystic form of renal cell carcinoma, and are usu-
enhanced images acquired in delayed excretory phase allow to ally surgically excised (Fig. 15.19). About half of complex
differentiate hydronephrosis from peripelvic cysts that can cysts are constituted by benign lesions such as complicated
cause stretching and compression of renal calyces. Renal sinus inflammatory cysts with thick fibrous walls and septa and the
cysts are rarely complicated with hemorrhage or infection. rare multilocular cystic renal tumors. Complex cystic masses
with solid enhancing nodules are very frequently malignant RCC by imaging alone. The presence of a central stellate
and require intervention. scar with high signal on T2-weighted images, a feature often
Hematomas and abscesses appear as fluid-containing encountered in oncocytomas, represents an unspecific find-
masses with imaging features similar to those seen in other ing that can also be found in RCC. Signs of local venous
organs; both lesions may be found inside the kidney, intrapa- invasion are helpful to differentiate benign from malignant
renchymal or subcapsular in location, or involve the perire- solid masses.
nal space. Renal metastases appear as solid masses with variable
enhancement depending on the pathological features of
Solid Masses the primary tumor and when solitary may resemble RCC in
The presence of contrast enhancement within a renal mass several aspects. The frequent multifocality and bilateral renal
allows to reliably differentiate solid tumors from cystic involvement in metastatic disease can suggest the correct
lesions. Renal cell carcinoma (RCC) of clear cell type is the diagnosis. Similarly, renal lymphoma usually encountered as
most common malignant tumor and can have a variable part of a disseminated lymphoproliferative disorder often
imaging appearance ranging from a predominantly cystic appear as multiple bilateral masses that characteristically
mass (Fig. 15.19), often with a small enhancing mural nod- show poor contrast enhancement. Less frequently lymphoma
ules, to a heterogeneous solid mass with intratumoral necro- presents as a solitary mass that may not be distinguishable
sis and hemorrhage. Renal cell carcinoma typically shows from other solid lesions. Other possible patterns of renal
low signal on T1-weighted images and high signal on involvement include direct invasion from retroperitoneal
T2-weighted images; necrotic changes show fluid-like sig- masses, diffuse renal enlargement, and perirenal soft tissue
nal, while hemorrhagic components usually appear hyperin- encasing the kidney.
tense on T1-weighted images due to the prevalence of blood Renal angiomyolipoma, the second most common benign
degradation components in the subacute stage. Some clear renal focal lesion after cysts, typically appears as a
cell RCCs have intratumoral deposit of microscopic fat with fat-containing mass. Angiomyolipoma can be reliably diag-
characteristic signal behavior in dual-phase chemical shift nosed when macroscopic fat is identified in a renal lesion;
imaging. After contrast administration, most RCCs tend to while this imaging feature is strongly suggestive of angio-
inhomogeneously enhance during the arterial phase reflect- myolipoma, it is not patognomonic since in rare cases RCCs
ing their hypervascular nature and often show washout of may contain macroscopic fat, typically associated with foci
contrast in the following contrastographic phases. The sec- of calcification or ossification that can help in the differential
ond most common variant of RCC, the papillary type, may diagnosis. While most angiomyolipomas are composed pre-
be differentiated from its clear cell counterpart because it dominantly of fat that can be associated with solid enhancing
frequently exhibits low signal intensity on T2-weighted component constituted by vessels and smooth muscle, some
images and demonstrates a lesser degree of contrast appear as lipid-poor angiomyolipomas. These tumors may
enhancement (Fig. 15.20) that sometimes requires assess- contain scant amounts of microscopic fat identifiable by
ment of subtraction images between precontrast and post- dual-phase chemical shift imaging; in some cases, no fat is
contrast acquisitions to be recognized. Intratumoral cystic detectable, and imaging appearance may overlap with other
changes and hemorrhage can also be found in papillary solid renal lesions. Small lipid-poor angiomyolipomas can
RCC. Renal oncocytoma is an uncommon benign solid show low signal intensity on T2-weighted images due to the
tumor that cannot be reliably distinguished by clear cell predominant smooth muscle composition, which can be a
a b c
Fig. 15.20 Papillary renal cell carcinoma. T2-weighted image (a), hypointense on T2-weighted image, nearly isointense to renal cortex on
T1-weighted image (b), and contrast-enhanced image (c) are shown. A T1-weighted image, and demonstrates minimal contrast enhancement.
solid mass located in the right kidney (white arrowhead) appears These features are suggestive of papillary-type renal cell carcinoma
useful distinguishing feature compared to clear cell RCC cysts usually resulting from prior hemorrhage, share the
that typically appears hyperintense. same imaging features of cystic lesions in different anatomi-
Transitional cell carcinomas are malignant tumors derived cal sites.
from the urothelium that may arise in renal calyces and pel- Adrenal adenomas, both functioning and nonfunction-
vis. These tumors may appear as vegetating lesions located ing, are common benign adrenal lesions that should be dis-
in excretory cavities or as solid masses that infiltrate renal tinguished from other benign and malignant solid masses.
parenchyma and typically determine less alteration of kidney The typical adrenal adenoma appears as a homogeneous
contours compared to renal cell carcinomas. well-marginated mass with intermediate signal on T1- and
T2-weighted sequences containing abundant intracellular
lipids that can be readily detected using dual-phase chemi-
cal shift imaging (Fig. 15.21). This imaging feature has
Key Learning Points been proven to accurately differentiate benign adenomas
• Simple and complicated renal cysts are frequently from metastases. Dynamic contrast-enhanced imaging, with
occurring lesions with typical imaging features. both early and delayed phase images, proved to be useful for
• Renal cell carcinomas may appear as complex cys- identifying lipid-poor adenomas not recognizable with
tic lesions or as solid masses. chemical shift imaging; after contrast administration, ade-
• Papillary-type renal cell carcinomas often show low nomas tend to show early enhancement followed by con-
signal intensity on T2-weighted images, whereas trast washout.
the clear cell type usually appears hyperintense. Myelolipoma is a mass lesion composed of a mixture of
• Benign renal oncocytoma cannot be reliably differ- mature adipose tissue and hematopoietic elements in vari-
entiated from renal cell carcinoma on imaging able proportions. Some myelolipomas are made almost
alone. entirely of adipose tissue, appearing similar to lipomas in
• Renal angiomyolipomas typically appear as macro- different anatomical sites, while others may show a very
scopic fat-containing masses. inhomogeneous composition with predominant solid tissue
• Renal transitional cell carcinomas often exhibit an intermixed with small areas of macroscopic fat. These patho-
infiltrative growth pattern with less alteration of kid- logical features correlate with specific imaging findings: adi-
ney contours compared to renal cell carcinomas. pose tissue signal appears very high on T1-weighted images
and is nulled on fat-suppressed images, while hematopoietic
tissue shows intermediate signal on T1- and T2-weighted
sequences, enhances after contrast administration, and often
15.4.4.6 Adrenal Masses loses signal on dual-phase chemical shift imaging due to the
Magnetic resonance imaging is particularly useful for char- microscopic fat content.
acterizing adrenal masses. Adrenal cysts, including true Pheochromocytomas are neuroendocrine tumors that can
cysts of developmental or parasitic etiology and pseudo- show variable signal characteristics; in the most typical
a b
Fig. 15.21 Adrenal adenoma. T1-weighted in-phase image (a) and out-of-phase image (b) are shown. Adenoma in the right adrenal gland (white
arrowhead) exhibits a significant signal loss on dual-phase imaging due to the large amount of intracellular fat
cases, they show a markedly high signal intensity on 15.4.4.7 Ovarian Masses
T2-weighted images resulting in the so-called “light bulb”
appearance. This feature is neither sensitive nor specific Cystic Lesions
since many pheochromocytomas have intermediate signal on Magnetic resonance imaging is a widely used diagnostic
T2-weighted images. Pheochromocytomas often have strong tool to evaluate mass lesions in the female pelvis, most
contrast enhancement with prolonged retention of contrast, commonly represented by cystic lesions of the ovaries.
although a washout enhancement pattern similar to adenoma Ovarian follicles and corpora lutea, often incidentally dis-
can sometimes be seen. Large tumors often appear heteroge- covered on imaging in premenopausal women, are normal
neous due to the presence of necrotic changes or hemorrhage structures developing during the menstrual cycle that have
with variable signal intensity depending on the stage of evo- to be distinguished from pathological cystic lesions.
lution. In a minority of cases, pheochromocytomas Normal follicles appear as simple cysts that can grow in
can appear as an entirely cystic lesion with a thin peripheral size up to 3 cm and then spontaneously rupture to release
rim of contrast enhancement that can assist in differential the ovocyte and transform into the corpus luteum. Corpus
diagnosis with simple cysts. luteum typically appears as a small oval mass less than
Adrenal metastases can have variable morphology from 3 cm in diameter with thickened walls and crenulated con-
well-defined round solid nodules to irregular heterogeneous tours that enhance after contrast administration, hyperin-
masses and frequently show bilateral adrenal involvement. tense on T2-weighted images but with lower signal intensity
Metastases commonly exhibit high signal intensity on compared to clear fluid. The most frequently encountered
T2-weighted images without significant signal loss on dual- ovarian cystic lesions larger than 3 cm are functional cysts,
phase chemical shift images. In rare cases, metastases from either follicular cysts, which resemble a simple cyst or cor-
certain types of tumors, such as renal cell carcinoma and pus luteum cysts that may exhibit high signal intensity on
hepatocellular carcinoma, may contain microscopic fat T1-weighted images due to hemorrhage. Endometrioma, a
showing signal loss in chemical shift imaging, but it is usu- localized form of endometriosis, typically appears as a cys-
ally less conspicuous compared to adenomas. Larger metas- tic lesion with hemorrhagic content that should be distin-
tases can show necrotic or hemorrhagic changes. guished from a complicated functional cyst. Endometriotic
Adrenal lymphomas appear as solid masses with imaging cysts usually show high signal intensity on T1-weighted
features that can be indistinguishable from metastases. images and low signal intensity on T2-weighted images
Adrenocortical carcinomas typically present as a large due to the presence of highly concentrated blood degrada-
irregular-shaped mass with predominantly high signal intention components from recurrent hemorrhage (Fig. 15.22).
sity on T2-weighted images and a heterogeneous appear- Sometimes, a fluid-fluid level can form in the endometri-
ance due to the coexistence of solid enhancing tissue and oma with low T2 signal intensity in the dependent portion
areas of necrosis and hemorrhage. Adrenocortical carcino- of the cyst due to sedimentation of blood cells and protein-
mas may contain microscopic fat detectable by dual-phase aceous material with a layered appearance.
chemical shift images and less frequently macroscopic fat A particular type of benign ovarian cystic mass is repre-
deposits; nevertheless, their marked inhomogeneous com- sented by mature cystic teratoma, a fat-containing lesion of
position allows in most cases distinction from other fat-con- germ cell origin. Mature teratoma classically presents as a
taining masses such as adenoma and myelolipoma. cyst with a fluid content mainly composed of sebaceous
Neurogenic tumors like neuroblastomas usually appear as material that has the same signal as fat in all imaging
solid masses with heterogeneous signal intensity and consequences. A solid protuberance projecting from the cyst
trast enhancement due to cystic changes, calcification, and wall (dermoid plug or Rokitansky nodule) with inhomoge-
hemorrhage. neous signal due to variable tissue composition is a typical
finding. In rare cases, the cystic content of mature teratomas
may appear similar to clear fluid and mimic a simple cyst of
other etiology; in these circumstances, a fat component
Key Learning Points within the cyst wall or within a Rokitansky nodule is usually
• Adrenal adenomas can often be differentiated from seen and can assist in the diagnosis. The presence of promi-
other solid masses, such as metastases, owing to nent solid enhancing components within a partially cystic
their microscopic fat content detectable on chemi- fat-containing mass is considered suspicious for immature or
cal shift imaging. malignant teratoma.
• Adrenal myelolipomas are macroscopic fat- Ovarian tumors of epithelial origin usually appear as
containing benign tumors. simple or complex cystic masses and account for the
• Adrenal pheochromocytomas usually show intense majority of ovarian neoplasms. The most frequent benign
contrast enhancement and high signal intensity on cystic ovarian tumors are represented by serous and muci-
T2-weighted images. nous cystadenomas. Serous cystadenoma typically appears
as a unilocular or multilocular cystic mass with homoge-
a b
Fig. 15.22 Ovarian endometriomas. T2-weighted image (a) and fat- signal intensity on T1-weighted image and low signal intensity on
suppressed T1-weighted image (b) are shown. Bilateral ovarian cystic T2-weighted images. These findings are typical of endometriotic cysts
masses (white arrowheads) demonstrate hemorrhagic content with high
a b
Fig. 15.23 Ovarian serous cystadenocarcinoma. T2-weighted image (a) and contrast-enhanced image (b) are shown. Bilateral adnexal complex
masses with solid and cystic components (white arrowheads) are typical of malignant ovarian tumors
neous fluid content, thin walls, and fine internal septa. Solid Masses
Mucinous cystadenoma usually presents as a multilocular Predominantly solid ovarian masses account for a small pro-
cystic mass with heterogeneous appearance due to variable portion of ovarian tumors. Fibrotic ovarian tumors of sex
signal characteristics of the fluid loculations that fre- cord-stromal origin, constituted by the spectrum of related
quently contain different amounts of proteinaceous or pathological entities of the fibroma-thecoma group, typically
mucinous material. In the evaluation of ovarian cystic appear as masses with predominantly low signal intensity on
masses, the presence of thick enhancing walls and septa- T2-weighted images. These lesions may contain areas of
tions (more than 3 mm thick) and intracystic papillary pro- high signal intensity on T2-weighted images attributable to
jections should suggest a borderline or malignant tumor. A stromal edema or cystic changes, more frequently found in
complex cystic mass with large solid components is highly tumors with prevalent thecoma features. Pure thecomas tend
suspicious for ovarian carcinoma (Fig. 15.23), especially to appear as masses with a heterogeneous high signal on
if associated with intralesional necrotic changes or perito- T2-weighted images and are often indistinguishable from
neal effusion. malignant solid tumors. Brenner tumors are benign masses
of transitional cell origin that can show similar signal charac- STIR or fat-suppressed T2-weighted images, it exhibits
teristics as fibromas when mainly solid or appearing as mul- intermediate signal intensity. The relative proportion of
tilocular cystic masses with solid fibrous components. A hematopoietically active red marrow is age-dependent, high-
useful distinguishing feature from other fibrous masses is the est at birth, and gradually decreasing from childhood to
presence of calcification in the solid portions of the tumor, adulthood. Physiological conversion of red marrow to yellow
better recognizable by CT. Granulosa cell tumors are rare marrow starts from the limbs and progresses centrally until
neoplasms of sex cord-stromal origin with indolent but the adult pattern is reached. In adults red bone marrow is
potentially malignant behavior and variable imaging fea- typically confined to axial skeleton, including the spine and
tures. These tumors can appear as predominantly solid the subchondral bone of the humeral and femoral proximal
masses but also as multilocular cystic masses with minimal extremities. Reconversion of yellow to red marrow, a physi-
or no solid component that may be difficult to distinguish ological reaction to a variety of stimuli, consists of the reap-
from more common ovarian epithelial tumors. A typical pearance of red marrow in adult skeleton in places formerly
imaging finding in multilocular granulosa cell tumors is the occupied by yellow marrow. Reconversion tends to appear
presence of intracystic hemorrhage with high signal inten- symmetrical, showing the typical characteristics of normal
sity areas on T1-weighted images and fluid-fluid levels. red marrow and need not to be confused with replacement of
the bone marrow by pathological tissue. Patchy areas of nor-
mal red marrow in vertebral bodies may at times mimic a
focal lesion. The occasional presence of a small adipose
Key Learning Points focus (so-called “bull’s eye” sign) inside red marrow areas
• Endometriomas typically appear as hemorrhagic may suggest the correct diagnosis.
cysts with low signal intensity on T2-weighted Bone marrow edema is a very common but unspecific
images. alteration that can be found in association with a wide variety
• Mature cystic teratomas typically appear as cystic of conditions including traumatic, inflammatory, vascular,
fat-containing masses. degenerative, neoplastic, and metabolic. Edema shows high
• Ovarian tumors of epithelial origin usually appear signal intensity on STIR and T2-weighted images and
as simple or complex cystic masses. Serous cystad- hypointensity on T1-weighted images, often appearing as
enomas often appear as unilocular simple cysts, hazy reticular signal alterations intermixed with normal bone
whereas mucinous cystadenomas usually appear as marrow. In very mild cases, bone edema is detectable on
multilocular cysts with fluid loculations having STIR and fat-suppressed T2-weighted images with minimal
variable signal intensity. Malignant ovarian tumors signal alterations on T1-weighted images. Bone marrow
typically contain papillary projections or solid edema associated with degenerative articular changes typi-
components. cally involves the subchondral bone. In degenerative disk
• Fibrotic ovarian neoplasms like fibroma-thecomas disease, variable signal changes of vertebral bodies along the
and Brenner tumors typically show low signal endplates are a common occurrence; they may be caused by
intensity on T2-weighted images. Granulosa cell reactive vascularized tissue fissuring the endplate and invad-
tumors may sometimes appear as multicystic ing the adjacent vertebral body (low signal intensity on T1-
masses with hemorrhage and fluid-fluid levels. and high signal intensity on T2-weighted and STIR images),
prevalent fatty infiltration (high signal intensity in both
T1-weighted and T2-weighted images with signal suppres-
sion on STIR), and subchondral bone sclerosis (low signal
15.4.4.8 E
valuation of Bone and Soft Tissue intensity in T1- and T2-weighted images).
Lesions A wide range of both benign and malignant diseases that
affect the bone may appear as focal lesions (solitary or mul-
Bone Marrow Signal Alterations tiple), diffuse signal alterations, or aggressive bone destruc-
Bone marrow is mainly composed of a mixture of hemato- tive processes with extraosseous involvement. Hemangioma
poietic tissue, more represented in red marrow, and adipose is a very common benign focal lesion often located in verte-
tissue that makes up the majority of yellow marrow. Yellow bral bodies that typically appear as a predominantly fat-
marrow has the same signal intensity of fat on T1-weighted, containing mass, with signal comparable to that of adipose
T2-weighted, and STIR images. Red marrow typically has a tissue. A minority of hemangiomas contain small amounts of
lower signal intensity than fat on T1-weighted images but fat and may be difficult to distinguish from malignant dis-
higher than muscle tissue or intervertebral disks. On ease because of their unspecific signal intensity (low signal
T2-weighted images, red marrow shows a lower signal than in T1-weighted and high signal in T2-weighted and STIR
fat and a slightly higher signal compared to muscle, while on images). Enchondroma is another type of benign bone tumor
a b
Fig. 15.24 Vertebral metastasis in patient with renal cell carcinoma. and lamina partially involving the posterior vertebral body (white aster-
T1-weighted image (a) and gradient-echo T2*-weighted image (b) are isk) with extraosseous extension
shown. A destructive bone lesion is seen in the right vertebral pedicle
often incidentally discovered in long bones. The typical undetectable on STIR images but are easily recognizable on
imaging appearance of enchondroma is a lobulated intra- T1-weighted images. Sometimes metastatic disease may
medullary mass with low signal intensity on T1-weighted appear as disseminated multinodular or diffuse bone marrow
images and markedly high signal intensity on T2-weighted replacement that is better seen on T1-weighted sequences,
images that may contain foci of very low signal due to calci- especially in the spine where signal intensity of vertebral
fication. Bone cystic lesions often have similar characteris- bodies and posterior elements may be lower than the signal
tics as cysts found in other anatomical sites. Aneurysmal intensity of normal intervertebral disks. The different pat-
bone cyst is a lesion that may arise in bone as an isolated terns of metastatic bone involvement are nonspecific and can
alteration or in association with other benign and malignant be found in other malignant infiltrative processes such as
tumors and is characterized by multiple loculations contain- leukemia/lymphoma and multiple myeloma.
ing fluid-fluid levels of variable signal intensity due to the
presence of hemorrhage. Soft Tissue Masses with Typical Imaging Features
In aggressive bone diseases such as osteomyelitis and Soft tissue masses are lesions originating from mesenchyme-
neoplasms, the normal bone marrow is replaced by patho- derived connective tissues. A few common soft tissue
logical tissue that shows high signal intensity on T2-weighted masses present with some specific imaging features that are
and STIR images, lower signal intensity compared with bone useful for their characterization. Lipoma is the most fre-
edema on T1-weighted images and typically exhibits con- quently encountered benign soft tissue lesion, typically
trast enhancement. In infectious processes like osteomyelitis appearing as a mass composed entirely of fat. A minority of
and spondylodiscitis, cortical bone destruction may be asso- lipomas may show thin internal septations that appear
ciated with the presence of fluid collections that can involve hypointense compared to fat or may contain non-adipose
surrounding soft tissues. In malignant neoplasms arising pri- tissue components mostly represented by degenerative
marily in the bones, the signal intensity of the tumor can be changes like fibrosis, liponecrosis, calcification, and myx-
variable, reflecting different pathological features. oid degeneration. These complex-appearing lipomas may
Extraosseous involvement with local soft tissue invasion is a be difficult to distinguish from well-differentiated liposar-
common occurrence in primary bone tumors but can also be comas (Fig. 15.4). The presence of thick nodular septa-
seen in metastatic bone disease (Fig. 15.24). Metastases tions or non-fatty solid areas should rise the suspicion of
often appear as well-demarcated focal lesions with variable malignancy. Poorly differentiated liposarcomas
pattern of skeletal involvement and signal characteristics often appear as completely solid lesions with minimal fat,
similar to other bone replacement processes. Less frequently at times detectable only on dual-phase chemical shift imag-
osteoblastic metastases may demonstrate low signal inten- ing, and have variable signal intensity reflecting pathologi-
sity in all imaging sequences due to the prevalence of scle- cal features (myxomatous variants typically show high
rotic bone over tumor tissue; in these cases they can be signal intensity on T2-weighted images). Lipomas may
arise in superficial or deep compartments, the latter includ- myxoid, and hemorrhagic changes can further complicate
ing mediastinal and retroperitoneal lipomatous tumors that the imaging appearance. Synovial sarcoma is a relatively
carry a worse prognosis compared to superficial masses. common malignant neoplasm often arising in soft tissues
Peripheral nerve sheath tumors are neurogenic masses near large joints, more frequently around the knee. It may
mainly represented by schwannomas and neurofibromas, show typical although non-pathognomonic imaging find-
the main difference being their relationship with nerve ings: marked heterogeneous signal on T2-weighted images
fibers. These tumors share similar imaging findings and due to the admixture of necrotic and hemorrhagic areas
cannot be easily differentiated from each other based on (very high T2 signal intensity), solid enhancing tissue
imaging alone. Both lesions typically appear as elongated (intermediate T2 signal intensity), and fibrotic changes
solid masses oriented parallel to the nerve with smooth with calcifications (low T2 signal intensity) creating the so-
contours, hyperintense on T2-weighted images and nearly called “triple sign”. In a minority of cases, synovial sar-
isointense compared to muscle on T1-weighted images, coma shows a multiloculated appearance (so-called “bowl
often surrounded by a rim of normal fat tissue (so-called of grapes”) due to clustering of multiple cystic areas that
split fat sign). In imaging planes oriented perpendicular to contain fluid-fluid levels with variable signal intensities
the tumor long axis, an area of low signal intensity on due to hemorrhage.
T2-weighted images can sometimes be appreciated in the
central portion (so-called target sign) attributable to collag-
enous fibrous tissue; this imaging appearance is more com-
monly seen in neurofibroma but is not a reliable Key Learning Points
distinguishing feature. It is not uncommon for peripheral • Red marrow shows lower signal intensity compared
nerve sheath tumors to show inhomogeneous signal on to yellow fatty marrow.
T2-weighted images with fluid-like intralesional compo- • Bone marrow edema shows high signal intensity on
nents due to myxoid degeneration and cystic changes. fat-suppressed T2-weighted and STIR images.
These lesions may demonstrate variable degrees of contrast • Benign bone lesions like cysts and enchondromas
enhancement. Elastofibroma dorsi is a benign mass com- typically show very high signal intensity on
posed of fibroelastic tissue (nearly isointense to muscle on T2-weighted images.
T1- and T2-weighted images) intermixed with linear fat • Vertebral hemangiomas usually appear as macro-
components that give the classic striated appearance. scopic fat-containing masses.
Hemangiomas arising in soft tissues may have variable sig- • Aggressive bone diseases, like infectious processes
nal characteristics depending on pathological features. and malignant primary and secondary neoplasms,
Most lesions are composed predominantly of dilated ves- usually show low signal intensity on T1-weighted
sels with slow-flowing blood, showing high signal intensity images.
on T2-weighted images and variable enhancement on post- • Lipomas typically appear as masses composed
contrast images. Internal foci of low T2 signal intensity entirely of macroscopic fat. Liposarcomas
may be generated by flow voids in areas of rapidly flowing often appear as fat-containing lesions with thick
blood or more typically by phleboliths. Hemangiomas may nodular septations and solid non-fatty components.
also contain fat, fibrous, or hemorrhagic components that • Schwannomas and neurofibromas usually appear as
have typical signal intensities. Gradient-echo T2*-weighted high signal intensity masses on T2-weighted
images are useful to detect the presence of hemosiderin images, sometimes with myxoid and cystic changes.
within hemangiomas and to help characterize other soft tis- • Soft tissue hemangiomas typically appear markedly
sue masses that contain large amounts of hemosiderin, such hyperintense on T2-weighted images. Signal voids
as giant cell tumors. Pure cystic lesions of soft tissues are from fast-flowing blood or phleboliths may be
readily recognized by their typical appearance and specific present.
location: synovial cysts, ganglia, and clear fluid collec- • Cystic and myxomatous lesions show very high sig-
tions. Myxomatous tumors, both benign and malignant, nal intensity on T2-weighted images and can be dif-
may at times appear as cyst-like masses due to high signal ferentiated from each other on contrast-enhanced
intensity on T2-weighed images and hypointensity on images.
T1-weighted images. The presence of contrast enhance- • Soft tissue sarcomas show variable signal
ment, often of low intensity, is useful to distinguish myxo- characteristics.
matous lesions from cysts. Soft tissue sarcomas in most • Some tumors, such as synovial sarcomas, may
cases appear as inhomogeneous predominantly solid sometimes show typical, although non-pathogno-
enhancing masses with unspecific s ignal intensity on T1- monic, imaging features.
and T2-weighted images. Intralesional cystic, necrotic,
Further Reading McInnis MC, Flores EJ, Shepard JA, Ackman JB. Pitfalls in the imag-
ing and interpretation of benign thymic lesions: how thymic MRI
can help. AJR Am J Roentgenol. 2016;206(1):W1–8.
Ackman JB, Wu CC. MRI of the thymus. AJR Am J Roentgenol.
Miller FH, Rini NJ, Keppke AL. MRI of adenocarcinoma of the pan-
2011;197(1):W15–20.
creas. AJR Am J Roentgenol. 2006;187(4):W365–74.
Albiin N. MRI of focal liver lesions. Curr Med Imaging Rev.
Mittal MK, Malik A, Sureka B, et al. Cystic masses of neck: a pictorial
2012;8(2):107–16.
review. Indian J Radiol Imaging. 2012;22(4):334–43.
Allen BC, Hosseinzadeh K, Qasem SA, et al. Practical approach
Nakazono T, White CS, Yamasaki F, et al. MRI findings of mediastinal
to MRI of female pelvic masses. AJR Am J Roentgenol.
neurogenic tumors. AJR Am J Roentgenol. 2011;197(4):W643–52.
2014;202(6):1366–75.
Nascimento D, Suchard G, Hatem M, et al. The role of magnetic reso-
Baliyan V, Das CJ, Sharma R, et al. Diffusion weighted imaging: tech-
nance imaging in the evaluation of bone tumours and tumour-like
nique and applications. World J Radiol. 2016;8(9):785–98.
lesions. Insights Imaging. 2014;5(4):419–40.
Bangiyev L, Rossi Espagnet MC, Young R, et al. Adult brain tumor
Nikken JJ, Krestin GP. MRI of the kidney—state of the art. Eur Radiol.
imaging: state of the art. Semin Roentgenol. 2014;49:39–52.
2007;17(11):2780–93.
Bitar R, Leung G, Perng R, et al. MR pulse sequences: what every
Nouh MR, Eid AF. Magnetic resonance imaging of the spinal marrow:
radiologist wants to know but is afraid to ask. Radiographics.
basic understanding of the normal marrow pattern and its variant.
2006;26(2):513–37.
World J Radiol. 2015;7(12):448–58.
Blake MA, Cronin CG, Boland GW. Adrenal imaging. AJR Am J
Ödev K, Arıbaş BK, Nayman A, et al. Imaging of cystic and cyst-like
Roentgenol. 2010;194(6):1450–60.
lesions of the mediastinum with pathologic correlation. J Clin
Chundru S, Kalb B, Arif-Tiwari H, et al. MRI of diffuse liver disease:
Imaging Sci. 2012;2:33.
characteristics of acute and chronic diseases. Diagn Interv Radiol.
Palas J, Matos AP, Ramalho M. The spleen revisited: an overview on
2014;20(3):200–8.
magnetic resonance imaging. Radiol Res Pract. 2013;2013:219297.
Del Grande F, Farahani SJ, Carrino JA, et al. Bone marrow lesions:
Pedrosa I, Sun MR, Spencer M, et al. MR imaging of renal masses:
a systematic diagnostic approach. Indian J Radiol Imaging.
correlation with findings at surgery and pathologic analysis.
2014;24(3):279–87.
Radiographics. 2008;28(4):985–1003.
Elsayes KM, Mukundan G, Narra VR, et al. Adrenal masses: MR
Puderbach M, Hintze C, Ley S, et al. MR imaging of the chest: a practi-
imaging features with pathologic correlation. Radiographics.
cal approach at 1.5T. Eur J Radiol. 2007;64(3):345–55.
2004;24(Suppl 1):S73–86.
Revelli M, Chiesa F, Del Prato A, et al. Role of respiratory-triggered
Foti PV, Attinà G, Spadola S, et al. MR imaging of ovarian masses:
diffusion-weighted MRI in the assessment of pleural disease. Br J
classification and differential diagnosis. Insights Imaging.
Radiol. 2016:20160289.
2016;7(1):21–41.
Sandrasegaran K, Lin C, Akisik FM, et al. State-of-the-art pancreatic
Freire M, Remer EM. Clinical and radiologic features of cystic renal
MRI. AJR Am J Roentgenol. 2010;195:42–53.
masses. AJR Am J Roentgenol. 2009;192:1367–72.
Schima W. MRI of the pancreas: tumours and tumour-simulating pro-
Ginat DT, Meyers SP. Intracranial lesions with high signal intensity
cesses. Cancer Imaging. 2006;6(1):199–203.
on T1-weighted MR images: differential diagnosis. Radiographics.
Shah LM, Hanrahan CJ. MRI of spinal bone marrow: part I, tech-
2012;32:499–516.
niques and normal age-related appearances. AJR Am J Roentgenol.
Hanrahan CJ, Shah LM. MRI of spinal bone marrow: part 2,
2011;197(6):1298–308.
T1-weighted imaging-based differential diagnosis. AJR Am J
Sharma V, Prabhash K, Noronha V, et al. A systematic approach
Roentgenol. 2011;197(6):1309–21.
to diagnosis of cystic brain lesions. South Asian J Cancer.
Hoang JK, Vanka J, Ludwig BJ, et al. Evaluation of cervical
2013;2(2):98–101.
lymph nodes in head and neck cancer with CT and MRI: tips,
Smirniotopoulos JG, Murphy FM, Rushing EJ, et al. Patterns of con-
traps, and a systematic approach. AJR Am J Roentgenol.
trast enhancement in the brain and meninges. Radiographics.
2013;200(1):W17–25.
2007;27(2):525–51.
Inaoka T, Takahashi K, Mineta M, et al. Thymic hyperplasia and thy-
Stadnik TW, Demaerel P, Luypaert RR, et al. Imaging tutorial: differ-
mus gland tumors: differentiation with chemical shift MR imaging.
ential diagnosis of bright lesions on diffusion-weighted MR images.
Radiology. 2007;243(3):869–76.
Radiographics. 2003;23(1):e7.
Juanpere S, Cañete N, Ortuño P, et al. A diagnostic approach to the
Takahashi K, Al-Janabi NJ. Computed tomography and magnetic
mediastinal masses. Insights Imaging. 2013;4(1):29–52.
resonance imaging of mediastinal tumors. J Magn Reson Imaging.
Kakkar C, Shetty CM, Koteshwara P, et al. Telltale signs of peripheral
2010;32(6):1325–39.
neurogenic tumors on magnetic resonance imaging. Indian J Radiol
Tani I, Kurihara Y, Kawasuchi A, et al. MR imaging of diffuse liver
Imaging. 2015;25(4):453–8.
disease. AJR Am J Roentgenol. 2000;174:965–71.
Kalb B, Sarmiento JM, Kooby DA, et al. MR imaging of cystic lesions
Thoeny HC. Imaging of salivary gland tumours. Cancer Imaging.
of the pancreas. Radiographics. 2009;29(6):1749–65.
2007;7(1):52–62.
Khosa F, Khan AN, Eisenberg RL. Hypervascular liver lesions on
Vancauwenberghe T, Snoeckx A, Vanbeckevoort D, et al. Imaging
MRI. AJR Am J Roentgenol. 2011;197(2):W204–20.
of the spleen: what the clinician needs to know. Singap Med J.
Lee KY, Oh YW, Noh HJ, et al. Extraadrenal paragangliomas of the body:
2015;56(3):133–44.
imaging features. AJR Am J Roentgenol. 2006;187(2):492–504.
Zimny A, Neska-Matuszewska M, Bladowska J, et al. Intracranial
Matos AP, Velloni F, Ramalho M, et al. Focal liver lesions: practi-
lesions with low signal intensity on T2-weighted MR images –
cal magnetic resonance imaging approach. World J Hepatol.
review of pathologies. Pol J Radiol. 2015;80:40–50.
2015;7(16):1987–2008.
Image-Guided and Radioguided
Surgery 16
Francesco Giammarile, Sergi Vidal-Sicart, Federica Orsini,
Renato A. Valdés Olmos, and Giuliano Mariani
Contents
16.2 Radioguided Surgery in Breast Cancer 355
16.2.1 The Clinical Problem 355
16.2.2 Pathophysiologic Issues on Sentinel Lymph Node Mapping 355
16.2.3 Procedures for Sentinel Lymph Node Mapping/Biopsy 356
16.2.4 Radioguided Occult Lesion Localization (ROLL) with 99mTc-MAA 363
16.2.5 Added Value of Intraoperative Portable Gamma Cameras 364
16.2.6 Future Perspectives 365
16.3 Radioguided Sentinel Lymph Node Biopsy in Cutaneous Melanoma 366
16.3.2 Lymphatic System of the Skin 366
16.3.3 Tracer Injection 366
16.3.4 Lymphoscintigraphy 367
16.3.5 Intraoperative Counting/Detection 369
16.3.6 Accuracy of Sentinel Lymph Node Biopsy and Long-Term Prognosis 372
16.3.7 Indications and Contraindications 373
16.4 Radioguided Sentinel Lymph Node Biopsy in Head and Neck Cancers 373
16.4.2 Lymphatic System of the Head and Neck 374
16.4.3 Modalities of Tracer Injection 375
16.4.5 Intraoperative Gamma Probe Counting/Detection 377
16.4.6 Retrieval and Analysis of Sentinel Lymph Nodes 378
16.5 Radioguided Sentinel Lymph Node Biopsy in Gynecological Cancers 379
16.5.2 Lymphatic System of Intrapelvic Female Organs 380
R. A. Valdés Olmos
F. Giammarile Nuclear Medicine Section and Interventional Molecular Imaging
Nuclear Medicine and Diagnostic Imaging Section, Division of Laboratory, Department of Radiology, Leiden University Medical
Human Health, Department of Nuclear Sciences and Applications, Centre, Leiden, The Netherlands
International Atomic Energy Agency, Vienna, Austria
Nuclear Medicine Department, Onze Lieve Vrouwe Gasthuis
S. Vidal-Sicart Hospital, Amsterdam, The Netherlands
Nuclear Medicine Department, University Hospital Clinic,
G. Mariani (*)
Barcelona, Spain
F. Orsini Research and Advanced Technologies in Medicine and Surgery,
Nuclear Medicine Unit, “Maggiore della Carità” University University of Pisa, Pisa, Italy
Hospital, Novara, Italy e-mail: giuliano.mariani@med.unipi.it

352 F. Giammarile et al.
16.5.3 odalities of Tracer Injection

M 380
16.5.5 Intraoperative Gamma Probe/Gamma Camera Detection 383
16.5.6 Retrieval and Analysis of Sentinel Lymph Nodes 384
References 386
Learning Objectives • To acquire basic knowledge on the clinical applications of

• To become familiar with the overall practice and rele- sentinel lymph node biopsy in patients with head and
vance of image-guided and radioguided surgery in cancer neck cancers.
patients. • To understand the pathophysiologic issues involved in the
• To acquire basic knowledge on the relevance of tumor rationale and practice of radioguided sentinel lymph node
status of regional lymph nodes in patients with solid
biopsy in patients with head and neck cancers.
epithelial tumors. • To become familiar with the procedures involved in lym-
• To understand the pathophysiologic bases of the sentinel phatic mapping and radioguided sentinel lymph node
lymph node concept in oncologic surgery. biopsy in patients with head and neck cancers, including
• To acquire basic knowledge on the clinical applications of injection of the radioactive agent for lymphoscintigraphy,
sentinel lymph node biopsy in patients with breast cancer. preoperative gamma camera imaging, and intraoperative
• To understand the pathophysiologic issues involved in the radioguided sentinel lymph node localization aided by the
rationale and practice of radioguided sentinel lymph node use of the handheld gamma-detecting probe.
biopsy in patients with breast cancer. • To become familiar with the advantages of intraoperative
• To become familiar with the procedures involved in lym- imaging with portable gamma cameras in patients with
phatic mapping and radioguided sentinel lymph node head and neck cancers.
biopsy in patients with breast cancer, including injection • To acquire basic knowledge on the clinical applications of
of the radioactive agent for lymphoscintigraphy, preoper- sentinel lymph node biopsy in patients with gynecologi-
ative gamma camera imaging, and intraoperative radiogu- cal cancers.
ided sentinel lymph node localization aided by the use of • To understand the pathophysiologic issues involved in the
the handheld gamma-detecting probe. rationale and practice of radioguided sentinel lymph node
• To understand the rationale and acquire knowledge on the biopsy in patients with gynecological cancers.
practice of radioguided occult lesion localization in patients • To become familiar with the procedures involved in lym-
with breast cancer with non-palpable breast lesions. phatic mapping and radioguided sentinel lymph node
• To become familiar with the advantages of intraoperative biopsy in patients with gynecological cancers, including
imaging with portable gamma cameras in patients with injection of the radioactive agent for lymphoscintigraphy,
breast cancer. preoperative gamma camera imaging, and intraoperative
• To acquire basic knowledge on the clinical applications of radioguided sentinel lymph node localization aided by the
sentinel lymph node biopsy in patients with cutaneous use of the handheld gamma-detecting probe.
melanoma. • To become familiar with the advantages of intraoperative
• To understand the pathophysiologic issues involved in the imaging with portable gamma cameras in patients with
rationale and practice of radioguided sentinel lymph node gynecological cancers.
biopsy in patients with cutaneous melanoma.
• To become familiar with the procedures involved in lym-
phatic mapping and radioguided sentinel lymph node
biopsy in patients with cutaneous melanoma, including 16.1 Introductory Background
injection of the radioactive agent for lymphoscintigraphy,
preoperative gamma camera imaging, and intraoperative Together with PET, radioguided surgery is the nuclear medi-
radioguided sentinel lymph node localization aided by the cine application that has experienced the most relevant
use of the handheld gamma-detecting probe. growth in the last 15 years. The simple term “radioguided
• To become familiar with the advantages of intraoperative surgery” actually constitutes a wide range of combined pro-
imaging with portable gamma cameras in patients with cedures based on synergistic collaboration between at least
cutaneous melanoma. two different specialties, nuclear medicine and surgery, but
16 Image-Guided and Radioguided Surgery 353
often involves other specialties as well; correct application of mary tumor; (2) systemic administration of a tumor-seeking
such procedures ensues obvious immediate and long-term radiopharmaceutical that preferentially accumulates in the
benefits to the patient. target lesion; and (3) direct intralesional administration of a
Radioguided surgery includes a set of pre-, intra-, and radiopharmaceutical that, by virtue of a relatively large size,
postoperative techniques and procedures that are designed to is retained virtually indefinitely at the injection site for the
optimize traditional oncologic surgery. All these technolo- so-called radioguided occult lesion localization.
gies may be encompassed in the “guided intraoperative scin- The sentinel lymph node (SLN) procedure is a diagnostic
tigraphic tumor targeting” (GOSTT) concept [1, 2]. As a staging procedure that is applied in a variety of tumor types.
common intrinsic feature, the GOSTT concept implies the The procedure aims to determine the tumor status of the
preoperative administration of a radiopharmaceutical that, SLN(s), defined as any lymph node on a direct drainage path-
though possibly with different modalities, concentrates pref- way from the primary tumor [3]. The concept is based on the
erentially in a certain target tissue to be removed. Preoperative premise that lymph flow from the primary tumor travels first to
imaging (most commonly with a conventional gamma the SLN and then sequentially to the other regional lymph
camera—sometimes with PET) provides useful information nodes—a process that underlies the concept of “orderly pro-
on the optimal surgical approach to be adopted for each indi- gression” of metastatic tumor cells along a certain path of lym-
vidual patient, while the intraoperative use of a handheld phatic drainage (Fig. 16.1). Therefore, the histopathologic
radioactivity-detecting probe (the gamma probe) further status of this node should reflect the histopathologic status of
guides the surgeon for the identification and removal of the the entire nodal basin. This concept translates into the clinical
target tissue that has preoperatively been “labeled” with the practice as follows:
Preferential accumulation of a radiopharmaceutical in the • If metastasis is found in the SLN(s), there is a high prob-
target lesion can be achieved by three main mechanisms: (1) ability that other lymph nodes in the same lymphatic
interstitial administration of an adequate radiopharmaceuti- basin harbor metastasis as well.
cal (typically a radiocolloid) that depicts the pattern of lym- • Current recommendations in case of metastatic SLN(s)
phatic drainage from the site of a solid epithelial tumor, to call for proceeding to completion lymphadenectomy of
identify the sentinel lymph node(s) draining from the pri- that particular basin (except in some specific instances
Fig. 16.1 The sentinel lymph node concept assumes that lymph drains exist at higher levels, with variable directions of lymph flow at interme-
from a solid tumor in an orderly fashion from lower-echelon to higher- diate levels. The pattern of lymph flow is even more complex when
echelon lymph nodes. Therefore, tumor cells migrating through lym- considering that lymph from the tumor can drain to more than one lym-
phatic channels are most likely to be entrapped and possibly originate phatic basin, each one repeating the basic pattern represented here for a
metastasis in the first node(s) they encounter, the sentinel node(s), single basin (reproduced with permission from: Giammarile F, Orsini F,
before spreading to higher-echelon lymph nodes along the same lym- Valdés Olmos RA, Vidal-Sicart S, Giuliano AE, Mariani G. Radioguided
phatic drainage pathway. Nevertheless, even in any given lymphatic surgery for breast cancer. In: Strauss HW, Mariani G, Volterrani D,
basin, there can be more than one sentinel lymph node, as lymph can Larson SM, eds. Nuclear oncology—From pathophysiology to clinical
drain from the site of the primary tumor via different lymphatic chan- applications. New York, NY: Springer; 2017:1363–1400)
nels toward the same basin. Furthermore, complex interconnections can
still under investigation—see further below in the section a 2nd

on breast cancer).
• Lymphadenectomy of a certain lymphatic basin is bur-
dened with a relatively high probability of immediate and SN
delayed local complications.
• Completion lymphadenectomy in a basin with metastatic
SLN(s) is performed either for purely staging purposes 3rd
(as in patients with breast cancer) or for treatment pur-
poses (as in patients with cutaneous melanoma). 2nd
• On the other hand, if no metastatic tumor cells are found
in the SLN(s), it is very unlikely that other lymph nodes 3rd
of that particular basin harbor metastasis.
• Therefore, completion lymphadenectomy can safely be
omitted, thus sparing to these patients unnecessary addi- b
tional surgery—with the associated complications.
SN
The sentinel lymph node concept takes on a special patho-
physiologic rather than a purely anatomic value. In particu-
lar, the sentinel lymph node is not always the node
anatomically closest to the tumor but rather the first node that
tumor cells migrating through the lymphatic channels drain-
ing the tumor site encounter along a certain drainage path-
way (Fig. 16.2).
In addition to being routinely applied in patients with
breast cancer and cutaneous melanoma, radioguided SLN
Fig. 16.2 (a) Lymphatic mapping for sentinel node biopsy assumes
biopsy is being validated in patients with a wide variety of that lymph from a primary cancer (depicted with a dark red colour)
other solid epithelial cancers. In particular, malignancies drains to a particular regional lymph node before reaching higher-tier
where the feasibility and/or clinical impact of radioguided lymph nodes along the same lymphatic pathway. (b) The lymph node
sentinel lymph node biopsy has been more thoroughly inves- anatomically closest to the cancer is not necessarily directly at risk of
receiving tumor cells. SN = sentinel lymph node, 2nd = second-tier
tigated include head and neck cancers and gynecological can- lymph node, 3rd = third-tier lymph node (modified from: Nieweg
cers (vulvar, cervical, and endometrial). Other less validated OE. The sentinel lymph node concept in oncologic surgery. In: Mariani
clinical applications include prostate cancer, cancers of the G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, eds. Atlas of
gastrointestinal tract (esophagus, stomach, colorectal, anus), Lymphoscintigraphy and sentinel node mapping—A pictorial case-
based approach. Milan: Springer-Verlag Italy; 2013:87–94)
non-small cell lung cancer, differentiated thyroid carcinoma,
and others.
Sentinel lymph node mapping in patients with penile can-
cer, an application with intermediate levels of evidence as • The rationale of sentinel lymph node biopsy in
regards clinical benefit to patients, is discussed in Chap. 35 of patients with epithelial solid cancers is based on the
this book. concept of “orderly progression” of metastatic
tumor cells along a certain lymphatic pathway.
• A negative sentinel lymph node biopsy makes it
highly unlikely that other lymph nodes in the same
lymphatic basin harbor metastasis.
Key Learning Points • Sentinel lymph node biopsy is currently the stan-
• Radioguided surgery includes a set of pre-, intra-, dard of care in patients with early breast cancer or
and postoperative techniques and procedures that cutaneous melanoma and extensively validated also
are designed to optimize oncologic surgery. in patients with head and neck cancers or with
• The fundamental common requisite for successful gynecological malignancies.
radioguided surgery is the use of a radiopharmaceu- • Application of sentinel lymph node biopsy in other
tical that, following either systemic or interstitial malignancies (prostate cancer, cancers of the gas-
administration, preferentially accumulates at the trointestinal tract, differentiated thyroid cancer,
lesion/tissue to be excised. non-small cell lung cancer, and others) is still under
clinical validation for feasibility and accuracy.
16.2 Radioguided Surgery in Breast Cancer

Key Learning Points
16.2.1 The Clinical Problem • In patients with breast cancer, accurate lymph node
staging is important for both prognosis and treat-
Breast cancer is the most frequent cancer diagnosed in women ment purposes.
worldwide. In patients with breast cancer, accurate lymph • Clinical examination and diagnostic imaging are
node staging is essential for both prognosis (of early-stage not accurate enough for reliable preoperative
disease) and treatment (for regional control of disease) [4]. assessment of axillary lymph node staging in
Clinical examination (i.e., palpation) is not accurate patients with early breast cancer.
enough for assessing the axillary lymph node status, and • De novo axillary lymph node dissection (a futile
preoperative imaging modalities including PET/CT with surgery in nearly 80% of patients with early breast
[18F]FDG and ultrasound have low sensitivities, especially cancer) is burdened with a high incidence of early
in case of micrometastatic disease [5]. Thus, the traditional and late postoperative complications.
staging approach for patients with breast cancer requires • Radioguided sentinel lymph node biopsy, a much less
axillary lymph node dissection (ALND) for extensive his- invasive procedure, identifies with high accuracy those
tologic evaluation of up to 15–20 lymph nodes. However, 20% of early breast cancer patients with lymph node
ALND results in a high incidence of postoperative compli- metastasis who actually need lymph node dissection.
cations that can reduce the quality of life. Furthermore, in
early breast cancer, nearly 80% of axillary dissections
reveal no metastasis and, therefore, could have been 16.2.2 Pathophysiologic Issues on Sentinel
avoided [6]. Lymph Node Mapping
Radioguided SLN biopsy has become a routine technique
in breast cancer management, contributing to development The breast embryologically originates from ectodermal tissue,
of less invasive surgical procedures [7]. A systematic review as a skin appendage, and therefore shares a pattern of lym-
showed an axillary 0.6% relapse rate in SLN-negative phatic drainage with the overlying skin. The mammary gland
patients and no benefits from ALND in terms of survival is interposed between the superficial (subdermal) and the deep
after a negative SLN biopsy [8]. (subcutaneous) lymphatic plexuses; the two systems are inter-
The main objective of SLN mapping (SLNM) and SLN connected by a dense network of lymphatic vessels (Fig. 16.3).
biopsy (SLNB) in breast cancer patients is axillary stag- Lymphatic vessels surrounding the mammary lobules pre-
ing. These procedures are an appropriate alternative to dominantly drain to the subareolar Sappey’s plexus, which is
routine staging ALND for patients with early-stage part of the superficial plexus of the skin. Most of the lymph
biopsy-proven breast carcinoma without cytologically or produced in the breast is therefore drained to the subareolar
histologically proven axillary lymph node metastases. region, progressing then toward the ipsilateral axillary nodes.
Appropriately identified patients with negative SLNB do A small fraction of the lymph produced in the breast (about
not need to undergo ALND. On the other hand, recent evi- 3%, mostly from the deeper portions of the breast paren-
dence suggests that ALND can safely be omitted also in chyma) drains instead to lymph nodes of the internal mam-
patients with early breast cancer even in the case of posi- mary chain, while an even smaller amount drains to other
tive SLNB, provided that some tumor-related features lymph nodes such as the intercostal and pectoral muscles, con-
indicate a nonaggressive pattern of growth in that particu- tralateral breast, or even abdominal lymph nodes.
lar patient [9–12].
Despite the widespread application of SLNB for early-
stage breast cancer, there is significant variation in perfor-
mance characteristics reported for such procedures. The Key Learning Points
ranges of rates for false-negative findings and for SLN iden- • Most lymphatic drainage from the mammary gland
tifications emphasize the variability of this procedure. flows toward axillary lymph nodes, after merging in
Learning curves for this technical procedure also vary. the subareolar Sappey’s plexus.
Nevertheless, once a multidisciplinary team is experienced • About 3% only of the lymph produced in the breast
with the procedure, reasonable levels of accuracy are (mostly originating from the deeper parenchyma)
achieved, with identification rates well above 95% reported drains to lymph nodes of the internal mammary chain.
routinely [13].
a epidermis localization for surgical removal of the detected lymph nodes

superficial plexus
(LNs). Although there is consensus on some broad aspects of
SLN protocols for breast cancer, consensus does not exist on all
dermis pre-collectors details. Controversies exist with regard to the particle size of the
radiocolloid, the optimal route for injection, the timing and type
superficial fascia
subcutaneous plexus
of scintigraphy and intraoperative detection, and whether or not
deep fascia extra-axillary SLNs should be considered for harvesting and
analysis. In addition, the specific radiotracer and technique used
are guided by local availability, regulations, and practices [14].
adipose tissue deep collectors
16.2.3.1 Procedures in Nuclear Medicine

Three main parameters define an optimal tracer administration
technique for radioguided SLNB: injection site, injected vol-
infraclavicular
b ume, and injected activity. A fourth parameter to be taken into
lateral apical account is the time elapsed between injection and surgery, as it
directly influences the amount of radioactivity to be injected.
Radiopharmaceuticals
subareolar plexus
The ideal radiotracer should show rapid transit to SLNs with
central cross-drainage
prolonged retention in the nodes. In general, the drainage,
subscapular
distribution, and clearance of radioactive colloids by the
parasternal lymphatic system may vary depending on the size of the par-
pectoral
ticles. Small particles are drained and cleared first; large par-
from thoracic wall
ticles are drained and cleared last and may be retained longer
to abdominal wall
at the injection site. Nevertheless, it has been shown the suc-
subcutaneous plexus
cess rate in the identification of axillary SLNs is not signifi-
Fig. 16.3 (a) Schematic representation of blood circulation and lym- cantly affected by the particle size of the radiotracer. Thus,
phatic circulation in the skin. For easier comprehension, the lymph and the selection of radiotracer is based more on local availabil-
blood vessel networks (which are actually embedded in each other) are ity than on differences in SLN detection. However, there is
represented separately, respectively, on the right (yellow) and on the left general agreement that a 100–200-nm-sized radiocolloid
(red and blue). Its embryologic origin in the ectoderm places the mam-
mary gland in an ideal space between the subcutaneous plexus and the should be considered the best compromise between fast lym-
deep lymphatic collectors (magnified insert). Branches of the periductal phatic drainage and optimal retention in SLNs [15].
plexus drain lymph mostly toward the skin surface (via the subareolar The radiopharmaceuticals most widely used for radioguided
plexus), while a minor component drains toward the deep collectors SLNB in breast cancer patients are 99mTc-sulfur colloid (particle
(which in turn drain toward the internal mammary chain). Dark red
arrows indicate the direction of lymph flow at different levels, merging size: 15–5000 nm, usually filtered to select a more restricted
toward the subcutaneous plexus. Lymphatic mapping agents injected range of particle size), 99mTc-nanocolloidal albumin (5–100 nm),
intradermally over the mammary gland drain to the subcutaneous plexus and 99mTc-antimony trisulfide (3–30 nm). New radiotracers have
that also receives most of the lymph draining from the mammary gland. recently been developed for SLNM, most notably 99mTc-til-
(b) Anatomic drawing of the lymphatic pathways and lymph node sta-
tions draining the mammary gland. Most of the lymph produced in the manocept (or Lymphoseek®), which is composed of a dextran
breast drains to the subareolar (or Sappey’s) plexus before merging with backbone with multiple mannose residues and linked to DTPA
the subcutaneous plexus of the overlying skin and then flowing mostly to for 99mTc-labeling. Accumulation of this tracer in lymph nodes
the axilla. Lymph from deeper portion of the mammary gland drains is not due to nonspecific phagocytosis by macrophages (as is the
either through the same pathway or through deep lymphatics that reach
the parasternal lymph nodes (internal mammary chain) and even the case for traditional 99mTc-radiocolloids) but to a ligand-receptor
contralateral side (reproduced with permission from: Mariani G, type of binding to the CD206 antigen expressed on the surface
Moresco L, Viale G, Villa G, Bagnasco M, Canavese G, et al. Radioguided of macrophages and of dendritic cells present in lymph nodes.
sentinel lymph node biopsy in breast cancer surgery. J Nucl Med. Such binding is non-saturable (at least with the amounts of
2001;42:1198–215)
radiotracer employed for SLNM) and results in a much more
stable and prolonged retention in LNs than the phagocytosis
16.2.3 Procedures for Sentinel Lymph Node mechanism that radiocolloids undergo. The potential advan-
Mapping/Biopsy tages of its small molecular size (equivalent to 7.1 nm) and the
receptor-targeted nature of the mannose moieties in 99mTc-
Generally, SLNM and SLNB involve a continuum of proce- tilmanocept include rapid transit from the primary site to the
dures consisting in (1) interstitial tracer injection, (2) preopera- SLN as well as selective accumulation in that node, with limited
tive scintigraphic imaging, and (3) intraoperative gamma probe pass-through to second-echelon nodes [16].
Activities and Volumes Lymphatic circulation within solid tumors (including

Published evidence supports the use of small volumes with breast cancer) is generally grossly abnormal, disrupted, and
high specific activity to improve SLN detection. Consensus inefficient. Thus, intratumoral administration requires the
on the activity to be administered in a SLN procedure has not injection of high activities (up to 370 MBq) and large vol-
been reached. In current practice, a total injected activity of umes (up to 4 mL). Injection of such large volumes can sub-
5–30 MBq (depending on the elapsed time between scintig- stantially increase interstitial pressure at the injection site,
raphy and surgery) is generally considered sufficient for sur- thus possibly altering the pattern of lymphatic drainage and
gery planned for the same day. Prior-day injection has been forcing drainage routes different from those prevailing in
shown to be technically feasible by adequately increasing the baseline conditions. Furthermore, most of the injected radio-
amount of radioactivity injected (up to 150 MBq), to com- activity is retained at the injection site, often causing interfer-
pensate for physical decay in the intervening period between ence in SLN detection caused by the important “shine-through”
tracer injection and intraoperative detection. effect during imaging as during intraoperative detection.
Finally, slow drainage from the tumor can cause poor scinti-
Injection Procedure graphic visualization of the lymphatic channels and possibly
The injection site/modality is another controversial issue. lead to failure of lymphoscintigraphic imaging and of intra-
The most commonly used injection sites can be classified operative SLN identification. All the above considerations
into two categories: deep (intratumoral and peritumoral) and explain why, at present, most nuclear medicine centers prefer
superficial (intradermal, subdermal, subareolar, and periare- peritumoral or superficial injections rather than the originally
olar) injection (Fig. 16.4). proposed intratumoral route of administration.
a b c
d e f
Fig. 16.4 Modalities of interstitial injection of the lymphatic mapping Olmos RA, Vidal-Sicart S, Giuliano AE, Mariani G. Radioguided sur-
agent for SLN mapping in patients with breast cancer. Superficial injec- gery for breast cancer. In: Strauss HW, Mariani G, Volterrani D, Larson
tions (a–d) and deep injections (e, f): intradermal (a), subcutaneous (b), SM, eds. Nuclear oncology—From pathophysiology to clinical applica-
subareolar (c), periareolar (d), peritumoral (e), and intratumoral (f) tions. New York, NY: Springer; 2017:1363–1400)
(reproduced with permission from: Giammarile F, Orsini F, Valdés
Peritumoral injection, performed by depositing two ali- ity control on the use of the appropriate tracer, failure of the
quots on each side of the tumor, is considered the gold stan- injection, failure of the radiopharmaceutical, and manage-
dard for accurate SLN detection, because the tracer is ment of the appropriate breast and axilla (injection of the
injected near the same lymph vessels draining the tumor and proper side, left or right breast). Reasons not to use preopera-
is able to reveal both extra-axillary and axillary drainage. tive lymphoscintigraphy are logistical or because there is no
However, this approach has been criticized especially in case definite evidence of a higher intraoperative success rate in
of non-palpable and multicentric tumors. The use of peritu- the harvesting of axillary SLNs.
moral injections requires careful investigation of a patient’s Although dynamic imaging acquired at a 1-min frame
prior imaging and medical records, particularly if the tumor rate can provide information useful to SLN localization, this
is non-palpable. modality is not often used in SLN procedures for breast can-
Intra-/subdermal administration is performed on the skin cer, whereas it has a greater role in other malignancies such
overlying the tumor, in such a way so as to produce a small as cutaneous melanoma (see further below in this chapter).
wheal. More than one injection could be performed in adja- Thus, the standard technique for lymphatic mapping in
cent sites. Periareolar administration is generally performed patients with breast cancer relies on sequential planar images
with two to four injections, each one at the edge of the areola acquired with an adequate delay after interstitial injection of
at Sappey’s plexus. The rationale of such administration the radiopharmaceutical, complemented with SPECT/CT
stems from the fact that lymph (therefore also the radiocol- imaging—when indicated. Delay of acquisitions after radio-
loid) is drained from the intra-/subdermal space to the subcu- tracer injection varies according to the radiopharmaceutical
taneous plexus, which is the merging point for the lymph used (smaller particles migrate from the injection site faster
originating from the underlying breast parenchyma. than larger particles), the injection modality (speed of migra-
Superficial injection sites have numerous advantages, includ- tion progressively increases passing from intratumoral to
ing simplicity, shorter time between injection and SLN iden- peritumoral, subcutaneous, and intra-/subdermal injection,
tification, and increased radiotracer uptake in LNs, resulting respectively), and the patient’s characteristics (lymphatic
in improved nodal identification rates. Nonetheless, it should drainage can be slower in elderly or overweight patients).
be considered that superficial injection allows almost exclu- Lymphatic drainage and lymph node visualization is usually
sive identification of axillary nodes. observed already 20–30 min after superficial injection, com-
Despite the discrepancies mentioned above, it is common pared to 2–3 h following intratumoral injection. The amount
experience that all injection modalities enable axillary SLN to of radiocolloid migrated to LNs represents about 1% of the
be identified accurately, and satisfactory SLN detection rates injected activity after superficial administration, whereas it is
have been reported for all injection approaches. Thus, if the about 0.1% after peritumoral administration.
goal of SLNM/SLNB is axillary staging only, superficial tracer A single-head or dual-head large field-of-view (FOV)
injection (periareolar, subareolar, subdermal, intradermal) may gamma camera is generally used for imaging, using low-
be preferable to a deep injection (peritumoral, intratumoral), energy, high-resolution collimators and centering the energy
due to better and quicker visualization of axillary SLN. window on the 140 keV photopeak of 99mTc (±5% or 10%).
Nevertheless, it should not be neglected that deep tracer Planar images are usually acquired at 15–30 min, then
injection results in improved detection of extra-axillary 2–4 h postinjection, and as needed thereafter up to 18–30 h
SLNs: after peritumoral administration, lymphoscintigraphy (in case of 2-day protocol). At least two views are acquired,
shows drainage to the internal mammary chain in 20–30% of i.e., anterior and 45° anterior oblique (right or left depending
the cases, while this fraction is much lower (<3%) after on side of the tumor); a right or left lateral view is also rec-
intra-/subdermal or periareolar administration. The combi- ommended. The typical acquisition time is 3–5 min, and a
nation of both injection techniques (deep and superficial) in 2 mm pixel size is recommended (or anyway the smallest
the same patient may lead to an overall improvement in SLN pixel size available on the gamma camera), usually with
detection. 256 × 256 matrix size and zoom factor 1 (or, rarely, 128 × 128
matrix size with zoom factor 2).
Imaging Procedures Figure 16.5 shows different patterns of lymphatic drain-
Lymphatic mapping enables to determine the number of age and SLN visualization obtained by planar lymphoscin-
lymph nodes that are on a direct drainage pathway from the tigraphy. It should be noted that planar imaging alone does
tumor and thus to locate the SLNs. Preoperative imaging is not provide the exact anatomical location of the detected
strongly recommended due to variability in breast lymphatic SLNs. The crudest way to yield some anatomical correlates
drainage toward the axillary and extra-axillary LNs. Thus, is to define the body contour by moving a 57Co (or 99mTc)
preoperative lymphatic mapping has the potential to both point source along the patient’s body during image acquisi-
improve accuracy (especially in extra-axillary LNs) and tion (Fig. 16.6). Alternatively, by placing a 57Co flood source
reduce morbidity relative to the use of handheld gamma beneath the patient’s body, a simultaneous emission and
probes alone [14]. Preoperative imaging also serves as qual- transmission image can be acquired where delineation of the
Fig. 16.5 Different patterns of lymphatic

drainage visualized during lymphoscintigraphy
a b
acquired 30–60 min after intradermal injection of
99m
Tc-nanocolloidal albumin in patients with
breast cancer. (a) Single lymphatic vessel leading
to a single SLN, with faint visualization of
higher-tier lymph nodes (right anterior oblique
view). (b) Two separate lymphatic vessels widely
diverging in their initial pathway and leading to
two separate but adjacent SLNs, with faint
visualization of higher-tier nodes (left anterior
oblique view). (c) Three separate lymphatic
vessels widely diverging in their initial pathway,
leading to two separate but very close SLNs (left
anterior oblique view). (d) Multiple lymphatic
vessels leading to at least three separate SLNs
(reproduced with permission from: Mariani G, c d
Moresco L, Viale G, Villa G, Bagnasco M,
Canavese G, et al. Radioguided sentinel lymph
node biopsy in breast cancer surgery. J Nucl
Med. 2001;42:1198–215)
Fig. 16.6 Delineation of body contour obtained by moving a 57Co panel) visualize migration of the radiocolloid to a single SLN in the left
point source along the body of the patient during acquisition of the pla- axilla (reproduced with permission from: Giammarile F, Orsini F,
nar images in a patient with cancer of the left breast; although appearing Valdés Olmos RA, Vidal-Sicart S, Giuliano AE, Mariani G. Radioguided
as a single large radioactive area in the breast, 99mTc-nanocolloidal albu- surgery for breast cancer. In: Strauss HW, Mariani G, Volterrani D,
min was injected at four spots periareorally. Images acquired in the Larson SM, eds. Nuclear oncology—From pathophysiology to clinical
anterior view (left panel) and in the left anterior oblique view (right applications. New York, NY: Springer; 2017:1363–1400)
body contour is based on attenuation of the 57Co γ-rays pass- provided by the fused SPECT/CT images provides useful
ing through the patient’s body (Fig. 16.7). By combining preoperative complementary information resulting in better
tomographic functional images registered with anatomic anatomical SLN localization (Fig. 16.9), reduced surgical
data from CT, SPECT/CT imaging provides better contrast time, and greater confidence of the surgeons with the tech-
and resolution than planar imaging and has the possibility to nique. However, the added value of SPECT/CT imaging
correct for attenuation and scatter (Fig. 16.8). In particular, seems to be limited to a small fraction of the breast cancer
fused SPECT/CT imaging considerably improves the topo- patients undergoing SLNB and is associated with additional
graphic SLN localization within an anatomical landscape, costs and requires extra time for imaging. For these reasons,
thus providing a valuable road map to the surgeon. It has also specific indications for the use of SPECT/CT imaging should
been shown that SPECT/CT can detect in a substantial num- be defined, so that the majority of patients who will not ben-
ber of patients additional SLNs not visualized on planar efit from this imaging technique are spared unnecessary
images. In the majority of cases, the anatomic information costs and inconvenience.
a b
Fig. 16.7 Delineation of the body contour during acquisition of planar SLNs in the right axilla and right internal mammary chain (reproduced
lymphoscintigraphy in a patient with cancer of the right breast. (a) The with permission from: Giammarile F, Orsini F, Valdés Olmos RA, Vidal-
57
Co flood source is placed opposite to the gamma camera head (black Sicart S, Giuliano AE, Mariani G. Radioguided surgery for breast can-
arrows in the lower part of the panel) during acquisition of the anterior cer. In: Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear
planar image. (b) Lymphoscintigraphy obtained after intratumoral oncology—From pathophysiology to clinical applications. New York,
injection of 99mTc-nanocolloidal albumin depicts lymphatic drainage to NY: Springer; 2017:1363–1400)
Fig. 16.8 Left panel: geometry of the gamma camera detectors for permission from: Giammarile F, Orsini F, Valdés Olmos RA, Vidal-
SPECT/CT acquisition in the same patient as in Fig. 16.7. Right panel: Sicart S, Giuliano AE, Mariani G. Radioguided surgery for breast can-
3D surface volume rendering of the SPECT/CT acquisition clearly cer. In: Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear
visualizing radiocolloid drainage to one right parasternal SLN (with a oncology—From pathophysiology to clinical applications. New York,
higher-tier lymph node) and to one right axillary SLN (reproduced with NY: Springer; 2017:1363–1400)
Key Learning Points periareolar) is currently the preferred modality in

• Radiopharmaceuticals employed for lymphatic most centers.
mapping in patients with breast cancer include • Preoperative lymphoscintigraphy is an integral com-
99m
Tc-radiocolloids with variable particle size (that ponent of the whole lymphatic mapping procedure.
are retained in lymph nodes because of nonspecific • Planar scintigraphy using variable angle views is
phagocytosis by macrophages) and 99mTc- generally sufficient for adequate visualization of
tilmanocept (which is retained in lymph nodes axillary lymph nodes draining the breast.
because of binding to the CD206 antigen expressed • In patients with breast cancer, SPECT/CT imaging
by macrophages and dendritic cells). has a certain added value especially when planar
• Particle size of radiocolloids is inversely related to scintigraphy fails to adequately visualize the lym-
the rate of drainage from the site of interstitial phatic drainage pathways.
administration. • When acquiring planar scintigraphy only, definition
• Although the optimal modality of tracer injection is of the body contour yields useful anatomical corre-
still debated, superficial injection (intra-/subdermal lates for the lymph nodes visualized during
over the skin projection of the tumor, subareolar, or lymphoscintigraphy.
a b
c d e
Fig. 16.9 Anatomical SLN localization obtained with SPECT/CT in a axilla (c), as well as behind the left clavicle (e). The CT component of
patient with cancer of the left breast. Following intratumoral injection of the SPECT/CT acquisition (d and f) provides purely anatomical evalua-
99m
Tc-nanocolloidal albumin, planar imaging (a) visualizes lymphatic tion of the radioactive lymph nodes, as indicated by dashed yellow cir-
drainage to SLNs in the left axilla, periclavicular area, and internal cles (reproduced with permission from: Giammarile F, Orsini F, Valdés
mammary chain. Fused SPECT/CT imaging displayed with 3D surface Olmos RA, Vidal-Sicart S, Giuliano AE, Mariani G. Radioguided sur-
volume rendering (b) better localizes the anatomical correlates of SLNs gery for breast cancer. In: Strauss HW, Mariani G, Volterrani D, Larson
visualized on planar imaging. Transaxial fused SPECT/CT sections SM, eds. Nuclear oncology—From pathophysiology to clinical applica-
localize such SLNs in the second intercostal space and level I of the left tions. New York, NY: Springer; 2017:1363–1400)
16.2.3.2 Procedures in the Surgical Suite When a hot SLN has been removed, the surgical bed
should be checked to confirm its removal and to evaluate
Blue-Dye SLN Localization remaining activity. Finally, to minimize false-negative
There is general agreement that combined administration of results, the open axilla should be palpated and suspicious
radiocolloid and blue dye using either superficial injection or lymph nodes harvested, even if these are neither hot nor blue.
deep injection enhances SLN detection. The blue dye can be The use of dedicated small field-of-view gamma cameras
injected around the primary tumor 10–20 min prior to sur- facilitates assessment of complete removal of SLN(s).
gery in a volume of 2–5 mL. Within 5–15 min the SLNs are
colored. Washout is evident after approximately 45 min. SLN Non-Visualization or Failed Intraoperative
Disadvantages of using blue dyes are as follows: (1) Detection
impossibility to evaluate extra-axillary nodes, (2) tempo- In approximately 1–2% of the patients, SLNs will not be
rary blue tattooing of the skin or areola (for patients under- detected preoperatively or intraoperatively, and the status of
going breast-conserving surgery), and (3) induction of axillary LNs cannot be determined with this approach. Old age,
anaphylactic reactions (that require resuscitation in 0.5– obesity, tumor location other than in the upper outer quadrant,
1.0% of patients and that contraindicate its use in pregnant and non-visualization of SLNs on preoperative
women) [17]. lymphoscintigraphy may be associated with failed SLN
localization. Deeply located SLNs are difficult to detect
Radioguided Surgery intraoperatively because of tissue attenuation; furthermore, the
Intraoperative SLN detection is usually guided by the use large amount of radioactivity retained at the injection site may
of a handheld gamma probe that provides both a readout cause adjacent SLNs to be hidden during imaging and during
count rate and an audible signal which is proportional to the the intraoperative search because of the shine-through effect.
count rate. Such probes are designed specifically for intra- Patients who have been submitted to prior breast surgery
operative use, and are therefore be able to detect the SLN or radiation may demonstrate lymphatic drainage to LNs
from the skin surface as well as within the exposed surgical in locations not typically seen in patients without a history of
cavity. prior surgery. The lymphatic duct to the original SLN may be
In principle, SLNB requires the removal of all SLNs obstructed by tumor growth, or the original SLN may be
receiving direct lymphatic drainage from the site of the entirely replaced by metastatic disease. As a consequence,
primary tumor. In practice, this is not always achieved. In lymphatic drainage may be either diverted to a non-SLN, or
cases with multiple radiolabeled LNs, it is often difficult to no lymph nodes may be visualized.
distinguish between true SLNs and second-tier LNs. The It is important to note that when a SLN is not detected
issue of how many LNs should be harvested when multiple intraoperatively, this corresponds to a failure of the method
radioactive LNs are found is still debated. In this regard, and not to a false-negative case (i.e., when axillary metastatic
while removing too few LNs may lead to miss potential relapse is observed despite a negative SLNB).
metastases in regional LNs, indiscriminate removal of all The majority of patients with preoperative lymphoscinti-
radioactive axillary LNs may cause morbidity similar to graphic SLN non-visualization will have at least one SLN
that experienced after conventional ALND (in addition to detected intraoperatively, either by a gamma probe alone or by
the unnecessarily increased burden for histopathologic a gamma probe combined with blue dye. While logistically dif-
analysis). Several operational definitions of the SLN have ficult in most centers, a second radiotracer injection, perhaps
evolved over time in order to decide exactly which LNs with a different injection site/modality, may be useful to visual-
should be removed to maximize the likelihood of locating ize previously non-visualized SLNs. In many patients in whom
the “true” biologic SLN and to minimize unnecessary planar lymphoscintigraphy does not result in SLN visualization,
removal of non-SLNs. Some authors base SLN identifica- SPECT/CT imaging does result in SLN detection.
tion on the absolute number of counts/sec recorded for the Although there is no definitive consensus on what to do if
presumed nodes, while others consider the ratio of the “in a SLN cannot be visualized, the current standards of care
vivo” or “ex vivo” radioactive counts in the SLNs relative recommend ALND when intraoperative SLN identification
to background or to neighboring non-SLNs. Empiric is not achieved.
thresholds are widely employed, corresponding to (1) 10% Current recommendations regarding the use of SLNB in
or 20% of the count rate in the first LN removed (which is patients with breast cancer as jointly defined by the EANM
usually the most radioactive) or (2) at least tenfold the and the SNMMI in their practice guidelines are summarized
background count, taken at a location remote from the in Table 16.1 [14].
injection site.
Table 16.1 Recommendations regarding use of sentinel lymph node

biopsy in patients with breast cancer according to the joint EANM and • In about 1–2% of patients with breast cancer, the
SNMMI practice guidelines
sentinel lymph node cannot be detected on preop-
Clinical condition Use of sentinel lymph node biopsy erative lymphoscintigraphy or intraoperatively with
T1 or T2 tumor Established gamma probe guidance.
Older age Established
• Factors associated with failure of radioguided senti-
Obesity Established
nel lymph node biopsy include old age, obesity,
Male breast cancer Established
DCISa with mastectomy Established tumor location, and non-visualization on preopera-
Before neoadjuvant systemic Established tive lymphoscintigraphy.
therapy
Multicentric/multifocal tumor Controversial
T3 or T4 tumor Controversial
DCISa without mastectomy Controversial (except for DCIS
with suspected or proven
16.2.4 Radioguided Occult Lesion Localization
microinvasion) (ROLL) with 99mTc-MAA
Suspicious, palpable axillary Controversial
lymph nodes Hook-wire localization of non-palpable lesions has been for
After neoadjuvant systemic Controversial many years the most widely used preoperative technique.
therapy
Although this is a reasonably effective technique, it involves
Pregnancy Controversial
some disadvantages. First, the entry site of the wire is often not
Evaluation of internal Controversial
mammary lymph nodes at the ideal location for surgical incision at the time of surgery.
Prior diagnostic/excisional Controversial This may lead to additional unnecessary dissection and subop-
breast biopsy timal cosmetic results. The most important disadvantage,
Prior axillary surgery Controversial however, is the inaccuracy of localizing the target lesion per-
Prior nononcologic breast Controversial cutaneously and during dissection. This results in non-negligi-
surgery
ble rates of reoperation after histologic demonstration of
Inflammatory breast cancer Not recommended
tumor involvement of the margins of the surgical specimen.
Controversial indications are those for which SLN biopsy is not univer-
sally accepted or for which the evidence behind the practice is limited
Intraoperative ultrasound imaging without preoperative
or entirely missing wire localization has also been used to map excision of non-
Modified from Ref. [14] palpable breast lesions; however, this technique has limita-
a
DCIS ductal carcinoma in situ tions, as it is feasible only in patients whose breast lesion is
detectable at ultrasound imaging.
The “radioguided occult lesion localization” (ROLL)
Key Learning Points approach has gained popularity for non-palpable tumor
• Radioguidance provided by preoperative lymphos- lesions, including breast cancer [18]. ROLL involves the
cintigraphy and by intraoperative gamma probe intralesional injection of a small amount of radioactive tracer
counting can be complemented by visual guidance that does not migrate from the site of interstitial injection,
based on administration of blue dye. typically 99mTc-MAA. Injection is performed on the same
• Biopsy of the sentinel lymph node(s) is guided day or on the day before surgery, under mammographic or
intraoperatively by a handheld gamma probe pro- ultrasound guidance, with injected activities ranging from 2
viding an instantaneous count rate. to 15 MBq. Surgeons identify the lesion intraoperatively as a
• Radiocolloids with small particle size might not be hot spot by using a handheld gamma probe, which allows
retained efficiently in the first (sentinel) lymph node accurate localization and removal of the lesion with minimal
encountered along a certain pathway of lymphatic excision of healthy tissue.
drainage, thus possibly visualizing multiple lymph ROLL is a well-tolerated and feasible technique for local-
nodes along that pathway. izing early-stage breast cancer during breast-conserving sur-
• Although the issue is still debated, different param- gery and is a suitable replacement for wire-guided
eters can be considered to decide which lymph localization. Reported advantages of the ROLL technique
nodes should be removed to maximize the likeli- include (1) easy and precise intraoperative localization of the
hood of harvesting the “true” sentinel lymph node breast lesion; (2) complete lesion resection, with free mar-
for analysis. gins and reduced needs for second operations in most of the
cases; (3) improved centering of the lesion within the surgi-
cal specimen; and (4) possibility to choose the most suitable pathways of lymph drainage and to guide intraoperatively
surgical approach (skin incision) minimizing surgical aggres- for the radioguided search of SLN(s). Most of the injected
siveness and optimizing the cosmetic results of surgery. activity actually remains retained at the site of interstitial
injection, thus making it possible to perform a ROLL proce-
dure using the same handheld gamma probe used for radiogu-
Key Learning Points ided SLNB.
• Conventional techniques for localization of non- Therefore, as a single combined procedure for intraopera-
palpable breast lesions are affected by definite tive localization of the breast lesion and for SLNB, the SNOLL
drawbacks that limit their accuracy and/or result in technique may improve the entire surgical procedure. The
suboptimal cosmetic outcome. majority of the studies published so far show a high percentage
• Radioguided occult lesion localization based on of successful tumor resection and intraoperative SLN
intralesional injection of radiolabeled particles that localization, with minimal failure rates [20].
do not appreciably move from the injection site When using the radioactive seeds for ROLL, the energy
(such as 99mTc-MAA) has gained popularity because window of the handheld gamma probe can be switched
of its high feasibility, overall accuracy, and optimal between the 27 keV energy peak of the 125I source and the
cosmetic results. 140 keV energy peak 99mTc, thus permitting to discriminate
γ-ray emissions of the two radionuclides.
16.2.4.1 ROLL with Radioactive Seeds Key Learning Points

Alternatives to hook-wire localization of occult breast • Intralesional placement of radioactive seeds in the
lesions include carbon trace as well as the use of sealed form of tiny sealed sources containing 125I is an
radioactive seeds. The seeds used to this purpose are similar alternative to intralesional injection of 99mTc-MAA
to those used for brachytherapy in patients with prostate can- for radioguided occult lesion localization.
cer, each seed consisting of a 4.5 × 0.8 mm titanium capsule • The procedure of radioguided sentinel lymph node
containing a tiny ceramic cylinder enriched with 125I. biopsy can be combined in the same patient with
Iodine-125 has a long decay time (half-life of 59.4 days) and the procedure of radioguided occult lesion
emits low-energy γ-photons (27 keV). The radioactive seed localization.
is placed in the center of the breast lesion using an 18 G • Different combinations of radioactive agents have
needle fixed in a needle holder under mammographic or been proposed for simultaneous radioguided senti-
ultrasound guidance. During surgery, excision of the lesion nel lymph node biopsy and radioguided occult
is guided by using a handheld gamma probe. lesion localization.
It has been shown that radioguided seed localization in
non-palpable breast lesions is at least equivalent to the hook-
wire technique in terms of ease of procedure, reliable resec-
tion of the target lesion, volume of breast tissue excised, and 16.2.5 Added Value of Intraoperative Portable
histologically negative tissue margins; furthermore, the pro- Gamma Cameras
cedure can be combined with the standard technique of
radioguided SLNM and SLNB [19]. Recently, several types of portable or handheld mini gamma
cameras have become available for clinical practice; while
16.2.4.2 Sentinel Node and Occult Lesion some of these portable gamma cameras are not specifically
Localization (SNOLL) designed for radioguided surgery, other models are focused
Different techniques have been described to identify the on different applications of SLNB [21].
SLNs in combination with ROLL, the so-called “sentinel Appropriate perioperative use of a portable gamma cam-
node and occult lesion localization” (SNOLL) procedure. In era enhances the reliability of the entire procedure of
particular, intratumoral injection of 99mTc-MAA for ROLL radioguided surgery, by providing high-resolution imaging
may be associated with subdermal injection of 99mTc- of the surgical field. The use of a perioperative imaging
nanocolloid for radioguided SLNM and SLNB. Another pos- device implies the possibility to better plan the surgical
sibility is to use a single intratumoral injection for both approach, to localize surgical targets in complex anatomical
ROLL and SNOLL in the same session. In fact, following areas just before making the surgical incision, to monitor
interstitial injection of even small-particle radiocolloids, the lymphatic basin before and after removal of the hot
only a minor fraction of the injected activity migrates through lymph nodes, and, above all, to verify completeness of SLN
the lymphatic channels, thus making it possible to image the excision.
16.2.6 Future Perspectives

Key Learning Points
Some of the competitive/complementary modalities that are • Portable small field-of-view gamma cameras can be
emerging for different applications of image-guided surgery employed intraoperatively for high-resolution
are particularly interesting, such as those based on indocyanine imaging of the surgical field.
green fluorescence, contrast-enhanced ultrasound with micro- • Intraoperative imaging per se does not guide search
bubbles, and superparamagnetic iron oxide nanoparticles. for the sentinel lymph node(s), but rather it is
Although such novel approaches have clinical potential, employed for assessing completeness of radiogu-
they are generally burdened by relatively high failure rates ided sentinel lymph node excision.
and/or false-negative results and at present cannot challenge • Newer and/or alternative approaches to sentinel
the existing standard procedures [22]. lymph node biopsy are based on indocyanine green
Particular attention is also being paid to the development fluorescence, contrast-enhanced ultrasound with
of hybrid imaging agents that in principle combine the microbubbles, and superparamagnetic iron oxide
advantages of a multimodality approach, such as radioactive nanoparticles.
and fluorescent agents for SLNM (Fig. 16.10). Important • Hybrid imaging agents (e.g., radioactive and fluo-
advances in this technology lead to the development of rescent) that combine the advantages of multimo-
hybrid intraoperative devices capable to detect simultane- dality intraoperative detection are currently under
ously both an optical signal (i.e., the fluorescent guide) and a clinical validation.
radioactive signal (i.e., intraoperative radioguidance).
a b c
d e f g
Fig. 16.10 Resection of an infraclavicular SLN in a patient with high-risk guidance (e). Ex vivo control of the fluorescence signal is finally performed
breast cancer using the hybrid tracer ICG-99mTc-nanocolloidal albumin. (f and g) (reproduced with permission from: Giammarile F, Orsini F, Valdés
The SLN is first located on the skin projection using a portable gamma Olmos RA, Vidal-Sicart S, Giuliano AE, Mariani G. Radioguided surgery
camera (a) and a handheld gamma probe (b). A portable near-infrared cam- for breast cancer. In: Strauss HW, Mariani G, Volterrani D, Larson SM,
era (c) is then used to depict the fluorescence signal on the screen (d). This eds. Nuclear oncology—From pathophysiology to clinical applications.
enables to remove the SLN under combined radioguidance and optical New York, NY: Springer; 2017:1363–1400)
16.3 R
adioguided Sentinel Lymph Node umbilicus to the level of the second lumbar vertebra, stating
Biopsy in Cutaneous Melanoma that lymph channels did not cross these lines. This concept
(the so-called Sappey’s rule) was universally accepted over the
16.3.1 The Clinical Problem next almost 100 years. After the introduction of clinical lym-
phoscintigraphy in the mid-twentieth century, it was frequently
The incidence of melanoma is growing worldwide. Currently, observed that Sappey’s rules did not always prove to be correct
the estimated incidence in Europe is 9 out of 100,000 inhab- (Fig. 16.11). In particular, there were “zones of ambiguity” on
itants, with a mortality of 2.3/100,000. Every year 4% more each side of Sappey’s lines where the direction of lymphatic
people are diagnosed with the disease, thus making mela- drainage was impossible to predict. Therefore, clinicians
noma the cancer with the greatest increase in incidence. began to use lymphoscintigraphy in patients with melanomas
Regional lymph nodes are frequently the first site of metas- located in those ambiguous areas to identify lymphatic basins
tasis before systemic spreading of the disease. The presence or potentially at risk to receive metastatic tumor cells.
absence of lymph node metastases, even in the form of micro- Since lymphatic cutaneous drainage is highly variable, lym-
metastasis, is an important prognostic factor in early-stage phoscintigraphy is currently a mandatory component of the
melanoma patients [23]. Occult metastases in the regional radioguided SLNB procedure. The reproducibility of this tech-
nodes (clinically non-palpable and difficult to identify also nique is very high, ranging from 84% to 96% in several studies.
with ultrasonography) are found in nearly 20% of patients The predictability of lymphatic drainage in cutaneous mel-
who present with melanoma with Breslow thickness >1 mm. anoma depends on the location of the primary lesions, being
Therefore, accurate histopathological evaluation is required approximately 98% in the lower limbs, 88% in the upper
for early detection of metastasis. In the past, regional lymph- extremities, 56% in the anterior thorax, and 39% in the poste-
adenectomy was routinely performed to stage clinically node- rior trunk. Lymphatic drainage is almost completely unpre-
negative melanoma patients. However, this procedure very dictable in the head and neck region. Nevertheless, learning
often constituted a clear, unnecessary overtreatment since the the physiologic and “lymphoscintigraphic” drainage patterns
majority of patient (usually >75%) with <4 mm Breslow constitutes at least a first estimation of the most likely draining
thickness melanomas were found not to have lymph node basins, depending on the location of the primary tumor.
metastases [24, 25]. Therefore, de novo lymphadenectomy
exposed many patients to unnecessary immediate and long-
term surgical complications, such as possible wound infection Key Learning Points
and delayed wound healing, lymphedema, hematoma, seroma, • In patients with cutaneous melanoma, sentinel
pain, etc. Due to the above considerations and to recent devel- lymph node biopsy has replaced de novo lymphad-
opments of cancer surgeries toward less aggressive proce- enectomy, which exposes patients to unnecessary
dures, over the last 20 years, lymphadenectomy has gradually immediate and long-term surgical complications
been replaced with SLNB, a minimally invasive method based and turns out to be futile surgery in >75% of patients
on the concept of “orderly” progression of metastatic cells with <4 mm Breslow thickness melanomas and a
through the lymphatic system that allows accurate assessment clinically negative lymph node basin.
of the lymph node status [26] without the burden of complica- • Widespread use of lymphoscintigraphy has demon-
tions associated with lymphadenectomy [27]. strated failure of the so-called Sappey’s rule to
accurately predict the pathway of lymphatic drain-
age for melanomas located in ambiguous areas
16.3.2 Lymphatic System of the Skin around Sappey’s lines (midline in the front and
back and horizontal line around the waist).
The skin has a very dense network of lymphatic capillaries • High intersubject variability in the patterns of lym-
and vessels. The paths followed by the collecting vessels on phatic cutaneous drainage makes the use of preop-
their way to the lymphatic basins vary from patient to patient erative lymphoscintigraphy a mandatory component
and from a skin location to another. These pathways can of radioguided sentinel lymph node biopsy in
sometimes be very complex and unexpected. In general, patients with cutaneous melanoma.
lymphatic vessels converge to form larger vessels or multiple
vessels (as, e.g., in the upper thigh). These collecting vessels
pass through the subcutaneous fat layer and penetrate into
the deep fascia after reaching the lymph node basin. 16.3.3 Tracer Injection
The pattern of lymphatic drainage from the skin has been
investigated with various modalities over the past centuries, Usually the radiotracer is administered intradermally with
mostly with postmortem studies. In the nineteenth century, four or more aliquots (depending on the anatomic site)
Sappey defined lines that passed down the midline front and injected around the primary melanoma or the excisional
back, as also along a horizontal line around the waist from the biopsy site. According to the EANM guidelines [28], after
Fig. 16.11 On the left, Metastatic dissemination Midline areas with ambiguous lymphatic drainage
metastatic dissemination in
melanoma: N1, metastasis in Sappey’s Concept
draining lymph node basin; Lymphoscintigraphy

M1a, distant skin,
subcutaneous, or nodal
metastases; M1b, lung
metastases; and M1c, all other
visceral metastases. On the
right, midline areas with
ambiguous lymphatic
drainage according to the
M1b
classical skin watershed
Sappey’s concept determined
postmortem (blue) and its
M1c
actual extension according to
recent lymphoscintigraphy
findings (light brown)
M1a
from: Vidal-Sicart S, Orsini F,
Giammarile F, Mariani G,
Valdés Olmos
N1
RA. Radioguided surgery for
malignant melanoma. In:
Strauss HW, Mariani G,
Volterrani D, Larson SM, eds.
Nuclear oncology—From
pathophysiology to clinical IN-TRANSIT
applications. New York, NY: METASTASIS
Springer; 2017:1401–32)
<2cm SATELLITES
PRIMARY LESION
injection into the dermis (within 1 cm from the melanoma or reflects the actual lymphatic drainage in that particular patient
the excisional biopsy scar), the volume of radiotracer should and enables to visualize the lymphatic basin at risk of metasta-
raise a wheal. To avoid contamination of the patient’s skin ses draining from the specific region of the skin where the
during the injections, the needle must be inserted in a tangent melanoma is located [29].
direction relative to the skin surface. All possible drainage regions must be covered during
The radiotracer may be injected the day before surgery or image acquisition. In the case of cutaneous melanoma,
on the same day of surgery (2-day or 1-day protocol, respec- dynamic lymphoscintigraphy is crucial to identify regional
tively). The injected activity varies from 5 MBq up to lymphatic basins and to distinguish the true SLN from higher-
120 MBq depending on the time elapsed between echelon, non-SLNs that can be visualized along the same
lymphoscintigraphy and surgery. The injected volume must lymphatic duct. After the dynamic series, static early and
be small (0.1–0.2 mL per aliquot) to avoid lymphatic col- delayed images are acquired, 30 min and 2 h postinjection,
lapse caused by sudden rise of interstitial pressure. respectively (Fig. 16.12). When using a lymphatic mapping
agent with fast migration, such as 99mTc-tilmanocept, SLNs
can be clearly visualized as early as 15–20 min postinjection
16.3.4 Lymphoscintigraphy (Fig. 16.13). Finally, the site of each SLN visualized during
lymphoscintigraphy must be marked on the skin with the aid
After injection, lymphatic imaging is performed to ascertain of a 57Co or 99mTc point source. The study is completed by
adequate drainage and uptake in the SLNs. Lymphoscintigraphy performing external counting with a handheld gamma probe
a b
c d
Fig. 16.12 Lymphoscintigraphy obtained in a patient with melanoma the aforementioned lymph node, as well as in other nodes upstream.
of the right lower thigh at sequential times (30 min and about 2 h, Second-echelon nodes are more frequently observed in the groin than in
respectively) after perilesional injection of 99mTc-nanocolloidal albu- other parts of the body, due to the particularly high velocity of lymph
min. Early (a and b) and delayed (c and d) planar images in anterior flow that occurs in the lower limb. In these cases, acquiring only delayed
(left panel) and oblique views (right panels). In the early images, two images would not be accurate, because it is difficult to distinguish
well-depicted lymphatic channels directly connect the injection site between the real SLNs and second-tier lymph nodes (reproduced with
with SLNs in right groin. The yellow arrow in (a) points to a faint lym- permission from: Vidal-Sicart S, Orsini F, Giammarile F, Mariani G,
phatic channel cranial to one of the two lower SLNs, leading to an addi- Valdés Olmos RA. Radioguided surgery for malignant melanoma. In:
tional lymph node, which could be a second-echelon node, but also a Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear oncol-
real SLN if this channel passes below the first depicted SLN directly ogy—From pathophysiology to clinical applications. New York, NY:
from the injection site. The yellow arrow in (c) indicates high uptake in Springer; 2017:1401–32)
99mTc-TiImanocept in melanoma
LAO 20 min Lateral 20 min Lateral 24 h
Fig. 16.13 Lymphoscintigraphy with 99mTc-tilmanocept obtained in a and the SLN to background count ratio was >30:1 intraoperatively
patient with melanoma of the left forearm. Planar images acquired with (modified from: Sondak VK, King DW, Zager JS, Schneebaum S, Kim J,
the left arm raised, site of injection remaining visible on the top of each Leong SP, et al. Combined analysis of phase III trials evaluating [99mTc]
image. A single axillary lymph node was visualized 20 min postinjec- tilmanocept and vital blue dye for identification of sentinel lymph nodes
tion, along with some activity in the draining lymphatic pathway. in clinically node-negative cutaneous melanoma. Ann Surg Oncol.
Radioactivity in the SLN visualized at 20 min was well retained at 24 h, 2013;20:680–8)
on the lymphatic basin visualized during lymphoscintigraphy,

a b
in order to confirm the exact projection on the skin surface of
the SLN location [30, 31].
By depicting the lymph node stations at risk for metasta-
sis, lymphoscintigraphy constitutes a lymphatic road map for
surgeons. SPECT/CT added to planar lymphatic mapping
allows a more accurate anatomical localization of SLNs and
frequently detects SLNs not seen on planar imaging
(Fig. 16.14); this imaging technique also reduces false-
positive findings such as skin contamination.
By providing anatomical landmarks to be recognized by
c d
the surgeon in the operation room, it has been recognized
that SPECT/CT improves and optimizes the overall surgical
procedure (Fig. 16.15). In principle, SPECT/CT is
recommended in all patients with melanoma as a comple-
mentary modality to planar imaging and is becoming manda-
tory in melanomas of the head and neck, of the trunk, or in
any areas with high probability of unexpected lymphatic
drainage [32–34].
Key Learning Points

• The radiopharmaceutical for lymphatic mapping
(either 99mTc-radiocolloid or 99mTc-tilmanocept) is
e f injected intradermally in multiple aliquots around
the primary melanoma or (more frequently) around
the excisional biopsy scar.
• In patients with cutaneous melanoma, lymphoscin-
tigraphy usually includes an early dynamic acquisi-
tion that helps to identify regional lymphatic basins
and to distinguish the true sentinel lymph node(s)
from higher-echelon lymph nodes that can be visu-
alized along the same pathway of lymphatic
drainage.
• Delayed static images acquired at 30 min and 2 h
postinjection are generally sufficient for adequate
preoperative imaging.
• SPECT/CT imaging provides an added value over
Fig. 16.14 Lymphoscintigraphy obtained after pericicatricial injection
of 99mTc-nanocolloidal albumin in a patient operated because of mela-
planar imaging, for more accurate anatomical local-
noma in the left posterior parietal scalp. (a) Anterior and (b) left oblique ization of radioactive sentinel lymph node(s).
planar views show several hot spots in the occipital and left cervical • SPECT/CT imaging is highly recommended in
area. (c) 3D surface volume rendering after SPECT/CT acquisition. (d) patients with melanomas of the head and neck
Skin marks in the areas where SLNs were considered to be. (e) Sagittal
fused SPECT/CT slice corresponding to the focus of uptake in the
region, of the trunk, or in areas with high probabil-
occipital area. (f) Transaxial fused SPECT/CT slice showing two sepa- ity of unexpected lymphatic drainage.
rate foci of radiocolloid uptake in this area. The focus with faint uptake
indicated by the green asterisk was a very tiny SLN (3 mm) that was
positive for metastasis. In this case SPECT/CT imaging constitutes an
added value over planar imaging alone that depicted only one hot spot
associated with a large-sized SLN in the occipital area, masking activity 16.3.5 Intraoperative Counting/Detection
in the adjacent tiny SLN (reproduced with permission from: Vidal-
Sicart S, Orsini F, Giammarile F, Mariani G, Valdés Olmos Intraoperative SLN detection is based on an acoustical and
RA. Radioguided surgery for malignant melanoma. In: Strauss HW,
Mariani G, Volterrani D, Larson SM, eds. Nuclear oncology—From
count-reading signal provided by the handheld gamma
pathophysiology to clinical applications. New York, NY: Springer; probe. Injection of a blue dye (patent blue, isosulfan, methy-
2017:1401–32) lene blue) may be employed to complement with intraopera-
a b
c d e
Fig. 16.15 Lymphoscintigraphy obtained after perilesional injection SPECT/CT acquisition, showing two tiny lymph nodes in para-aortic
of 99mTc-nanocolloidal albumin in a patient with melanoma located in and paravertebral areas (green lines). (e) Transaxial fused SPECT/CT
the left-central periumbilical area (yellow arrow in upper left panel). (a) slice confirming obvious radiocolloid uptake in these locations—cor-
The planar anterior view shows several hot spots in both axillae and responding to SLNs; due their deep location, these SLNs were not har-
para-costal areas and two hot spots central in the abdomen. All of them vested for analysis (reproduced with permission from: Vidal-Sicart S,
were considered as SLNs. (b) 3D surface volume rendering after Orsini F, Giammarile F, Mariani G, Valdés Olmos RA. Radioguided
SPECT/CT acquisition, showing the anatomical location of these surgery for malignant melanoma. In: Strauss HW, Mariani G, Volterrani
nodes. (c) Sagittal fused SPECT/CT slice centered on the foci of D, Larson SM, eds. Nuclear oncology—From pathophysiology to clini-
“abdominal” uptake. (d) Transaxial slice of the CT component of the cal applications. New York, NY: Springer; 2017:1401–32)
tive visual guidance the radioguidance provided by the department as the last phase of lymphoscintigraphy. After
radionuclide procedure. According to current recommenda- incision in the marked area, in vivo radioactivity is mea-
tions, all stained and/or radioactive lymph nodes retrieved sured in comparison with background activity (average
during surgery should be harvested for analysis. count rates of the surrounding lymph node basin). After
The standard device for intraoperative SLN detection is SLN identification, the gamma probe is used to assess the
a non-imaging, handheld gamma probe which enables to residual activity in the surgical field, so as to remove poten-
identify the area with the highest activity (counts per sec- tial additional SLNs.
ond) in each lymphatic region and its correspondence with Consensus has not yet been achieved on the best proce-
the skin mark(s) previously made in the nuclear medicine dure to define a count-rate threshold that would enable to
Fig. 16.16 Lymphoscintigraphy
obtained after pericicatricial injection a b
of 99mTc-nanocolloidal albumin in a
patient operated because of
melanoma of the left arm—just below
the shoulder. 3D surface volume
rendering images shown in (a) and (c)
depict the site of radiocolloid
injection and SLNs in the left axilla
and left supraclavicular area (a, c).
Transaxial fused SPECT/CT slices
shown in (b) and (d) depict exact
anatomical locations of such SLNs,
demonstrating that in the left axilla
there are actually two adjacent but
separate radioactive lymph nodes.
(e) Image produced by the freehand
SPECT system, where the radioactive c d
foci are displayed directly onto the
patient’s body (the infrared-based
tracking device positioned on the
patient’s body is clearly seen in the
right upper chest; a similar device is
fixed on the handheld gamma probe).
(f) The freehand SPECT device
allows virtual reconstruction
providing real-time information about
depth within the of the radioactive
foci (reproduced with permission
from: Vidal-Sicart S, Orsini F,
Giammarile F, Mariani G, Valdés
Olmos RA. Radioguided surgery for
malignant melanoma. In: Strauss
HW, Mariani G, Volterrani D, Larson
e f
SM, eds. Nuclear oncology—From
pathophysiology to clinical
applications. New York, NY:
Springer; 2017:1401–32)
identify the true SLNs and therefore to avoid unnecessary Recently, portable small FOV gamma cameras have been
harvesting of non-SLNs, based on the absolute number of introduced to improve intraoperative detection of SLNs in
counts in the radioactive lymph nodes. The ratio of in vivo addition to the handheld gamma probe. Although these
or ex vivo radioactive counts in the lymph node relative to devices are too bulky for direct handling in the exposed sur-
background has also been proposed to the same purpose. gical field, they allow real-time SLN visualization and there-
Finally, the 10% rule recommends that all lymph nodes fore the possibility to assess completeness of lymph node
counting 10% or higher than the ex vivo hottest SLN resection. These dedicated small FOV portable gamma cam-
should be harvested, whereas, according to other authors, eras can be used in combination with fluorescence cameras
a 20% threshold would be optimal for this purpose. for simultaneous dual-signature lymph node detection after
Whatever is the protocol adopted, the surgeon must har- injection of a hybrid (such as ICG-99mTc-nanocolloid) tracer
vest for analysis also any additional lymph node(s) in the for optical and radioactive guidance. Development of hybrid
surgical field that appear enlarged or suspicious on palpa- instrumentation for simultaneous detection of the radioactive
tion, even if they do not prove to be radioactive or blue- and the optical signal with a single small FOV imaging
stained [28]. device is also underway.
The freehand SPECT technique has recently been intro-

duced to allow for intraoperative 3D navigation by using an • Intraoperative high-resolution imaging with porta-
infrared system to track the handheld gamma probe relative to ble, small field-of-view gamma cameras does not
the patient’s body. While the surgeon is scanning the area of guide per se search for the sentinel lymph node(s),
interest with the probe, relative positions of the patient and of but rather it is employed for assessing completeness
the counting device are continuously being recorded by the of radioguided sentinel lymph node excision.
tracking system, thus permitting direct visualization of the • Hybrid imaging agents (e.g., radioactive and fluo-
distribution of radioactivity in the surgical field and providing rescent) that combine the advantages of multimo-
assessment of exact depth of the lesion (Fig. 16.16). dality intraoperative detection are currently under
clinical validation.
16.3.5.1 H arvesting and Analysis of Sentinel • Sentinel lymph node(s) harvested during surgery
Lymph Nodes must be analyzed using both conventional hema-
Most protocols discourage the use of intraoperative frozen toxylin and eosin staining and immunohistochemis-
section pathology to evaluate SLNs immediately after try with antibodies against the S100 antigen and the
harvesting, because this procedure may miss the lymph
MART-1 antigen and/or against tyrosinase.
node’s subcapsular region (the area most commonly involved • Molecular analysis based on PCR techniques is
in metastases). Furthermore, at variance with breast cancer, extremely sensitive in the detection of melanoma
immunohistochemistry is necessary for accurate pathologic cells (or cell debris containing nuclear material)
analysis of melanomas (see further below). Finally, due to within an excised sentinel lymph node.
the process, some tissue will be discarded, and occult meta-
static cells may be missed. A possible alternative is touch-
preparation cytology.
There is not yet a universally accepted protocol for pro- 16.3.6 Accuracy of Sentinel Lymph Node
cessing SLNs in melanoma patients. In addition to conven- Biopsy and Long-Term Prognosis
tional staining with hematoxylin and eosin (H&E), diagnostic
sensitivity and specificity of pathologic analysis are greatly SLNB identifies about 20% of patients with clinically occult
enhanced by performing immunohistochemistry with anti- nodal metastasis at diagnosis, thus preventing the develop-
bodies against the S100 antigen (one of the most sensitive ment of regional lymph node involvement by performing
markers for melanoma) and against the melanoma antigen selective lymph node dissection at the involved lymphatic
recognized by T cells (MART-1, one of the most specific basin; on the other hand, it avoids futile surgery in patients
markers for melanoma) and/or against tyrosinase [35]. with pathologically negative SLNs, i.e., those patients who
Finally, the extremely high sensitivity of molecular analysis would not benefit from lymphadenectomy.
using PCR-based techniques greatly facilitates the detection Currently, regional lymphadenectomy is performed only in
of melanoma cells (or cell debris containing nuclear mate- patients with metastatic SLN, although recent studies based on
rial) within an excised SLN [36, 37]. long-term follow-up raise some concern that the false-negative
rate is actually higher than it was reported by initial validation
studies based on regional lymphadenectomy—which was per-
Key Learning Points formed irrespective of the SLNB findings. On the other hand,
• Injection of a blue dye may be employed to comple- there is wide variability in the values reported for the false-
ment with intraoperative visual guidance the negative rate, ranging from 5.6% to a surprisingly high value
radioguidance provided by the radionuclide of 21%; these data suggest that in some patients micrometas-
procedure. tases were not correctly identified during the SLN procedure
• According to current recommendations, all stained and then progressed over time to palpable disease [28, 38].
and/or radioactive lymph nodes retrieved during The Multicenter Selective Lymphadenectomy Trial
surgery should be harvested for analysis. (MSLT-I) demonstrated that, in selected patients with
• Although the best count-rate threshold identifying intermediate-thickness cutaneous melanoma, clinical man-
the true sentinel lymph node(s) has not yet been agement based on SLNB findings prolongs disease-free sur-
univocally defined, different parameters have been vival and melanoma-specific overall survival, whereas SLNB
proposed, mostly based on the ratio of count rate of has not been demonstrated to achieve similar clinical benefits
additional radioactive lymph nodes relative to the in patients with higher Breslow thickness [39].
hottest lymph node retrieved (10% rule or 20% Particular attention must be paid to melanomas with lym-
rule). phatic drainage to multiple basins, as frequently observed in
the torso and head and neck region. In fact, patients with trun-
cal melanomas and multiple lymphatic basin drainage have

less favorable survival than patients with a single draining • There might be a non-negligible rate of false-
lymphatic basin. Truncal melanomas are also associated with negative sentinel lymph node biopsies in patients
an increased risk of nodal metastasis. Furthermore, SLNs in with cutaneous melanoma.
uncommon sites must be harvested for analysis, since they may • Clinical management based on sentinel lymph node
contain metastasis at nearly the same frequency (about 20%) as findings prolongs disease-free survival and
SLNs in common lymphatic drainage basins. melanoma-specific overall survival in patients with
When SLNs in both common and uncommon drainage intermediate-thickness cutaneous melanoma,
basins cannot be harvested because of surgical difficulties, whereas evidence is not stringent in patients with
close clinical and ultrasound-based monitoring of the l ymphatic higher Breslow thickness.
stations at risk is recommended during follow-up [40–42]. • Sentinel lymph node biopsy is indicated in mela-
noma patients with clinically lymph node-negative
lymphatic stations and intermediate Breslow thick-
16.3.7 Indications and Contraindications ness (1.01–4 mm).
• Controversy remains when the primary melanoma
The SLNB procedure must be performed after histological is <1 mm or >4 mm; different factors should be
confirmation that the primary lesion is indeed a malignant considered in these patients, based on histology of
melanoma. In the common clinical practice, it is generally the primary tumor.
combined with wide local excision of the scar from prior
biopsy of the suspected cutaneous lesion, with 1–2 cm mar-
gins depending on the melanoma Breslow thickness.
According to current guidelines, SLNB is indicated for 16.4 R
adioguided Sentinel Lymph Node
staging melanoma patients with clinically lymph node- Biopsy in Head and Neck Cancers
negative lymphatic stations and intermediate Breslow
thickness (1.01–4 mm). 16.4.1 The Clinical Problem
When thickness of the primary melanoma is <1 mm or
> 4 mm, some controversy remains, and each individual case New cases of oral cavity, pharynx, and larynx cancers in the
should be considered taking into account different factors. In United States in 2017 were estimated at about 63,000 (73%
particular, SLNB should be offered to selected patients with men, 27% women), 27% of them being located in the phar-
thin melanoma and high-risk features (ulceration, mitotic ynx, 26% in the tongue, 21% in the mouth, 21% in the lar-
rate ≥1 mm2, or >50% regression), especially in the sub- ynx, and 5% at other sites. These tumors account for about
group of patients with Breslow thickness between 0.75 and 3.5% of all new cancers cases, and about 13,350 deaths were
0.99 mm. It may also be considered in patients where exact expected for these cancers in 2017 [44]. Squamous cell car-
thickness cannot be reliably assessed because of inadvertent cinoma (or its variants) is the most frequent histologic type
shaving or cauterization before definitive biopsy of the sus- of these tumors (over 90% of the cases).
pected lesion. SLNB can be considered also for melanomas Metastasis to neck lymph nodes is a major determinant
with >4 mm Breslow thickness, in order to obtain more accu- for the prognosis of oral, oropharyngeal, and other head and
rate information and for local control. neck cancers, as the disease-free survival rate decreases to
Contraindications for SLNB include poor general health approximately 50% when even a single lymph node harbors
status, local or systemic spread of disease ascertained by other metastasis. The extent of lymph node involvement can be
diagnostic procedures, and prior extensive surgery in the region considered as an indirect index of the systemic tumor burden
of the primary tumor or of the lymph node basin. In case of and is a crucial factor of tumor staging, being closely corre-
concurrent primary melanoma and satellitosis or “in-transit” lated to overall survival and constituting a main determinant
metastases, SLNB should not be considered due to the fact that of treatment planning [45].
these patients are staged as having stage III disease [28, 43]. Preoperative evaluation of patients with newly diagnosed
head and neck cancers includes physical examination,
ultrasound, CT, MRI, and/or PET/CT; all these modalities
Key Learning Points have suboptimal sensitivity for the detection of microscopic
• Sentinel lymph node biopsy identifies about 20% of lymph node involvement, as occult metastasis is found in
patients with clinically occult nodal metastasis at 15–30% of patients with clinically node-negative head and
diagnosis and avoids futile surgery in patients with neck squamous cell carcinoma [46].
negative sentinel lymph node(s). Although debate is still ongoing about performing pro-
phylactic lymph node neck adenectomy versus a wait-and-
see approach [47], selective neck dissection is generally

recommended in all patients with clinically negative lymph • Preoperative evaluation with even the most
node status. However, lymph node metastasis is found in advanced imaging techniques has suboptimal sensi-
only about 30% of the patients with early oral squamous cell tivity for the detection of microscopic lymph node
carcinoma so treated, thus resulting in overtreatment in over involvement.
70% of patients. Prophylactic neck lymphadenectomy has • Although selective neck dissection is recommended
been developed based on the common pathways for spread in patients with clinically negative lymph node sta-
of all head and neck cancers to regional nodes, and it consists tus, lymph node metastasis is found in only about
of surgical removal of those nodes most commonly involved 30% of the patients with early oral squamous cell
with metastasis specifically from head and neck cancers. The carcinoma, thus resulting in overtreatment in over
extent of lymphadenectomy varies according to location of 70% of such patients.
the primary tumor. For example, in case of oral cancers, it • Radioguided sentinel lymph node biopsy consti-
includes all lymph nodes of levels I, II, and III and some- tutes a valid alternative to selective neck dissection,
times also the superior part of level V. In case of pharyngeal as it personalizes identification of lymphatic neck
and laryngeal cancers, selective neck dissection includes lev- drainage and it allows detection of occult cervical
els II, III, IV, and VI when appropriate [48]. metastasis in patients with early head and neck
Due to the anatomical complexity and unpredictable indi- cancers.
vidual variability of lymphatic drainage in the individual • The procedure has already been sufficiently stan-
patients (that can be found in up to 20–30% of patients), the dardized in patients with oral and oropharyngeal
therapeutic value of selective neck dissection is limited by squamous cell carcinoma with accessible subsites,
the occurrence of 10–20% of “skip metastases” that have whereas it is still under experimental in head and
bypassed the expected first nodal basin. neck cancers arising at other sites.
In this scenario, radioguided SLNB constitutes an alterna-
tive to selective neck dissection, as it guides surgery to a per-
sonalized identification of lymphatic neck drainage and it
allows detection of occult cervical metastasis in patients with 16.4.2 Lymphatic System of the Head
early head and neck cancers. In particular, a negative SLNB and Neck
prevents the unnecessary removal of functional lymph nodes
and limits the extent of neck dissection surgery. Compared to The head and neck region contains over 300 lymph nodes,
selective neck dissection, patients also usually experience constituting approximately 30% of lymph nodes in the whole
fewer complications such as sensory disturbances (skin human body. Their anatomy is quite complex due to the close
numbness); overall duration of surgery is also considerably proximity with different tissues and vital organs—often of
shortened [49]. small size.
Application of SLNB technique has been investigated Concerning in particular lymphatic anatomy of the cervical
mostly for those head and neck cancers whose anatomical region, a useful schematic classification into specific anatomic
location allows direct and easy access to the tumor for injec- subsites groups them into seven levels on each side of the neck,
tion of the lymphatic mapping agents. For this reason the Robbins’ classification [52] (Fig. 16.17). In the anterosuperior
procedure has already been sufficiently standardized in region of the neck, submental and submandibular lymph nodes
patients with oral and oropharyngeal squamous cell carci- are included in level I, respectively, as sublevel IA and sublevel
noma with accessible subsites, while it is still experimental IB, divided by the anterior belly of the digastric muscle. In the
in head and neck cancers arising at other sites. lateral region of the neck, upper jugular nodes extend from the
In patients with early oral and oropharyngeal cancer, base of the skull up to the inferior border of the hyoid bone and
SLNB has shown results comparable to those of selective are classified as level II; the vertical plane defined by the poste-
neck dissection in terms of regional control, as demonstrated rior surface of the submandibular gland divides lymph nodes
by multiple validation trials [50, 51]. located anteriorly (sublevel IIA) from nodes located posteriorly
(sublevel IIB). The middle jugular lymph nodes are included in
level III, the space from the hyoid bone up to the inferior border
Key Learning Points of the cricoid cartilage. Level IV includes lower jugular nodes,
• Metastasis to neck lymph nodes is a major determi- from the inferior border of the cricoid cartilage to the clavicle.
nant for the prognosis and for treatment planning in The posterior border of the sternocleidomastoid muscle consti-
patients with oral, oropharyngeal, and other head tutes the posterior border of levels II, III, and IV. Those lymph
and neck cancers. nodes located posteriorly to the posterior border of the sterno-
cleidomastoid muscle are classified as level V; in turn, sublevel
Fig. 16.17 Lymph node levels

of the neck according to
Robbins’ classification
VA includes the spinal accessory nodes, whereas sublevel VB

includes the nodes following the transverse cervical vessels • The patterns of lymph drainage from the oral cavity,
and the supraclavicular nodes. Pretracheal, paratracheal, pre- oropharynx, larynx, and hypopharynx exhibit
cricoid, and perithyroidal nodes (including the lymph nodes numerous inter- and intraindividual variations, even
along the recurrent laryngeal nerves) constitute level VI. Finally, from the same primary tumor site.
level VII includes the superior mediastinal lymph nodes [53]. • Lymph originating in midline structures (tongue and
The patterns of lymph drainage from the oral cavity, floor of the mouth) frequently drains to contralateral
oropharynx, larynx, and hypopharynx exhibit numerous inter- lymph nodes.
and intraindividual variations, even from the same primary
tumor site. For example, drainage from the anterior floor of the
mouth and lingual apex is expected to submandibular nodes; 16.4.3 Modalities of Tracer Injection
however, bilateral drainage to higher-level lymph nodes and to
the middle jugular chains is not infrequently observed. A variety of radiopharmaceuticals have been used for imag-
Moreover, lymphatic drainage from a primary tumor located in ing the pattern of lymph flow in the head and neck region. In
the midline may be directed to either the left or the right side. On addition to the established radiocolloid lymphatic mapping
the other hand, lateralized malignancies of the tongue or floor of agents (primarily 99mTc-albumin nanocolloid and 99mTc-sulfur
the mouth often show pure contralateral drainage [54–56]. colloid), a new tracer targeting the mannose receptors located
on macrophages and other cells in lymph nodes, 99mTc-til-
manocept, has more recently been approved; the particularly
small particles of this agent result in rapid clearance from the
Key Learning Points site of interstitial injection, while avid binding to the CD206
• Anatomy of the lymphatic system in the head and receptors expressed on macrophages’ membranes results in
neck region (which constitutes about 30% of all reduced or absent drainage to second-echelon lymph nodes
lymph nodes in the whole human body) is quite and efficient SLN retention, thus facilitating SLN detection
complex. both preoperatively and intraoperatively [57, 58].
• Robbins’ classification levels for each side of the As an alternative to the radioactive label, near-infrared fluo-
neck constitute a well-validated basis for staging rescent tracers (such as indocyanine green—ICG) have been
patients with cancers of the head and neck. used that are potentially helpful for SLN biopsy [59]; moreover,
the combination of fluorescent tracers with conventional 99mTc-
radiocolloids is an attractive option for improving the SLN 16.4.4 Lymphoscintigraphy

detection rate in patients with head and neck malignancies, by
combining radioguidance with optical guidance [60]. Preoperative lymphoscintigraphy provides images of
The technique of tracer injection may have a relevant lymphatic drainage in the tumor region. SLNs are visual-
impact on the acquisition time and image quality during pre- ized as “hot spots” along a certain pathway of lymphatic
operative lymphatic mapping. In case of oropharyngeal can- drainage. Accurate gamma camera imaging plays an
cers, the tracer should be injected intramucosally in the healthy important role in lymphatic mapping, since the resulting
mucosa surrounding the malignant lesion or scar margin, con- images are used to direct the surgeon to the site(s) of
sidering that in the subepithelial stroma there is a high concen- SLN(s).
tration of lymphatic capillaries, which provides a larger area With the patient positioned as comfortably as possible on
for faster lymph drainage. Intratumoral or deep injections the imaging table (usually in the supine position—to repro-
should be avoided, because bleeding at the injection site results duce positioning during surgery), acquisitions are generally
in lower image quality and difficult SLN identification. performed using a large FOV gamma camera to visualize all
Tracer injection is usually performed with small syringes the possible routes of lymphatic drainage. Dynamic acquisi-
with minimal dead space; alternatively, 0.1 mL of air may be tion for 20–30 min starting immediately after radiotracer
drawn into the syringe behind the radiocolloid suspension to injection shows the drainage pattern and helps to distinguish
ensure complete administration. A 25 G or 27 G needle should SLNs from second-echelon nodes. SLNs are generally iden-
be used, total injected activities varying from 15 to 120 MBq, tified 15–60 min after radiotracer injection as one or more
depending on size and location of the primary tumor. Furthermore, hot spots to which lymphatic drainage passes and may be
the injected activity should be adjusted according to the timing of multiple, ipsilateral, and/or contralateral to the primary
lymphoscintigraphy with respect to surgery; in particular, greater tumor, in one or more areas of the neck. Static images in the
activities are required if surgery is scheduled the day after lym- anterior and lateral views are subsequently acquired; the
phoscintigraphy (2-day protocol), in order to ensure that the skin projection of SLNs can be marked with the aid of an
remaining activity exceeds 10 MBq at the time of surgery. external radioactive marker, such as a 57Co-source pen. If
Small volumes (0.1–0.2 mL per aliquot) should be injected SLNs are not clearly visualized, static imaging can be
at a short distance from the primary lesion, in order to avoid repeated later 2–4 h postinjection or even just before
masking of SLN(s) possibly located in the vicinity of the injec- surgery.
tion site; small volume is recommended also to minimize spill- Repeat radiotracer injection and imaging may be consid-
ing of the radiotracer (due to resistance to injection of, e.g., the ered in case of totally absent visualization of lymph nodes;
tongue tissue) causing contamination of the field. The number however, proceeding to neck dissection is preferred in these
of aliquots to be injected varies from two to four, depending on cases in order to avoid erroneous staging information caused
size and location of the lesion. The use of a local anesthetic for by a false-negative SLN biopsy. It must be considered that
topical application (10% xylocaine spray) a few minutes before accurate preoperative localization and marking of the skin
tracer injection is recommended for oral cavity tumors. projection of SLN location correlate well with the precision
The most frequent pitfall possibly occurring during lym- of the surgical procedure.
phoscintigraphy is skin or mucosal radioactive contamina- Although SLNs can often be adequately localized with
tion due to spillage of the lymphatic mapping agent either planar imaging alone, the use of SPECT/CT imaging facili-
during injection or immediately thereafter. Oral contamina- tates in the majority of cases the procedure of SLNB, particu-
tion can be reduced by inviting the patient to use a mouthwash larly when considering the complex anatomical region of the
or to rinse the mouth before swallowing. head and neck and the proximity of some primary lesions to
SLNs and to other important anatomic structures. SPECT/CT
is useful to identify additional SLNs (otherwise missed on
Key Learning Points planar imaging, especially when radioactive lymph nodes are
• The lymphatic mapping radiopharmaceutical adjacent to the injection site) (Fig. 16.18), as well as to
(either 99mTc-colloid or 99mTc-tilmanocept) is gener- exclude ambiguous imaging in case of radiotracer leakage or
ally injected intramucosally in multiple aliquots contamination. Furthermore, SPECT/CT imaging allows
around the tumor site. planning of the best surgical approach, thanks to the anatomi-
• The use of hybrid radioactive and fluorescent lym- cal details provided by CT in hybrid images and thanks to
phatic mapping agents (potentially combining the volume rendering with 3D images; for instance, it can clarify
advantages of radioguidance with those of visual if SLNs are located superficially underneath the skin or hid-
guidance) is currently under clinical investigation. den below in deep tissues or in proximity or not to vital vas-
• For cancers of the mouth, cautions must be adopted cular and neural structures [61, 62]. Nevertheless, it must be
to minimize spilling of the radiotracer causing con- emphasized that SPECT/CT imaging does not replace planar
tamination of the field. lymphoscintigraphy—but rather it complements planar
imaging.
a b
c d e f
Fig. 16.18 Lymphoscintigraphy obtained after perilesional, submuco- slices (c and e) and corresponding CT slices (d and f) better localize
sal injection of 99mTc-nanocolloidal albumin in a patient with carcinoma SLNs in the submandibular area (yellow dashed circle in d) and in the
of the tongue. (a) Planar imaging shows bilateral drainage of the lym- lower neck (yellow dashed circles in f) (reproduced with permission
phatic mapping agent; however, due to proximity with the site of radio- from: Orsini F, Puta E, Valdés Olmos RA, Vidal-Sicart S, Giammarile F,
colloid injection (large area of intense activity), it is difficult to ascertain Mariani G. Radioguided surgery for head and neck. In: Strauss HW,
the exact number and anatomical location of the radioactive lymph Mariani G, Volterrani D, Larson SM, eds. Nuclear oncology—From
nodes. (b) 3D surface volume rendering depicts an additional SLN, left pathophysiology to clinical applications. New York, NY: Springer;
from the injection site (yellow arrow). Transaxial fused SPECT/CT 2017:1433–49)
16.4.5 Intraoperative Gamma Probe

Key Learning Points Counting/Detection
• In patients with head and neck cancers, lymphos-
cintigraphy should include an early dynamic acqui- A handheld γ-detecting probe is routinely used for intraop-
sition, which helps to distinguish true sentinel erative detection of the SLN(s) in the surgical field. The
lymph nodes from second-echelon nodes visualized injection of blue dye at the time of surgery in addition to
along the same pathway of lymphatic drainage. radioguidance with the intraoperative gamma probe is
• Static images in the anterior and lateral views are optional and may be a useful adjunct to aid SLN localization
subsequently acquired, images being recorded even and harvesting. After SLN removal, the resection site and
up to 2–4 h after injection if necessary when senti- other cervical regions are explored with the gamma probe for
nel lymph nodes are not clearly visualized. any significant residual radioactivity; in this case further
• SPECT/CT imaging facilitates in the majority of radioactive lymph nodes may be retrieved and harvested. An
cases preoperative lymphatic mapping, because ex vivo counting ratio of 10:1/20:1 with respect to back-
of the complex anatomical region of the head and ground identifies a hot spot as a radioactive SLN to be ana-
neck and the proximity of some primary lesions lyzed for the presence of metastatic tumor cells.
to SLNs and to other important anatomic Sometimes the SLNs can be difficult to identify due to
structures. close proximity with the peritumoral injection site (the
so-called shine-through effect), which can limit detection 16.4.6 Retrieval and Analysis of
of SLNs, especially in patients with tumors of the mouth Sentinel Lymph Nodes
floor. Also after removing radioactive SLNs, activity
remaining at the injection site can impair measurement of Histopathologic analysis of SLNs is extremely effective and
residual activity in the excision surgical field. For these can detect the presence of micrometastasis (<2 mm) and of
reasons, the primary cancer should be resected before clusters of several cells or even isolated tumor cells. By
searching for the SLN(s). focusing on only a few lymph nodes, the pathologist can
Deeply located SLNs can be difficult to detect because of completely dissect and examine at 50–100 µm intervals each
tissue attenuation; in these cases SPECT/CT imaging con- SLN. Sections may be stained with hematoxylin and eosin
siderably helps in evaluating depth of radioactive lymph (H&E) for conventional light microscopy; if such examina-
nodes. A number of portable and handheld mini gamma tion is negative, adjacent sections may be used for
cameras have been developed to provide direct intraopera- immunohistochemistry.
tive visualization of radioactive foci, with the purpose of Complete neck dissection is performed if the SLNB pro-
improving detection of SLNs. Using these devices, the cedure detects invasion of at least one sample (10–15% of
entire lymph node excision procedure in the head and neck cases). If bilateral drainage is seen on lymphoscintigraphy,
area can be directly monitored in the surgical room. both hemi-necks are usually operated on. Most lymph node
Moreover, by acquiring images of absent or residual radio- procedures are performed in a second step following exhaus-
activity after excision of the SLNs, the dedicated mini tive SLNs pathologic analysis indicating micro- or macro-
gamma camera can help to assess completeness of the pro- metastasis. Other techniques, such as frozen sections, imprint
cedure [63, 64]. cytology, and molecular analysis with RT-PCR for cytokera-
Recently the use of intraoperative gamma cameras has tin, have been described. The clinical role of these methods
been combined with fluorescence cameras for synchronous remains uncertain due to the fact that, at present, the prog-
SLN signal detection using the abovementioned hybrid nostic significance of micrometastasis and isolated tumor
lymphatic mapping radiotracers combined with indocya- cells is unknown [66].
nine green [65]. A further technology called freehand
SPECT (fhSPECT) has been introduced for navigational
surgery, combining the acoustic information of a conven- 16.4.7 Accuracy of Sentinel Lymph Node
tional gamma probe and intraoperative 3D images with Biopsy and Long-Term Prognosis
real-time visualization of radiotracer distribution within the
surgical field. After the SLN concept in oral cancer was validated in several
studies in which all patients underwent elective neck dissec-
tion following SLNB, several long-term prognostic studies
have been performed. In 91–99% of the cases, the SLN was
Key Learning Points successfully harvested; a sensitivity of 86–95% and a nega-
• The injection of blue dye at the time of surgery in tive predictive value of 88–100% have been reported; 9–37%
addition to radioguidance with the intraoperative of patients were upstaged as a result of the SLNB findings
gamma probe may be a useful adjunct to aid senti- [45, 67–73].
nel lymph node localization and harvesting. These data confirm that the technique is a reliable means
• Any lymph node with an ex vivo counting ratio of of personalizing treatment by detecting lymph node metasta-
10:1/20:1 versus background should be harvested ses, provided that the team is experienced and adheres to
for analysis. good practice rules. It must be considered that, compared to
• Intraoperative high-resolution imaging with porta- the conventional surgical approach, the SLNB technique is
ble, small field-of-view gamma cameras does not less invasive, more cost-effective, and beneficial to the
guide per se search for the sentinel lymph node(s), patient’s quality of life, by reducing morbidity and improv-
but rather it is employed for assessing completeness ing cosmetic outcome.
of radioguided sentinel lymph node excision. Failure to visualize and harvest SLNs may be related to
• The so-called “freehand SPECT” technology allows their close proximity to the injection site (e.g., floor of mouth
navigational surgery, combining the acoustic infor- tumors). In fact, lower sensitivity rates are reported for floor
mation of a conventional gamma probe and intraop- of mouth cancers as compared to other oral subsites: 80–86%
erative 3D images with real-time visualization of versus 94–97%, respectively. In these patients, the close spa-
radiotracer distribution within the surgical field. tial relation between the injection site and the SLN seems to
result in a lower identification rate due to the “shine-through
effect.”
Another cause of false-negative SLNB is massive metas-

tasis in lymph nodes that may block lymphatic drainage Key Learning Points
along that particular pathway and reroute it to other lym- • Complete neck dissection is performed in case of a
phatic pathways; this event causes non-visualization of the tumor-positive sentinel lymph node biopsy, as
true SLNs (containing metastasis) and possible visualization observed in about 10–15% of cases.
of alternative routes of lymphatic drainage to metastasis-free • If bilateral drainage is seen on lymphoscintigraphy,
lymph nodes. both hemi-necks are usually operated on, even if the
As to histopathologic analysis of SLNs, the significance sentinel lymph node on one side does not harbor
of isolated tumor cells is still controversial. When it was con- tumor cells.
sidered as an indication for neck dissection, a significant dif- • Sentinel lymph node biopsy is a reliable means of
ference was found at 3 years between cases with isolated personalizing treatment in patients with head and
tumor cells alone versus micro- or macrometastasis, suggest- neck cancers—particularly squamous cell carci-
ing better prognosis in the former case [68]. Other studies noma of the oral cavity.
suggested that in the classification of SLNs as harboring iso- • Possible causes of failure or of false negativity of
lated tumor cells, micro- and macrometastasis could be sentinel lymph node biopsy include close proximity
important to evaluate if treatment of the neck is always nec- to the injection site and massive metastasis in lymph
essary after a positive SLNB. Additional non-SLN metasta- nodes that may block lymphatic drainage along a
ses were found in 31% of neck dissections following a pathway of lymphatic drainage and reroute it to
positive SLNB, respectively, in 13%, 20%, and 40% of other lymphatic pathways.
patients with isolated tumor cells, micrometastasis, and mac- • Sentinel lymph node biopsy is currently indicated
rometastasis in the SLN [74]. in patients with early oral and selected oropharyn-
geal squamous cell carcinomas and clinically nega-
tive lymph node status, staged as clinical T1 or T2.
16.4.8 Indications and Contraindications
Current indications for SLNB include early oral and selected

oropharyngeal squamous cell carcinomas, with negative
lymph node status by palpation and imaging evaluation (CT, 16.5 R
adioguided Sentinel Lymph Node
MR imaging, or PET/CT), staged as clinical T1 or T2 and N0 Biopsy in Gynecological Cancers
cancer [75].
T1 and T2 patients have resectable tumors with size 16.5.1 The Clinical Problem
<4 cm in maximum diameter. In fact, in case of larger
tumors, it is difficult to perform tracer injection all around Vulvar cancer has a low incidence in the most developed
the tumor; furthermore, large tumors tend to drain to mul- countries; for instance, about 6000 new cases overall were
tiple lymphatic basins and in the majority of patients estimated in the United States for the year 2016. The pres-
require a neck dissection for access to the primary tumor or ence of lymph node metastasis is the most important prog-
for defect reconstruction. The indication, generally nostic factor, because the 5-year survival rate drops from
achieved in a multidisciplinary oncology team meeting, 95% to 60% when lymph nodes harbor metastases. On the
may be extended to patients whose necks underwent prior other hand, the surgical complications of bilateral inguinal
surgery or radiation therapy. However, these applications of lymphadenectomy lead to high morbidity in these patients,
the SLNB technique, while clinically attractive, remain who are generally >70 years old and frequently present with
largely unexplored; at the moment, levels of evidence in other clinical comorbidities.
this regard are low, since the prior interventions can distort There were nearly 13,000 new cases of cervical cancer in
the normal lymphatic pathways and give rise to unexpected the United States in 2016, with more than 4000 deaths.
patterns of metastasis [76, 77]. Prognosis of early-stage cancer depends on the lymph node
In pregnant women and children, the urgency and the status, tumor size, depth of stromal invasion, and presence of
necessity of staging the neck should be discussed. In particu- lymphovascular involvement. Pelvic lymphatic metastases
lar, SLNB protocols should be modified to minimize risks of are found in 11–21% of patients with FIGO stage IB and in
radiation exposure and of blue-dye administration. SLNB 40% of patients with FIGO stage IIB, with possible spread to
can safely be performed in lactating women, but it is advised the para-aortic region. Considering these figures, SLNB has
that breastfeeding be discontinued for about 2 days follow- been proposed as an alternative to de novo pelvic lymphad-
ing the procedure [75]. enectomy in an attempt to reduce surgical morbidity.
Endometrial cancer is the most frequent gynecological

cancer, with 60,000 new cases estimated in the United States Key Learning Points
in 2016 and nearly 10,500 deaths. Lymph node metastasis is • Gynecological malignancies where sentinel lymph
present in 10% of stage I patients, and the 5-year survival node biopsy can be employed include vulvar can-
rate drops to 50% when pelvic or para-aortic node lymph cer, cervical cancer, and endometrial cancer.
nodes contain metastases. The treatment of choice for endo- • Lymph originating in the vulva normally drains to
metrial cancer is modified radical abdominal hysterectomy the inguino-femoral lymph nodes, proceeding from
and bilateral salpingo-oophorectomy. However, the role of superficial to deep inguinal nodes and to pelvic
lymphadenectomy remains controversial. Radical pelvic and lymph nodes.
para-aortic lymphadenectomies represent the standard treat- • The expected lymphatic drainage from the cervix is
ment in the high-risk group (grade 3, >50% myometrial inva- bilaterally to the parametria, to the external iliac nodes,
sion or papillary serous, carcinosarcoma, and clear-cell and to the obturator lymph node, progressing thereon
cancer histology). Currently, there is no international con- to the common iliac and para-aortic lymph nodes.
sensus on the optimal extent of surgery for endometrial can- • Lymph from the two lower thirds of the corpus uteri
cer staging. Thus, the SLN procedure is being explored as drains in a similar manner as lymph from the cer-
possible alternative option to de novo pelvic and para-aortic vix, whereas lymph from the upper third drains
lymphadenectomy [78, 79]. directly to the para-aortic lymph nodes.
In general, radioguided SLNB in patients with gyneco-
logical cancers involves a more complex approach than in
patients with breast cancer or melanoma, due to a variety of
anatomical and pathophysiological factors. 16.5.3 Modalities of Tracer Injection
Administration of the lymphatic mapping agent (e.g., a

16.5.2 Lymphatic System of Intrapelvic 99m
Tc-colloid) in patients with vulvar malignancies is per-
Female Organs formed through three to four intradermal/intramucosal injec-
tions (0.1–0.2 mL each aliquot) around the primary lesion or
Lymphatic drainage from the vulva normally flows to the the excision scar (Fig. 16.19).
inguino-femoral lymph nodes, either on one side only or In patients with cervical cancer, lymphatic mapping is
bilaterally. In particular, lymph originating from the labia performed by peritumoral or periorificial injection of
flows to the superficial inguinal nodes, proceeding then to four radiopharmaceutical aliquots. In the case of previ-
the deep inguinal nodes and to the pelvic lymph nodes, ous conization, pericicatricial injection at the four quad-
whereas some midline structures, such as the clitoris, can rants is recommended. Usually, a total volume of 2 mL is
drain directly to deep lymph nodes. Tumors located in the applied.
midline usually drain to both inguinal sides. One of the most controversial issues for accurate lym-
The cervix is a deep midline structure whose lymph drains phatic mapping in patients with endometrial cancer con-
laterally to the parametria, to the external iliac lymph nodes, cerns the best injection site/modality. Different approaches
and to the obturator lymph nodes. The expected drainage is have been proposed to this purpose, such as cervical injec-
bilateral migration to the external iliac lymph nodes. tion, endometrial peritumoral injection assisted by hyster-
Obturator fossa lymph nodes are the most common group, oscopy, and myometrial/subserosal injection. The cervical
followed by progression of lymph to the common iliac and approach (similar to the technique adopted for cervical can-
para-aortic lymph nodes as the second-echelon group. Direct cer, although with somewhat deeper punctions) is the easi-
para-aortic drainage has been observed in 1% of cases. est way to inject the lymphatic mapping agent. It is
Parametrial lymph nodes can be very difficult to evaluate performed periorificially into the four quadrants.
during lymphoscintigraphy and radioguided SLNB, due to Endometrial injection during hysteroscopy allows direct
their proximity to the site of injection of the lymphatic map- injection around the tumor. If it is performed the day before
ping agent. surgery, this procedure requires particular coordination
Lymphatic drainage from the corpus uteri is in some among the gynecology, anesthesia, and nuclear medicine
respect different from cervical drainage. In particular, lymph departments. If it is performed shortly prior to the start of
from the two lower thirds of the corpus drains in a similar surgery, it is problematic to acquire lymphoscintigraphy
manner as lymph from the cervix. Whereas, lymph from the and SPECT/CT. The third option, radiotracer injection into
upper third of the corpus uteri drains directly to the para- the corpus uteri in a myometrial or subserosal location, is
aortic lymph nodes [79]. usually adopted during surgery. However, it is also possible
Fig. 16.19 Modality of injection of the lymphatic mapping agent in in the majority of these patients, it is important to inject the radiocolloid
vulvar cancer. Submucosal injection in a patient with a midline lesion around the tumor, although this can be difficult in the medial part (right
in the left labia minora. Since bilateral lymphatic drainage is expected panel)
to inject the tracer guided by transvaginal ultrasonography then obtained (Fig. 16.20). The lymph node that first receives
the day prior to surgery [78]. a direct lymphatic channel from the tumor or shows an
increase in uptake in the delayed images is usually consid-
ered as a SLN.
Key Learning Points In patients with cervical cancer (Fig. 16.21) or with
• In patients with vulvar cancer, the lymphatic map- endometrial cancer (Fig. 16.22), planar imaging is based on
ping agent is administered through 3–4 intradermal/ 3–5 min anterior and lateral views acquired at 30 min (early)
intramucosal injections around the primary lesion and 60–120 min (delayed) after radiotracer injection. Early
or the excision scar. images can depict the lymphatic duct(s) and the first-
• In patients with cervical cancer, the lymphatic map- draining lymph node(s). Delayed images may discriminate
ping agent is injected peritumorally or periorifi- the SLN from second-echelon nodes. Planar images cannot,
cially, at the four quadrants. however, provide precise anatomical landmarks [78]. This
• The optimal modality of injection in patients limitation is overcome by SPECT/CT imaging obtained just
with endometrial cancer is still debated, and dif- after delayed imaging. In vulvar cancer, lymphatic drainage
ferent approaches are being explored (e.g., cervi- is mainly directed to Daseler’s medial inguinal region
cal injection, endometrial injection during (83%), while drainage to the lateral inferior groin is mini-
hysteroscopy). mal (0.5%) [79].
SPECT/CT is crucial to adopt the optimal approach to
inguinal lymphadenectomy in patients with positive
SLNB. SPECT/CT imaging is mandatory in case of cervical
16.5.4 Lymphoscintigraphy and endometrial malignancies, since it allows correction for
tissue attenuation and may thus lead to detection of addi-
In the case of vulvar cancer, a dynamic acquisition is recorded tional SLN(s). Moreover, by providing accurate anatomical
immediately after tracer injection, usually for about 10 min. localization, it plays an important role in planning surgery
Early (15 min) and delayed (2 h) static planar images are and shortening the operative time [80].
Fig. 16.20 Lymphoscintigraphy obtained after perilesional, submuco- However, a faint uptake on the right inguinal side is observed as well
sal injection of 99mTc-nanocolloidal albumin in the same patient with (red arrow). The 3D surface volume rendering image obtained after
vulvar cancer as depicted in Fig. 16.19. The delayed planar image (left SPECT/CT acquisition more accurately visualizes the radioactive
panel) shows a predominantly left inguinal lymphatic drainage. lymph nodes (right panel)
Fig. 16.21 Lymphoscintigraphy obtained after perilesional injection of slice (lower right panel) depict more precisely the location of the radioac-
99m
Tc-nanocolloidal albumin in the patient with cervical cancer. Upper left tive lymph nodes, one of which (considered a true SLN) is located in the
panel: planar imaging shows bilateral lymphatic drainage with two radio- right parametrium (reproduced with permission from: Paredes P, Vidal-
active lymph nodes in the right pelvis and one in the left pelvis. Note a hot Sicart S. Preoperative and intraoperative lymphatic mapping for radiogu-
spot near the injection site, depicted on the right-hand side. Upper right ided sentinel node biopsy in cancers of the female reproductive system. In:
panel: 3D surface volume rendering obtained after SPECT/CT acquisition Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, eds.
shows the anatomical distribution of the radioactive lymph nodes. Coronal Atlas of lymphoscintigraphy and Sentinel Node Mapping—A pictorial
fused SPECT/CT slice (lower left panel) and transaxial fused SPECT/CT case-based approach. Milan: Springer 2013:249–68)
Fig. 16.22 Lymphoscintigraphy obtained after cervical injection of anatomical location of the radioactive lymph nodes near major blood
99m
Tc-nanocolloidal albumin in the patient with endometrial cancer. The vessels (reproduced with permission from: Paredes P, Vidal-Sicart
transaxial fused SPECT/CT slice (upper left panel) shows radiocolloid S. Preoperative and intraoperative lymphatic mapping for radioguided
uptake in SLNs located in both external iliac areas; the lymph nodes are sentinel node biopsy in cancers of the female reproductive system. In:
depicted in the CT component of the SPECT/CT acquisition (green cir- Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, eds.
cles in the lower left pane). Three-dimensional surface volume rendering Atlas of lymphoscintigraphy and Sentinel Node Mapping—A pictorial
obtained after SPECT/CT acquisition provides a good overview of the case-based approach. Milan: Springer 2013:249–68)
cases the same SLNs are identified by the blue dye and by the
Key Learning Points radioactive lymphatic mapping agent, radioguided SLNB has
• In patients with vulvar cancer, an early dynamic in general better performance than optical guidance alone
acquisition helps to identify the lymphatic basins at with blue dyes. Nevertheless, current standard practice in
risk and also to recognize higher-echelon lymph patients with vulvar or cervical cancer includes the simultane-
nodes visualized after visualization of the “true” ous injection of both types of lymphatic mapping agent,
sentinel lymph node along a certain pathway of radiopharmaceutical and blue dye, for combined radioguid-
lymphatic drainage; static planar images acquired ance and optical guidance. The use of fluorophores (mainly
up to about 2 h postinjection are usually sufficient indocyanine green, ICG) has recently been advocated in this
for adequate lymphoscintigraphic mapping in these scenario, with better results than with the blue dye [78, 79].
patients. In patients with vulvar tumors, skin marks on the cutane-
• Lymphoscintigraphy can be more complex in ous projection of the SLN(s) can indicate the best location
patients with cervical cancer or endometrial cancer, for surgical incision and subsequent use of the handheld
in whom SPECT/CT imaging acquired in addition gamma probe for exploring the surgical field. Conversely, in
to early and delayed planar imaging provides the cervical and endometrial tumors, laparoscopy is the most
anatomical correlates that are especially useful to frequent surgical approach; therefore, cutaneous marks are
guide the surgeon to retrieve the sentinel lymph not useful as a guide. Moreover, specifically designed gamma
node(s). probes must be used during laparoscopy. In particular, these
probes must be inserted through a 10–12-mm-diameter tro-
car, must be long enough to reach the area of interest (up to
a distance of 25 cm from the abdominal surface), and must
16.5.5 Intraoperative Gamma Probe/Gamma have suitable maneuverability so to cover an area with a
Camera Detection radius of 15–20 cm. Laparoscopic gamma probes have been
developed taking these issues into consideration, in particu-
SLN detection using blue dyes alone in patients with vulvar lar the limitation on maneuverability that may impede SLN
or cervical cancer ranges from 75% to 95%. Although in most identification. In these probes the angle of the detector rela-
tive to the tip varies from 0° to 45–90°, according to specific

uses in different phases of the same surgical session. by proximity with radioactivity either at the injec-
In clinical practice, possible interferences during laparo- tion site or at physiologic sites of accumulation
scopic gamma probe scanning must be kept in mind. In par- other than lymph nodes.
ticular, the uterus is usually enlarged in patients with • Visual guidance, especially with the use of a fluo-
endometrial cancer, thus increasing the interference on SLN rescent agent, turns out to be particularly useful to
detection. Furthermore, laparoscopic gamma probe scanning overcome these interferences.
along the intra-abdominal lymphatic pathways can be affected • Other imaging modalities possibly employed dur-
by activity accumulated in the ureter or by liver activity due to ing surgery (such as dedicated small field-of-view
radiocolloid uptake in the reticuloendothelial system [78–82]. gamma cameras or freehand SPECT) can be useful
Due to anatomical complexity of the area, activity remain- to assess completeness of sentinel lymph node
ing at the injection site can mask activity coming from the excision.
SLN, as in the case of, e.g., an inguino-femoral SLN adja-
cent to the radiotracer injection site in a patient with vulvar
cancer or parametrial SLNs in cervical/endometrial cancer. 16.5.6 Retrieval and Analysis of Sentinel
In such instance, intraoperative visual guidance (e.g., with Lymph Nodes
blue dye—or with fluorescent agents) is particularly useful.
Furthermore, intraoperative use of gamma probes can be The optimal method for histopathological examination of
supplemented with the use of portable gamma cameras or SLNs in patients with gynecological cancers has not been
other techniques such as intraoperative freehand SPECT. The established. Considering that routine staining with hema-
main advantages of using a portable gamma camera for SLN toxylin and eosin (H&E) may not allow the identification of
detection in gynecological cancers are (1) greater sensitivity micrometastasis (<2 mm in diameter), ultrastaging histo-
for localizing parametrial lymph nodes, (2) better discrimi- logic techniques can be used when H&E staining is nega-
nation of interference from liver activity when resecting tive for metastasis. In particular, immunohistochemical
para-aortic lymph nodes, and (3) better ability to ascertain staining can be employed for enhanced sensitivity in detect-
the completeness of SLN excision. ing tumor cells.
Intraoperative freehand SPECT can provide some advan- There is current consensus on the rationale of performing
tages also in this setting, as it yields virtually real-time infor- lymphadenectomy of the basin where the SLN contains
mation on precise SLN’s depth and on completeness of SLN micrometastasis, although it can be argued that the benefit of
resection. Moreover, SPECT/CT data obtained during lym- increased detection of metastatic involvement in the SLN
phoscintigraphy with a large field-of-view gamma camera outweighs potential harms, including overtreatment of
can be uploaded and included in the image display to provide patients with micrometastasis [83].
anatomical landmarks and the possibility of intraoperative In patients with cervical cancer, hysterectomy is per-
navigation [79, 80]. formed only when no metastases are found in the pelvic
lymph nodes; ultrastaging with immunohistochemistry has
resulted in increased detection of micrometastases in
Key Learning Points 25–30% of patients in whom conventional H&E staining had
• In patients with vulvar or cervical cancer, sentinel revealed no SLN metastasis. In this regard, the recurrence-
lymph node biopsy is often performed combining free survival rate is significantly reduced in patients with
preoperative lymphatic mapping and intraoperative micrometastases, thus suggesting that even this low meta-
gamma probe guidance with visual guidance using static load is an independent prognostic factor in cervical
a blue dye (or more recently a fluorescent agent). cancer.
• The most frequent approach to sentinel lymph node The added value of immunohistochemistry is relevant
biopsy in patients with cervical or endometrial can- also in patients with endometrial cancer, since nearly 20% of
cer is laparoscopic surgery, using specifically patients are upstaged because of the detection of microme-
designed gamma probes that can be inserted in the tastases. In low-risk endometrial cancer (grades 1–2, with
abdominal cavity through a laparoscopic trocar and less than 50% myometrial invasion), ultrastaging increases
reach deep intra-abdominal sites. by about 50% the number of positive SLNs versus standard
• Intraoperative gamma probe guidance can be some- histology. Nevertheless, there is no consensus yet on how to
what hampered by interferences in counting caused treat patients with endometrial cancer and micrometastasis
only in SLN(s) [78].
16.5.7 Accuracy of Sentinel Lymph Node 50% or endometrial serous and clear-cell carcinoma) require
Biopsy and Long-Term Prognosis staging based on pelvic and para-aortic lymph node
dissection. Randomized clinical trials demonstrated that
Lymphatic mapping with SLNB may modify the manage- there is no advantage in performing systemic
ment of patients with gynecologic tumors by allowing accu- lymphadenectomy in patients with early-stage endometrial
rate stratification according to lymph node metastatic status cancer, and the debate whether lymphadenectomy should be
and possibly sparing in many patients the potential morbidity the standard approach continues. Other studies showed that
of unnecessary lymphadenectomy. SLNB predicts lymph node status in patients with early-
The rate of radioguided SLN identification is high in stage endometrial cancer and that lymph node involvement
patients with vulvar and cervical cancer, in whom lymph was more accurately detected by ultrastaging.
node status plays a key role for the selection of further sur- The SENTI-ENDO multicenter study assessed more than
gery or changes in patient’s management. In the current clini- 120 patients with endometrial cancer using a double-tracer
cal practice, patients with early vulvar cancer are treated with cervical injection. SLN was detected in 88.8% of patients
de novo bilateral inguino-femoral lymphadenectomy, which and 17% presented pelvic lymph node metastases. The
is burdened by high comorbidity; however, only 10–26% of negative predictive value was 97%, with three cases of false-
the patients actually have inguinal metastases. In a recent negative results. It was concluded that SLNB could be an
meta-analysis, the overall SLN detection rate per groin was alternative to lymphadenectomy in patients with low- or
86.9% with combined blue dye and radiocolloid. The false- intermediate-risk endometrial cancer. However, the real inci-
negative rate resulting from 25 studies using the combined dence of metastasis in para-aortic lymph nodes could be
dual tracer procedure was 6.6%, with an average 3.4% recur- underestimated [78, 79, 81, 82].
rence rate in the groin after a negative SLNB; nevertheless,
the GROINSS-V study reported that women with multifocal
disease had a higher rate of false-negative SLNB, 11.8% [84]. Key Learning Points
In these patients SLNB resulted in a lower incidence of • Analysis of the harvested sentinel lymph nodes in
morbidity than de novo inguino-femoral lymphadenectomy. patients with gynecological malignancies is based
The wound breakdown rate was 11.7% versus 34%, cellulitis on conventional hematoxylin and eosin staining,
4.5% versus 21.3%, and lymphedema 1.9% versus 25.2%. although immunohistochemistry is more sensitive
Cervical cancer is usually treated with radical hysterec- for the detection of micrometastasis <2 mm or iso-
tomy and pelvic lymphadenectomy. However, the great lated tumor cells.
majority of stage Ib patients does not benefit from lymphad- • There is no univocal consensus on the optimal thera-
enectomy, as only 15% of them have lymphatic metastasis at peutic strategy to be adopted on the basis of the
diagnosis. During surgery, a positive SLNB avoids hysterec- results of sentinel lymph node biopsy in patients with
tomy or trachelectomy, while para-aortic lymphadenectomy gynecological malignancies, especially when micro-
is performed because of the higher risk of further lymph metastasis only is found in sentinel lymph node(s).
node involvement; afterwards, chemoradiation therapy • Nonetheless, sentinel lymph node biopsy has a defi-
becomes the treatment of choice. On the contrary, when nite impact in patients with gynecological
SLNB is negative for metastasis, hysterectomy or trachelec- malignancies.
tomy can be performed as planned. • The false-negative rate of sentinel lymph node
A meta-analysis showed that the highest SLN detection biopsy in patients with vulvar cancer is low when
rates were obtained with the combined radiotracer and blue- using the combined radioguidance and visual blue-
dye procedure. The pooled SLN detection rate was 92.2% dye guidance procedure.
with a 95% negative predictive value for cervical tumors • Sentinel lymph node biopsy yields important infor-
<2 cm. Bilateral SLN visualization is associated with fewer mation in patients with cervical cancer, with high
false-negative results. detection rate and low false-negative rates when
A novel lymphatic mapping modality that employs ICG using the combined radioguidance and visual blue-
dye with near-infrared fluorescence imaging is gaining dye guidance procedure.
widespread use because of the possibility of real-time • Visual guidance with a fluorescent agent is a useful
imaging during surgery and higher overall and bilateral SLN complement to radioguidance for sentinel lymph
detection rates than with the blue dye [78]. node biopsy in patients with cervical cancer.
Early-stage, low-risk endometrial cancer requires pelvic • Sentinel lymph node is feasible and can represent
lymphadenectomy for staging, although this procedure is a an alternative to lymphadenectomy in patients with
matter of controversy. Patients in the high-risk group (grade low- or intermediate-risk endometrial cancer.
3 endometrioid carcinoma with myometrial infiltration over
16.5.8 Indications and Contraindications virtual/augmented reality environment radioguided surgery. Q J

Nucl Med Mol Imaging. 2014;58:207–15.
3. Nieweg OE, Tanis PJ, Kroon BBR. The definition of a sentinel
According to the EANM guidelines published in 2014, the node. Ann Surg Oncol. 2001;9:538–41.
indications for SLNB in patients with gynecological cancers 4. Zurrida S, Veronesi U. Milestones in breast cancer treatment. Breast
can be summarized as follows: J. 2015;21:3–12.
5. Liu Y. Role of FDG PET-CT in evaluation of locoregional nodal
disease for initial staging of breast cancer. World J Clin Oncol.
• Early cervical cancer (Ia2/Ib1, IIa1 stages). 2014;5:982–9.
• Stage I and II high-risk endometrial cancer, i.e., endome- 6. Beek MA, Verheuvel NC, Luiten EJ, Klompenhouwer EG, Rutten
trioid cancer with the following features: >50% myome- HJ, Roumen RM, et al. Two decades of axillary management in
breast cancer. Br J Surg. 2015;102:1658–64.
trial invasion or poorly differentiated (grade 3) or serous 7. Veronesi U, Paganelli G, Galimberti V, Viale G, Zurrida S, Bedoni
papillary, clear-cell, or carcinosarcoma histological M, et al. Sentinel node biopsy to avoid axillary dissection in breast
subtype. cancer patients with clinically negative lymph-nodes. Lancet.
• Squamous cell vulvar carcinoma Ib/II <4 cm in size, with- 1997;349:1864–7.
8. Giuliano AE, Gangi A. Sentinel node biopsy and improved patient
out presurgical lymph node metastases. care. Breast J. 2015;21:27–31.
• Although based on limited experience, SLNB in vulvar 9. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW,
melanoma is also accepted with the same indications as in Blumencranz PW, et al. Axillary dissection vs no axillary dissection
cutaneous melanoma. in women with invasive breast cancer and sentinel node metastasis:
a randomized clinical trial. JAMA. 2011;305:569–75.
10. Li CZ, Zhang P, Li RW, Wu CT, Zhang XP, Zhu HC. Axillary
Contraindications for SLNB in patients with gynecologi- lymph node dissection versus sentinel lymph node biopsy alone for
cal cancers include the following conditions: early breast cancer with sentinel node metastasis: a meta-analysis.
Eur J Surg Oncol. 2015;41:958–66.
11. Giuliano AE, Ballman K, McCall L, Beitsch P, Whitworth PW,
• Suspected extrauterine involvement. Blumencranz P, et al. Locoregional recurrence after sentinel lymph
• Presence of pathological pelvic or para-aortic lymph node dissection with or without axillary dissection in patients
nodes on radiological or other imaging investigation. with sentinel lymph node metastases: long-term follow-up from
• Prior history of surgery or radiotherapy of the lymph node the American College of Surgeons Oncology Group (Alliance)
ACOSOG Z0011 randomized trial. Ann Surg. 2016;264:413–20.
areas under evaluation. 12. Giuliano AE, Ballman KV, McCall L, Beitsch PD, Brennan
• General contraindication to surgical treatment (related to MB, Kelemen PR, et al. Effect of axillary dissection vs no
age or medical conditions). axillary dissection on 10-year overall survival among women
with invasive breast cancer and sentinel node metastasis: the
ACOSOG Z0011 (Alliance) randomized clinical trial. JAMA.
2017;318:918–26.
Key Learning Points 13. Valero MG, Golshan M. Management of the axilla in early breast
• Sentinel lymph node biopsy in patients with gyne- cancer. Cancer Treat Res. 2018;173:39–52.
14. Giammarile F, Alazraki N, Aarsvold JN, Audisio RA, Glass E,
cological malignancies is indicated for early cervi- Grant SF, et al. The EANM and SNMMI practice guideline for
cal cancer, stage I and II high-risk endometrial lymphoscintigraphy and sentinel node localization in breast cancer.
cancer, and clinically lymph node-negative squa- Eur J Nucl Med Mol Imaging. 2013;40:1932–47.
mous cell vulvar cancer <4 cm in size. 15. Lyman GH, Giuliano AE, Somerfield MR, Benson AB 3rd, Bodurka
DC, Burstein HJ, et al. American Society of Clinical Oncology
• Besides established general contraindications to guideline recommendations for sentinel lymph node biopsy in
surgery, sentinel lymph node biopsy is contraindi- early-stage breast cancer. J Clin Oncol. 2005;23:7703–20.
cated when extrauterine involvement is suspected 16. Vera DR, Wallace AM, Hoh CK. A synthetic macromolecule for
and there are preoperative evidence of pathological sentinel node detection: 99mTc-DTPA-mannosyl-dextran. J Nucl
Med. 2001;42:951–9.
pelvic or para-aortic lymph nodes and prior history 17. Albo D, Wayne JD, Hunt KK, Rahlfs TF, Singletary SE, Ames FC,
of surgery or radiotherapy of the lymph node areas et al. Anaphylactic reactions to isosulfan blue dye during sentinel
under evaluation. lymph node biopsy for breast cancer. Am J Surg. 2001;182:393–8.
18. Paganelli G, Luini A, Veronesi U. Radioguided occult lesion local-
ization (ROLL) in breast cancer: maximizing efficacy, minimizing
mutilation. Ann Oncol. 2002;13:1839–40.
19. Jakub JW, Gray RJ, Degnim AC, Boughey JC, Gardner M, Cox
References CE. Current status of radioactive seed for localization of non-
palpable breast lesions. Am J Surg. 2010;199:522–8.
1. Zaknun JJ, Giammarile F, Valdés Olmos R, Vidal-Sicart S, Mariani 20. Monti S, Galimberti V, Trifirò G, De Cicco C, Peradze N, Brenelli
G. Changing paradigms in radioguided surgery and intraopera- F, et al. Occult breast lesion localization plus sentinel node biopsy
tive imaging: the GOSTT concept. Eur J Nucl Med Mol Imaging. (SNOLL): experience with 959 patients at the European Institute of
2012;39:1–3. Oncology. Ann Surg Oncol. 2007;14:2928–31.
2. Valdés Olmos RA, Vidal-Sicart S, Giammarile F, Zaknun JJ, Van 21. Vermeeren L, Valdes Olmos RA, Klop WM, Balm AJ, van den
Leeuwen FW, Mariani G. The GOSTT concept and hybrid mixed/ Brekel MW. A portable gamma-camera for intraoperative detec-
tion of sentinel nodes in the head and neck region. J Nucl Med. tases by intraoperative radioactive threshold and molecular analysis
2010;51:700–3. in patients with melanoma. J Nucl Med. 2008;49:1769–75.
22. Ahmed M, Purushotham AD, Douek M. Novel techniques for senti- 38. Manca G, Romanini A, Rubello D, Mazzarri S, Boni G, Chiacchio
nel lymph node biopsy in breast cancer: a systematic review. Lancet S, et al. A critical reappraisal of false negative sentinel lymph node
Oncol. 2014;15:e351–62. biopsy in melanoma. Q J Nucl Med Mol Imaging. 2014;58:105–13.
23. Nieweg OE, Uren RF, Thompson JF. Radioguided sentinel lymph 39. Morton DL, Thompson JF, Cochran AJ, Mozzillo N, Nieweg OE,
node mapping and biopsy in cutaneous melanoma. In: Hermann Roses DF, et al. Final trial report of sentinel node biopsy versus
K, Nieweg OE, Povosky SP, editors. Radioguided surgery. nodal observation in melanoma. N Engl J Med. 2014;370:599–609.
Switzerland: Springer; 2016. p. 152–63. 40. McMasters KM, Chao C, Wong SL, Wrightson WR, Ross MI,
24. Lens MB, Dawes M, Newton-Bishop JA, Goodacre T. Tumour Reintgen DS, et al. Interval sentinel lymph nodes in melanoma.
thickness as a predictor of occult lymph node metastases in patients Arch Surg. 2002;137:543–9.
with stage I and II melanoma undergoing sentinel lymph node 41. Jimenez RE, Panageas K, Busam KJ, Brady MS. Prognostic impli-
biopsy. Br J Surg. 2002;89:1223–7. cations of multiple lymphatic basin drainage in patients with trun-
25. Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Atkins MB, cal melanoma. J Clin Oncol. 2005;23:518–24.
Byrd DR, et al. Final version of 2009 AJCC melanoma staging and 42. Ipenburg NA, Nieweg OE, Uren RF, Thompson JF. Outcome

classification. J Clin Oncol. 2009;27:6199–206. of melanoma patients who did not proceed to sentinel node
26. Mariani G, Gipponi M, Moresco L, Villa G, Bartolomei M,
biopsy after preoperative lymphoscintigraphy. Ann Surg Oncol.
Mazzarol G, et al. Radioguided sentinel lymph node biopsy in 2017;24:117–26.
malignant cutaneous melanoma. J Nucl Med. 2002;43:811–27. 43. Wong SL, Balch CM, Hurley P, Agarwala SS, Akhurst TJ, Cochran
27. Kretschmer L, Thoms KM, Peeters S, Haenssle H, Bertsch A, et al. Sentinel lymph node biopsy for melanoma: American
HP, Emmert S. Postoperative morbidity of lymph node exci- Society of Clinical Oncology and Society of Surgical Oncology
sion for cutaneous melanoma-sentinel lymphadenectomy ver- joint clinical practice guideline. Ann Surg Oncol. 2012;19:3313–24.
sus complete regional lymph node dissection. Melanoma Res. 44. Siegel RL, Miller KD, Ahmedin Jemal DVM. Cancer statistics
2008;18:16–21. 2017. CA Cancer J Clin. 2017;67:7–30.
28. Bluemel C, Herrmann K, Giammarile F, Nieweg OE, Dubreuil J, 45. Gil Z, Carlson DL, Boyle JO, Kraus DH, Shah JP, Shaha AR, et al.
Testori A, et al. EANM practice guidelines for lymphoscintigraphy Lymph node density is a significant predictor of outcome in patients
and sentinel lymph node biopsy in melanoma. Eur J Nucl Med Mol with oral cancer. Cancer. 2009;115:5700–10.
Imaging. 2015;42:1750–66. 46. Alex JC, Sasaki CT, Krag DN, Wenig B, Pyle PB. Sentinel lymph
29. Niebling MG, Pleijhuis RG, Bastiaannet E, Brouwers AH, van Dam node radiolocalization in head and neck squamous cell carcinoma.
GM, Hoekstra HJ. A systematic review and meta-analyses of senti- Laryngoscope. 2000;110:198–203.
nel lymph node identification in breast cancer and melanoma, a plea 47.
Burusapat C, Jarungroongruangchai W, Charoenpitakchai
for tracer mapping. Eur J Surg Oncol. 2016;42(4):466–73. M. Prognostic factors of cervical node status in head and neck squa-
30. Vitali GC, Trifirò G, Zonta M, Pennacchioli E, Santoro L, Travaini mous cell carcinoma. World J Surg Oncol. 2015;13:51.
LL, et al. Lymphoscintigraphy in clinical routine practice: repro- 48. NCCN Clinical Practice Guideline in Oncology. Head and Neck
ducibility and accuracy in melanoma patients with a long-term Cancer (Version 2.2017). ttps://www.nccn.org/professionals/physi-
follow-up. Eur J Surg Oncol. 2014;40:55–60. cian_gls/pdf/head-and-neck.pdf
31. Ahmadzadehfar H, Hinz T, Wierzbicki A, Höller T, Habibi E, 49. Murer K, Huber GF, Haile SR, Stoeckli SJ. Comparison of morbid-
Wenzel J, et al. Significance of multiple nodal basin drainage in ity between sentinel node biopsy and elective neck dissection for
patients with truncal melanoma. Q J Nucl Med Mol Imaging. treatment of the n0 neck in patients with oral squamous cell carci-
2016;60:274–9. noma. Head Neck. 2011;33:1260–4.
32. Veenstra HJ, Klop WM, Speijers MJ, Lohuis PJ, Nieweg OE,
50. Liu M, Wang SJ, Yang X, Peng H. Diagnostic efficacy of sentinel
Hoekstra HJ, et al. Lymphatic drainage patterns from melano- lymph node biopsy in early oral squamous cell carcinoma: a meta-
mas on the shoulder or upper trunk to cervical lymph nodes and analysis of 66 studies. PLoS One. 2017;12(1):e0170322. https://
implications for the extent of neck dissection. Ann Surg Oncol. doi.org/10.1371/journal.pone.0170322.
2012;19:3906–12. 51. Yamauchi K, Kogashiwa Y, Nakamura T, Moro Y, Nagafuji H,
33. Jimenez-Heffernan A, Ellmann A, Sado H, Huić D, Bal C,
Kohno N. Diagnostic evaluation of sentinel lymph node biopsy
Parameswaran R, et al. Results of a prospective multicenter inter- in early head and neck squamous cell carcinoma: a meta-analysis.
national atomic energy agency sentinel node trial on the value of Head Neck. 2015;37:127–33.
SPECT/CT over planar imaging in various malignancies. J Nucl 52. Robbins KT, Clayman G, Levine PA, Medina J, Sessions R, Shaha
Med. 2015;56:1338–44. A, et al. Neck dissection classification update: revisions proposed by
34. Vidal M, Vidal-Sicart S, Torres F, Ruiz DM, Paredes P, Pons
the American Head and Neck Society and the American Academy
F. Correlation between theoretical anatomical patterns of lym- of Otolaryngology-Head and Neck Surgery. Arch Otolaryngol
phatic drainage and lymphoscintigraphy findings during sentinel Head Neck Surg. 2002;128:751–8.
node detection in head and neck melanomas. Eur J Nucl Med Mol 53. AJCC. Cancer staging handbook. 7th ed. New York: Springer;
Imaging. 2016;43:626–34. 2010.
35. Batistatou A, Cook MG, Massi D, ESP Dermatopathology
54. Creighton F, Bergmark R, Emerick K. Drainage patterns to non-
Working Group. Histopathology report of cutaneous melanoma traditional nodal regions and level IIB in cutaneous head and neck
and sentinel lymph node in Europe: a web-based survey by the malignancy. Otolaryngol Head Neck Surg. 2016;155:1005–11.
Dermatopathology Working Group of the European Society of 55. de Bree R, Nieweg OE. The history of sentinel node biopsy in head
Pathology. Virchows Arch. 2009;454:505–11. and neck cancer: From visualization of lymphatic vessels to senti-
36. Mocellin S, Hoon DS, Pilati P, Rossi CR, Nitti D. Sentinel lymph nel nodes. Oral Oncol. 2015;51:819–23.
node molecular ultrastaging in patients with melanoma: a sys- 56. Shoaib T, Soutar DS, MacDonald DG, Gray HW, Ross GL. The
tematic review and meta-analysis of prognosis. J Clin Oncol. nodal neck level of sentinel lymph nodes in mucosal head and neck
2007;25:1588–95. cancer. Br J Plast Surg. 2005;8:790–4.
37. Manca G, Romanini A, Pellegrino D, Borsò E, Rondini M,
57. Azad AK, Rajaram MV, Metz WL, Cope FO, Blue MS, Vera DR,
Orlandini C, et al. Optimal detection of sentinel lymph node metas- et al. γ-Tilmanocept, a new radiopharmaceutical tracer for cancer
sentinel lymph nodes, binds to the mannose receptor (CD206). J 73. Pedersen NJ, Hebbelstrup Jensen DH, Hedbäck N, Frendø M, Kiss
Immunol. 2015;195:2019–29. K, Lelkaitis G, et al. Staging of early lymph node metastases with
58. Agrawal A, Civantos FJ, Brumund KT, Chepeha DB, Hall NC, the sentinel lymph node technique and predictive factors in T1/T2
Carroll WR, et al. [99mTc]Tilmanocept accurately detects sentinel oral cavity cancer: a retrospective single-center study of 253 con-
lymph nodes and predicts node pathology status in patients with oral secutive patients. Head Neck. 2016;38(Suppl 1):1033–40.
squamous cell carcinoma of the head and neck: results of a phase III 74. Den Toom IJ, Bloemena E, van Weert S, Karagozoglu KH, Hoekstra
multi-institutional trial. Ann Surg Oncol. 2015;22:3708–15. OS, de Bree R. Additional non-sentinel lymph node metastases in
59. Peng H, Wang SJ, Niu X, Yang X, Chi C, Zhang G. Sentinel node early oral cancer patients with positive sentinel lymph nodes. Eur
biopsy using indocyanine green in oral/oropharyngeal cancer. Arch Otorhinolaryngol. 2017;274:961–8.
World J Surg Oncol. 2015;13:278. 75. Alkureishi LW, Burak Z, Alvarez JA, Ballinger J, Bilde A, Britten AJ,
60. van den Berg NS, Brouwer OR, Klop WM, Karakullukcu B, Zuur et al. Joint practice guidelines for radionuclide l ymphoscintigraphy
CL, Tan IB, et al. Concomitant radio- and fluorescence-guided sen- for sentinel node localization in oral/oropharyngeal squamous cell
tinel lymph node biopsy in squamous cell carcinoma of the oral carcinoma. Eur J Nucl Med Mol Imaging. 2009;36:1915–36.
cavity using ICG-99mTc-nanocolloid. Eur J Nucl Med Mol Imaging. 76. Bluemel C, Rubello D, Colletti PM, de Bree R, Herrmann

2012;39:1128–36. K. Sentinel lymph node biopsy in oral and oropharyngeal squa-
61. Khafif A, Schneebaum S, Fliss DM, Lerman H, Metser U, Ben- mous cell carcinoma: current status and unresolved challenges. Eur
Yosef R, et al. Lymphoscintigraphy for sentinel node mapping J Nucl Med Mol Imaging. 2015;42:1469–80.
using a hybrid single photon emission CT (SPECT)/CT system in 77. Flach GB, Broglie MA, van Schie A, Bloemena E, Leemans CR, de
oral cavity squamous cell carcinoma. Head Neck. 2006;28:874–9. Bree R, et al. Sentinel node biopsy for oral and oropharyngeal squa-
62. den Toom IJ, van Schie A, van Weert S, Karagozoglu KH, Bloemena mous cell carcinoma in the previously treated neck. Oral Oncol.
E, Hoekstra OS, et al. The added value of SPECT-CT for the iden- 2012;48:85–9.
tification of sentinel lymph nodes in early stage oral cancer. Eur J 78. Giammarile F, Bozkurt MF, Cibula D, Pahisa J, Oyen WJ, Paredes
Nucl Med Mol Imaging. 2017;44:998–1004. P, et al. The EANM clinical and technical guidelines for lympho-
63. Sieira-Gil R, Paredes P, Martí-Pagés C, Ferrer-Fuertes A, García- scintigraphy and sentinel node localization in gynaecological can-
Díez E, Cho-Lee GY, et al. SPECT-CT and intraoperative portable cers. Eur J Nucl Med Mol Imaging. 2014;41:1463–77.
gamma-camera detection protocol for sentinel lymph node biopsy 79. Collarino A, Donswijk ML, van Driel WJ, Stokkel MP, Valdés
in oral cavity squamous cell carcinoma. J Craniomaxillofac Surg. Olmos RA. The use of SPECT/CT for anatomical mapping of
2015;43:2205–13. lymphatic drainage in vulvar cancer: possible implications for the
64. Borbón-Arce M, Brouwer OR, van den Berg NS, Mathéron H, Klop extent of inguinal lymph node dissection. Eur J Nucl Med Mol
WM, Balm AJ, et al. An innovative multimodality approach for sen- Imaging. 2015;42:2064–71.
tinel node mapping and biopsy in head and neck malignancies. Rev 80. Hoogendam JP, Veldhuis WB, Hobbelink MG, Verheijen RH, van
Esp Med Nucl Imagen Mol. 2014;33:274–9. den Bosch MA, Zweemer RP. 99mTc SPECT/CT versus planar lym-
65. Hellingman D, Vidal-Sicart S, de Wit-van der Veen LJ, Paredes P, phoscintigraphy for preoperative sentinel lymph node detection in
Valdés Olmos RA. A new portable hybrid camera for fused opti- cervical cancer: a systematic review and metaanalysis. J Nucl Med.
cal and scintigraphic imaging: first clinical experiences. Clin Nucl 2015;56:675–80.
Med. 2016;41:39–43. 81. Ouldamer L, Bendifallah S, Body G, Canlorbe G, Touboul C,
66. Hermanek P, Hutter RV, Sobin LH, Wittekind C. International
Graesslin O, et al. Call for surgical nodal staging in women with
Union Against Cancer. Classification of isolated tumor cells and ESMO/ESGO/ESTRO high-intermediate risk endometrial cancer:
micrometastasis. Cancer. 1999;86:2668–73. a multicentre cohort analysis from the FRANCOGYN study group.
67. Kovács AF, Stefenelli U, Seitz O, Middendorp M, Diener J, Sader Ann Surg Oncol. 2017;24:1660–6.
R, et al. Positive sentinel lymph nodes are a negative prognostic 82. Elisei F, Crivellaro C, Giuliani D, Dolci C, De Ponti E, Montanelli
factor for survival in T1–2 oral/oropharyngeal cancer – a long-term L, et al. Sentinel-node mapping in endometrial cancer patients:
study on 103 patients. Ann Surg Oncol. 2009;16:233–9. comparing SPECT/CT, gamma-probe and dye. Ann Nucl Med.
68. Alkureishi LW, Ross GL, Shoaib T, Soutar DS, Robertson AG, 2017;3:93–9.
Thompson R, et al. Sentinel node biopsy in head and neck squa- 83. Klapdor R, Hertel H, Soergel P, Jentschke M, Hillemanns

mous cell cancer: 5-year follow-up of a European multicenter trial. P. Application of sentinel lymph node dissection in gynecological
Ann Surg Oncol. 2010;17:2459–64. cancers: results of a survey among German hospitals. Arch Gynecol
69. Broglie MA, Haile SR, Stoeckli SJ. Long-term experience in
Obstet. 2017;295:713–20.
sentinel node biopsy for early oral and oropharyngeal squamous 84. Te Grootenhuis NC, van der Zee AG, van Doorn HC, van der Velden
cell carcinoma. Ann Surg Oncol. 2011;18:2732–8. J, Vergote I, Zanagnolo V, et al. Sentinel nodes in vulvar cancer:
70. Govers TM, Hannink G, Merkx MA, Takes RP, Rovers
long-term follow-up of the GROningen INternational Study on
MM. Sentinel node biopsy for squamous cell carcinoma of the oral Sentinel nodes in Vulvar cancer (GROINSS-V) I. Gynecol Oncol.
cavity and oropharynx: a diagnostic meta-analysis. Oral Oncol. 2016;140:8–14.
2013;49:726–32.
71. Den Toom IJ, Heuveling DA, Flach GB, van Weert S, Karagozoglu
KH, van Schie A, et al. Sentinel node biopsy for early oral cavity
cancer: the VU University Medical Center experience. Head Neck.
2015;37:573–8.
72. Schilling C, Stoeckli SJ, Haerle SK, Broglie MA, Huber GF,

Sorensen JA, et al. Sentinel European Node Trial (SENT): 3-year
results of sentinel node biopsy in oral cancer. Eur J Cancer.
2015;51:2777–84.
Part II
Organ/Apparatus-Specific Applications
Radionuclide Imaging for Non-tumor
Diseases of the Brain 17
Duccio Volterrani, Giampiero Giovacchini,
and Andrea Ciarmiello
Contents
17.1 Anatomy and Physiology 391
17.1.1 Brain Metabolism 392
17.1.2 Blood-Brain Barrier 392
17.1.3 Brain Cells 392
17.1.4 Blood Flow, Metabolism, and Neuronal Activity 393
17.2 Dementias 393
17.2.1 [18F]FDG PET/CT 394
17.2.2 Amyloid Imaging with PET/CT 394
17.2.3 rCBF SPECT 396
17.2.4 Movement Disorders 398
17.2.5 Dopamine Transporter SPECT 399
17.2.6 18F-DOPA PET/CT 401
17.2.7 Differential Diagnosis of Degenerative Parkinsonisms 401
17.2.8 SPECT Imaging of Brain Perfusion 404
17.2.9 rCBF SPECT in Epilepsy 406
17.2.10 rCBF SPECT with Acetazolamide Test 407
17.2.11 Brain Death Scintigraphy 407
17.2.12 CSF Imaging 410
References 411
Learning Objectives • Learn advantages and disadvantages of different tech-

• Understand the basic notions on anatomy and physiology niques such as SPECT, SPECT/CT and PET/CT for diag-
of the central nervous system (CNS). nosing and characterizing non-tumor disease of the brain.
• Understand the pathophysiology of Alzheimer’s disease • Learn properties of radiotracers used for brain imaging
(AD), dementia with Lewy bodies (DLB), frontotemporal and the rationale to their use in different clinical settings
dementia (FTD), Parkinson’s disease (PD), Parkinson (dementias, Parkinson’s disease, multiple system atrophy,
plus syndromes, and epilepsy. progressive supranuclear palsy, corticobasal degenera-
tion, DLB movement disorders, epilepsy).
• Understand the principles of image interpretation for
radionuclide imaging of non-tumor disease of the brain.
D. Volterrani (*)
Research and Advanced Technologies in Medicine and Surgery, 17.1 Anatomy and Physiology
e-mail: duccio.volterrani@med.unipi.it
The central nervous system (CNS) is formed by the brain
G. Giovacchini · A. Ciarmiello and the spinal cord. Both structures are composed of gray
Nuclear Medicine Department, “S. Andrea” Hospital,
substance (neurons) and white substance (nerve fibers).
La Spezia, Italy

The telencephalon, diencephalon, cerebellum, and brain nuclei, and fibers connecting with the cerebellum. The sub-
stem are parts of the brain. The telencephalon includes the stantia nigra is a small nucleus within the midbrain homing
cerebral cortex. Nine tenths of the cortex is neocortex, which dopaminergic neurons, which are the major source of affer-
includes areas of sensation (except for olfaction), motor ent fibers to the putamen and the caudate nucleus.
areas, and large expanses of association areas where superior Degeneration of the substantia nigra is the basis of the main
functions take place. The neocortex of the frontal lobe has a disorders of motor functions occurring in parkinsonisms.
special role in motor activities, in determining mood, and in Along with the SNC, it is worth mentioning the auto-
judgment and foresight. nomic nervous system, which includes the whole of cells and
In the frontal lobe, the precentral gyrus and the paracen- fibers that innervate the internal organs and glands, perform-
tral lobule on the medial face include the primary motor area. ing functions that are out of voluntary control. The efferent
Anterior to the primary motor, there is the premotor area; route (from CNS to innervated organs) is always composed
lesions at this level cause apraxias (difficulty performing of two neurons, a neuron at the CNS level (cranial or spinal
purposeful movements or speech). A supplementary motor head) and a ganglionic neuron that is homed in central or
area is in the dorsal wall of the lateral sulcus within the fron- peripheral ganglia or in the wall innervated.
tal lobes. The large expanse of the cortex called the frontal
pole (which includes the superior, middle, and inferior fron-
tal gyrus) represents the prefrontal cortex that, when involved 17.1.1 Brain Metabolism
by neurodegenerative processes, is responsible for cognitive
impairment and behavioral changes. The left frontal lobe has The brain is a highly active metabolic organ that requires a
also specialized language functions located in the medial continuous supply of oxygen and of its main energy sub-
surface of the supplementary motor area and in the expressive strate, glucose, which provides about 95% of the energy
speech area (Broca). Expressive aphasia (hesitant and dis- required for its functions. Neurons cannot metabolize fatty
torted speech) is caused by lesions in Broca’s area. acids in the same way as other tissues do and do not have an
Sensory areas occupy mainly the postcentral gyrus and intracellular energy reserve. Therefore, although it repre-
paracentral lobule of the parietal lobe, whereas the somes- sents only 2% of body weight, the brain receives about 20%
thetic association cortex is mainly in the superior parietal lob- of the heart output. During anoxia, brain cells quickly start to
ule and in the precuneus. Destructive lesions of these areas suffer and die within a few minutes, with irreversible conse-
produce a defect in understanding the significance of sensory quences on the affected brain functions.
information, called agnosia. The vision area surrounds the
calcarine sulcus and extends over the occipital lobe.
Association areas within the parietal and temporal lobes 17.1.2 Blood-Brain Barrier
are responsible for many of the unique qualities of the human
brain and form the basis of learning at intellectual level. The blood-brain barrier (BBB) is a characteristic barrier of
The paleocortex receives stimuli from the olfactory sys- brain capillaries that is designed to defend the brain from
tem, while the archicortex includes the limbic system, which possible toxic circulating substances. Crossing through the
is involved in memory (hippocampus) and in behavioral and BBB occurs passively or via active transport mechanisms for
emotional changes (amygdala). small liposoluble molecules, while macromolecules and
The diencephalon is the central core of the brain and highly charged substances are not allowed to pass. Integrity
includes the thalamus that consists of different nuclei receiv- of the BBB is reduced in case of trauma, ischemia, inflam-
ing and projecting stimuli to the sensory cortex. The nuclei mation, and infection.
of the base are gray matter formations located deep in the
white matter of the cerebral hemispheres. They include the
caudate nucleus, the lenticular nucleus (putamen and globus 17.1.3 Brain Cells
pallidus), the claustrous, and the amygdala, and they are
involved in determining motion parameters. In the brain, there are two main categories of cells, neurons
The cerebellum is located in the lower back portion of the and neuroglia. Neurons are perennial cells carrying electrical
cranial cavity beneath the occipital lobe and consists of two signals. Synapses are the connections among neurons, allow-
hemispheres and a central portion (vermis). It is mainly ing the passage of the nervous pulse through the release of
involved in the coordination of movements; lesions at this neurotransmitters within the synaptic space. There are sev-
level are responsible of symptoms such as gait ataxia, dys- eral highly specialized neuronal populations using specific
metria, and asynergy. neurotransmitters: gamma-aminobutyric acid (GABA, gen-
The brain stem (bulb, pons, and midbrain) includes erally with an inhibitory action), glutamate (mainly excit-
ascending and descending tracts, cranial nerve and several atory action), biogenic amines (serotonin, noradrenaline,
17 Radionuclide Imaging for Non-tumor Diseases of the Brain 393
dopamine, acetylcholine, with mixed actions), and neuro-

peptides (endorphins). major source of afferent fibers to the putamen and
Neuroglia is composed of cells (astrocytes) that do not the caudate nucleus.
transmit electrical impulses but play a primary role in sup- • Blood flow and brain metabolism, on one hand, and
porting and protecting neurons. the presence of the blood-brain barrier, on the other
side, make the brain a unique organ.
• Neurons have no intracellular metabolic stores and
17.1.4 Blood Flow, Metabolism, therefore require a continuous supply of oxygen
and Neuronal Activity and glucose for their functions. Any increase in syn-
aptic activity in a particular brain area results in a
Blood flow and brain metabolism, on one hand, and the pres- rapid local increase in blood flow and regional
ence of BBB, on the other side, make the brain a unique metabolism.
organ. In fact, there is no similar structure to BBB in other • The main determinant of glucose consumption is
organs, and there is a very close correlation between blood the synaptic stimulus mediated by glutamate.
flow, metabolism, and neuronal activity. This close correla- Astrocytes act likely to ensure the relationship
tion is due to the fact that neurons have no intracellular meta- between blood flow to different areas of the brain
bolic stores and thus require a continuous supply of oxygen and their levels of synaptic activity.
and glucose for their functions. Thus, an increase in synaptic
activity in a particular brain area results in a rapid local
increase in blood flow and regional metabolism. Thanks to
such a close relationship, variations in blood flow and/or
metabolism can be used as a surrogate for variations in neu- 17.2 Dementias
ronal activity.
The main determinant of metabolic consumption is the The term dementia is used for a variety of neurodegenera-
synaptic stimulus mediated by glutamate, the leading neuro- tive disorders associated with considerable disability and
nal excitatory neurotransmitter. However, astrocytes are also high costs to individuals and to society. The most common
likely to act as “intermediaries” to ensure the relationship neurodegenerative disorders responsible for dementia, such
between blood flow to different areas of the brain and their as Alzheimer’s disease (AD), dementia with Lewy bodies
levels of synaptic activity. In fact, by detecting the presence (DLB), and frontotemporal dementia (FTD), have different
of glutamine at the synaptic level, astrocytes probably stimu- underlying pathogenetic mechanisms, and each is associ-
late vasodilation, thus increasing the blood flow. ated with a distinctive course and its own set of possible
Several techniques and many radiopharmaceuticals have complications [1].
been used for clinical and experimental studies of the The diagnosis of dementia is currently established in terms
CNS. This chapter will focus on those techniques that can be of probability and is based on clinical features. However,
applied in clinical routine, particularly on those employed in accuracy of the clinical diagnostic criteria is limited c ompared
the evaluation of patients with neurodegenerative disorders with postmortem histopathology. Even though medications
such as dementias and movement disorders. are available to reduce some symptoms, there are no specific
treatments of the underlying disease. An accurate diagnosis
of dementia is crucial to avoid inappropriate interventions
Key Learning Points and provide guidance for appropriate non-pharmacological
• The central nervous system is formed by different management.
structures which are composed of gray substance The recent development of imaging biomarkers as an
(neurons) and white substance (nerve fibers). optional tool for the diagnosis of dementias may lead to bet-
• The neocortex includes areas of sensation, motor ter clinical management of the affected patients. Moreover,
areas, and large expanses of association areas where such biomarkers may also make it possible to achieve early
superior functions take place. etiological characterization of the prodromal stage of
• The basal ganglia, such as the caudate nucleus, len- dementia, also known as mild cognitive impairment (MCI).
ticular nucleus (putamen and globus pallidus), This is a particularly important aspect, since the disease-
claustrous, and amygdala, are involved in determin- modifying drugs currently under development are expected
ing motion parameters. Dopaminergic neurons of to be more effective in the early disease phase. Neuronal
the substantia nigra within the midbrain are the damage or dysfunction and specific pathological features of
dementias are the targets of the most widely validated PET
biomarkers [2, 3].
17.2.1 [18F]FDG PET/CT Despite some overlap in the patterns of distribution of the
brain areas with reduced metabolism, [18F]FDG PET can be
Assessment of neurodegeneration using [18F]FDG PET/CT used for the differential diagnosis of the major neurodegen-
has been extensively evaluated in dementia patients. In fact, erative disorders, with higher diagnostic accuracy than clini-
the cerebral metabolic rate of glucose (CMRglc) reflects syn- cal judgment [6]. Even though the hypometabolic pattern in
aptic activity and density. patients with DLB can be similar to that observed in AD,
involvement of the occipital cortices is usually more severe—
17.2.1.1 Technique while the temporal lobes are less affected [6].
The activity to be injected should be adjusted according to
available instrumentation and patient’s age. In adults the
average reference activity is around 185 MBq. 17.2.2 Amyloid Imaging with PET/CT
In consultation with the referring physician and consider-
ing the clinical query, it is advisable to instruct the patient Βeta-amyloid (Aβ) plaques are present in the cortical gray
about whether or not to discontinue any drug acting on the matter in all cases of AD, in 50–70% of patients with DLB,
CNS and on the need to keep fasting for at least 6 h before and in normal elderly persons with percentages that increase
the scan. Prior to administration of [18F]FDG, blood glucose with age. Since the advent of the first amyloid tracer, the
should be measured as done for oncological studies. For 11
C-labeled Pittsburgh compound B ([11C]PiB) [7], several
patients with known glucose intolerance, the exam should be studies have been published assessing the amyloid status of
preferably scheduled early in the morning. individuals with dementia.
Acquisition in 3D mode is highly recommended and In order to overcome the limitations associated with the
should start about 40 min after [18F]FDG administration and short half-life of 11C (20 min), 18F-labeled Aβ-targeting trac-
should last 15–20 min. The FOV is usually set at 30 cm and ers have been developed, including 18F-3′-F-PiB (flute-
the matrix of reconstruction at 128 × 128 (or 256 × 256 in metamol, Vizamyl®, GE Healthcare), 18F-AV-45 (florbetapir,
newer scanner). Amyvid®, Ely Lilly), 18F-AV-1 or 18F-BAY94-9172 (florbeta-
ben, Neuraceq®, Piramal), and 18F-AZD4694 or NAV4694
17.2.1.2 Image Analysis (Fig. 17.3). A high correlation between uptake of [11C]PiB
Careful evaluation of all sections (transaxial, coronal, and and of 18F-labeled ligands has been demonstrated [7], sup-
sagittal) is mandatory. The coronal sections are fundamen- porting the translation of the [11C]PiB PET findings into the
tal for evaluating lateral and deep temporal regions and for domain of 18F-labeled tracers [8] (Fig. 17.4).
evaluating the ventricular system along this axis. Attention PET studies with [11C]PiB studies have shown remarkable
should be paid to the upper transaxial sections for the pos- differences in tracer uptake between AD patients and age-
sible presence of anatomical asymmetries of the grooves matched controls, with very high sensitivity and high nega-
especially in elderly patients (due to cerebral atrophy); tive predictive value. On the other hand, specificity and
these areas can therefore be the sites of significant asym- positive predictive value decline with age, with diagnostic
metries of tracer uptake. Such asymmetries should be eval- accuracy for AD not higher than 75–80% in subjects over
uated with reference to extension, degree, and involvement 80 years of age. However, if the purpose of the diagnostic
of adjacent sections and correlated with the corresponding assessment is to evaluate the presence of Aβ plaques, the
CT images. accuracy of amyloid imaging remains high at all ages.
If a standard control dataset is available, a stereotactic Whereas a positive amyloid PET scan does not establish with
standardization and voxel-based analytical approach can be certainty a diagnosis of AD, it can help to exclude or support
employed to detect changes of metabolism in an observer- the presence of AD in the context of a more comprehensive
independent manner—for improved diagnostic accuracy. evaluation of the patient [8].
Metabolic deficits in AD patients are present in the neo- An important clinical application of amyloid PET imag-
cortical association areas, with reduced [18F]FDG uptake ing is the differential diagnosis between AD and FTD, as Aβ
detected predominantly in the temporoparietal regions, the pathology is not present in the latter. In a recent study includ-
posterior cingulate cortex, and the frontal cortex [4] ing 62 AD patients and 45 FTD patients, [11C]PiB and [18F]
(Fig. 17.1). Metabolic reductions in the medial occipital cor- FDG PET showed similar diagnostic accuracy. However,
tex are typically observed in DLB, while metabolic impair- [11C]PiB was more sensitive and showed higher inter-rater
ments in FTD typically involve the frontal or frontotemporal reliability and agreement between qualitative and quantita-
regions [5] (Fig. 17.2). [18F]FDG PET has been shown to tive evaluation [9].
correctly distinguish AD subjects from healthy controls with According to the amyloid cascade hypothesis [10], accu-
high sensitivity and specificity, using either clinical assess- mulation of Aβ during early stages of the disease precedes a
ment or pathological confirmation as reference standard. cascade of neuropathological events that eventually leads to
Fig. 17.1 Brain PET with [18F]FDG in a patient with Alzheimer’s disease. Bilateral reduction of [18F]FDG uptake is observed mainly in the lateral
and mesial temporal cortex, parietal cortex, and precuneus. Mild hypometabolism is clearly seen within the left prefrontal cortex
AD. Thus, intense efforts have been directed at evaluating [11C]PiB uptake than non-converting patients, comparable to
the possible role of amyloid imaging in the prognostic that of AD patients.
assessment of patients with MCI and of cognitively normal The findings of several longitudinal studies, which set out
elderly individuals. Even though several studies support the to compare [11C]PiB retention, [18F]FDG uptake, and cogni-
notion that many healthy elderly individuals with a positive tive impairment in AD and MCI patients, suggest a bidirec-
[11C]PiB PET are in a preclinical phase of AD, this hypoth- tional interrelationship between neurodegeneration assessed
esis requires further longitudinal investigation [11]. On the by [18F]FDG PET and Aβ deposition evaluated by [11C]PiB
other hand, current data suggest that amyloid PET could cor- PET [13]. Amyloid-positive regions maintain neuronal activ-
rectly stratify MCI patients, identifying those who will con- ity in an early preclinical phase, followed by progressive
vert to AD [12]. The hypothesis is that MCI patients who will neuronal degeneration coupled with cognitive decline.
convert to AD during follow-up have a significantly higher Whereas [18F]FDG PET can be used to determine the degree
Fig. 17.2 Brain PET with [18F]FDG in a patient with frontotemporal dementia. Asymmetrical reduction in metabolic activity (more pronounced
on the right hemisphere) is observed within the frontal, insular, parietal, temporolateral, and anterior cingulate cortex
and progression of neurodegeneration [14], amyloid PET to the timing of the onset of dementia symptoms relative to
imaging appears to be a powerful tool for the early diagnosis parkinsonism [16].
of AD pathology that may lead to a greater level of diagnos-
tic confidence in particular cases, such as MCI patients,
patients with possible AD and atypical clinical presentation, 17.2.3 rCBF SPECT
or patients with atypically young-onset dementia.
In DLB, amyloid deposits are associated with AD-like Brain perfusion imaging with the two most commonly employed
atrophy, predominantly involving the parahippocampal area single-photon tracers, 99mTc-hexamethylpropylenamine (99mTc-
and lateral temporal and parietal cortices [15]. In a study HMPAO) and 99mTc-ethylcysteinate dimer (99mTc-ECD), has
comparing DLB patients with patients affected by Parkinson’s been extensively used for the diagnosis, therapeutic manage-
disease dementia (PDD), the presence of amyloid plaques in ment, and follow-up of dementia patients [17]. In patients with-
the DLB patients was linked to the cognitive impairment and out cerebrovascular disease, hypoperfusion is related to impaired
Florbetaben Florbetapir Flutemetamol

NeuraCeq ® Amyvid ® Vizamyl ®
Fig. 17.3 Patterns of normal distribution for the three different amyloid 18F-labeled radiotracers commercially available
Fig. 17.4 Left panel:

transaxial slice of
18
F-florbetapir PET in a
patient without significant
brain uptake of the amyloid
tracer. Right panel: transaxial
slice in a patient with high
cortical uptake of
18
F-florbetapir due to the Negative Positive
presence of a pathologic
burden of amyloid
neuronal activity and is coupled with decreased metabolism

detected by [18F]FDG PET, showing similar patterns of reduced • The diagnosis of dementia is currently established
regional uptake in different types of dementia (Fig. 17.5). in terms of probability and is based on clinical
However, [18F]FDG PET performs better in discriminating AD features.
patients from controls, with 100% sensitivity compared to 90% • Neuronal damage or dysfunction and specific path-
for brain perfusion SPECT [18]. ological features of dementias are the targets of the
Thus, whenever it is possible to choose between PET and most widely validated PET biomarkers.
SPECT, PET should be preferred. Nevertheless, SPECT can • Assessment of neurodegeneration and dysfunction
be considered a reliable alternative. using [18F]FDG PET/CT has been extensively eval-
uated in dementia patients.
• 18F]FDG PET has been shown to correctly distin-
guish AD subjects from healthy controls with high
Key Learning Points
sensitivity and specificity. [18F]FDG PET can be
• AD, DLB, and FTD are the most common neurode-
used for the differential diagnosis of the major neu-
generative disorders responsible for dementia, they
rodegenerative disorders.
have different underlying pathogenetic mecha-
• Βeta-amyloid (Aβ) plaques are present in the corti-
nisms, and each is associated with a distinctive
cal gray matter in all cases of AD, in 50–70% of
course.
Fig. 17.5 Transaxial sections at different levels as obtained from a Perfusion is maintained in the occipital and sensory-motor areas (repro-
brain perfusion SPECT with 99mTc-ECD in a patient with Alzheimer’s duced with permission from Volterrani D, Erba PA, Mariani G, Eds.
disease: reduced perfusion (mirroring reduced metabolism) is detected Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
in the temporoparietal and frontal areas bilaterally (white arrowheads). Springer Italy; 2010)
(MSA), progressive supranuclear palsy (PSP), cortico-

patients with DLB, and in normal elderly persons basal degeneration (CBD), and DLB. All these disorders
with percentages that increase with age. are currently lumped into two groups, the synucleinopa-
• A positive amyloid PET scan does not establish thies and the tauopathies. Other parkinsonian syndromes
with certainty a diagnosis of AD, but it can help to due to vascular, toxic, postencephalitic, or iatrogenic
support the presence of AD. A negative amyloid causes occur less frequently and are often easier to
PET excludes AD. Amyloid PET could also stratify diagnose.
MCI patients, detecting those who will convert to The clinical diagnosis of PD and the discrimination with
AD. and between the various parkinsonian syndromes are based
• An important clinical application of amyloid PET on standardized and reproducible clinical diagnostic criteria.
imaging is the differential diagnosis between AD However, new biomarkers have a role in supplementing clin-
and FTD. ical assessment, becoming significant especially in individu-
als with early onset of symptoms and in atypical clinical
presentation where clinical diagnosis is less
17.2.4 Movement Disorders straightforward.
High-field magnetic resonance imaging (MRI) is gaining
The term “parkinsonism” defines a variety of different increasing importance as a morphologic imaging approach
underlying pathologies that can cause symptoms such as in the diagnosis of PD and all parkinsonian disorders.
bradykinesia, rest tremor, rigidity, and gait abnormalities. The neuropathological marker of PD is the loss of dopa-
The disorders that produce these symptoms referred to as minergic neurons of the substantia nigra (in particular of the
degenerative parkinsonisms are Parkinson’s disease (PD) pars compacta) and of nigrostriatal terminals, associated
and the so-called atypical parkinsonian or Parkinson-plus with the presence of intracellular Lewy inclusion bodies,
syndromes. These latter include multiple system atrophy also in neurons of the cortex of most non-demented PD
patients. Serotonergic cells in the median raphe, noradrener- nervous system stimulants ephedrine and p hentermine. Anti-
gic cells in the locus coeruleus, and cholinergic cells in the parkinsonian drugs do not interfere with 123I-FP-CIT binding
nucleus basalis are also involved—although to a lesser to DaT to any significant degree. Since 123I-FP-CIT is supplied
extent. In Parkinson-plus syndromes, the degenerative pro- in an ethanol solution, caution is recommended when admin-
cess is relatively widespread involving other brain regions istering the tracer to alcoholic patients. The recommended
besides the substantia nigra such as the basal ganglia, brain activity is 111–185 MBq, typically 185 MBq.
stem, midbrain, cerebellum, and cortex. SPECT acquisition should be started when the ratio of
Thus, the prominent biomarker for molecular imaging of striatal to occipital 123I-FP-CIT binding is stable, between 3
PD and all Parkinson-plus syndromes consists in the demon- and 6 h after radiotracer injection. The field of view should
stration of the nigrostriatal dopaminergic pathway depletion, include the whole brain, using the smallest possible rotational
which is related to the degeneration of the substantia nigra radius (<16 cm). A 128 × 128 matrix is recommended with a
neurons. The loss of dopamine neural terminals may be zoom factor yielding a pixel size of 3–4 mm. Step-and-shoot
assessed in vivo with different positron-emitting and single- mode with angle increments of 3° over a circular orbit of 360°
photon-emitting ligands that enable to assess either the incor- is recommended (with total acquisition time between 30 and
poration of metabolic substrates involved in the synthesis of 45 min). A minimum of 1.5 million total counts should be
dopamine within the nigrostriatal terminals related to specific collected. Iterative reconstruction is preferred, with the low-
enzymatic activities, or the density of dopamine transporters pass Butterworth filter. Attenuation correction should be per-
(DaT) that are specific cell membrane transport presynaptic formed using an acquired transmission or CT scan or a
sites for dopamine reuptake, or the density of vesicle mono- calculated correction matrix according to Chang’s method.
amine transporters (VMAT2) in dopamine terminals. Visual and semiquantitative analysis can be used for inter-
pretation. In visual interpretation, normal striata on transax-
ial images should look comma-shaped with symmetric
17.2.5 Dopamine Transporter SPECT well-delineated borders. The head of the caudate and the
putamen usually have high contrast versus background
Cocaine analogs that bind selectively and with high affinity (Fig. 17.6). In an abnormal SPECT, reduced uptake can be
to DaT have been proposed for both PET and SPECT, pro- seen on one or both sides, especially within the putamen,
viding an accurate approach to assess dopamine nigrostriatal
pathway degeneration occurring in degenerative parkinson-
isms and also in patients presenting with minor clinical
symptoms. Various ligands labeled with either iodine-123,
such as 123I-β-CIT, 123I-IPT, and 123I-FP-CIT, or with carbo-
nium-11, such as [11C]FE-CIT, have been developed. A
99m
Tc-labeled cocaine analog, 99mTc-TRODAT-1, has been
also introduced for DaT imaging with SPECT. In particular,
SPECT with 123I-β-CIT and 123I-FP-CIT demonstrated a clear
loss of striatal DaT in patients with PD, with high target-to-
nontarget ratios [19, 20]. Especially 123I-FP-CIT has been
increasingly used in clinical practice, mainly because of its
commercial availability in both the European Union and the
United States (DatScan®, GE Healthcare) and because of fast
kinetics which allows imaging within a few hours after i.v.
administration.
17.2.5.1 Technique
Preparation of the patient involves administration of Lugol
solution (100 mg) or potassium perchlorate (400 mg) at least
30–60 min before tracer injection, to block thyroid uptake of
radioiodide released during tracer degradation. An antidepres-
sant drug, sertraline, may increase binding to DaT and must
therefore be discontinued a few days before the scan. However,
sertraline is not expected to interfere with visual interpretation
Fig. 17.6 Transaxial slice of brain SPECT with 123I-FP-CIT in a
of images. Cocaine, amphetamines, and methylphenidate patient without nigrostriatal degeneration, showing a normal pattern of
decrease 123I-FP-CIT binding to DaT, as also do the central uptake within both striata
Fig. 17.7 Brain SPECT with 123I-FP-CIT in two patients with mild and dates, so that non-specific uptake in the cerebral cortex becomes more
advanced Parkinson’s disease, respectively. The four slices on the left obvious (reproduced with permission from Volterrani D, Erba PA,
show reduction of uptake mainly in the right putamen, an index of Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
degeneration of nigro-putaminal fibers. The four slices on the right Applicazioni. Milan: Springer Italy; 2010)
show markedly reduced [123I]FP-CIT uptake in both putamen and cau-
which is usually more severely affected than the caudate

nucleus (Fig. 17.7). Semiquantitative analysis allows a more
objective measurement of striatal binding ratios. Manual
ROIs, semiautomated ROIs, or 3D anatomical VOIs can be
used. Left and right caudate and putamen ratios must be
quantified separately using an occipital ROI (or VOI) as
background ROIs (Fig. 17.8). A reference database of age-
matched normal subjects can help for the interpretations in
patients who are difficult to classify.
17.2.5.2 Clinical Indications

Several studies have demonstrated the high accuracy of DaT
imaging with 123I-FP-CIT SPECT in differentiating patients
with PD and Parkinson-plus syndromes from those with
essential tremor or other neurological diseases that do not
involve the nigrostriatal dopaminergic pathway [21]. Recently,
it has been demonstrated that SPECT with 123I-FP-CIT is clini-
cally useful in differentiating DLBD from Alzheimer’s dis-
ease. Instead, 123I-FP-CIT is not designed to distinguish among
PD, MSA, and PSP. 123I-FP-CIT SPECT is also used for the
early diagnosis of parkinsonian syndromes and for differenti-
ating presynaptic parkinsonian syndromes from parkinson-
isms without presynaptic dopaminergic loss, such as
drug-induced parkinsonism or psychogenic parkinsonism. Fig. 17.8 Brain SPECT with 123I-FP-CIT: semiquantitative analysis by
On the other hand, the effectiveness of 123I-FP-CIT SPECT means of predefined ROIs that are positioned on caudate and putamen
nuclei and on the occipital region (reproduced with permission from
as a screening or confirmatory test and for monitoring disease Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
progression or response to therapy has not been established. Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
17.2.6 18F-DOPA PET/CT ratios. Absolute quantitative analysis can also be performed,
based on a 90-min dynamic acquisition.
Uptake of 18F-6-fluorodopa (18F-DOPA) within the presynap- 18
F-DOPA PET can be considered as a biomarker for
tic dopaminergic terminals involves several steps. After in vivo measurement of striatal dopamine levels and AADC
crossing the BBB via the large neutral amino acid transport- activity, but potentially less sensitive in reflecting dopami-
ers, 18F-DOPA is actively transported into the nigrostriatal nergic neuronal loss, particularly in early PD. The clinical
terminals, where the aromatic amino acid decarboxylase indications for 18F-DOPA PET are the same as for DaT
enzyme (AADC) converts 18F-DOPA to 18F-dopamine, which SPECT.
is transported via the VMAT2 into the storage vesicles.
Reduced uptake of 18F-DOPA is related to the loss of
neurons in the substantia nigra and consequently of the 17.2.7 Differential Diagnosis of Degenerative
nigrostriatal pathways. However, several studies reported a Parkinsonisms
direct correlation between 18F-DOPA uptake and the striatal
and nigral dopamine levels and AADC activity that can be Presynaptic dopaminergic imaging allows the differentiation
considered the rate-limiting step in the dopamine synthesis. between the absences of nigrostriatal degeneration on one
Thus, while PET with 18F-DOPA has demonstrated its side and PD or Parkinson-plus syndromes on the other side.
potential for PD diagnosis [22], the uptake of 18F-DOPA in In fact, presynaptic dopaminergic nerve terminals are
the striatum is only partly related to the degree of degenera- involved in most degenerative parkinsonian disorders.
tion of the substantia nigra, because of the upregulation of Although the degree of caudate involvement or patterns of
AADC activity in survived cells in order to increase dopa- symmetry/asymmetry or caudate-to-putamen ratios can pro-
mine turnover [23]. vide some differentiation on a group basis, this imaging
For routine clinical use, a static acquisition is performed strategy in the differential diagnosis of parkinsonian syn-
between 60 and 90 min after tracer administration (mean dromes and in particular in distinguishing between PD and
activity 185 MBq), when striatal uptake reflects the activity various Parkinson-plus syndromes is limited.
of the enzyme AADC and transport of the tracer into the On the basis of postmortem investigations that have
vesicles of the presynaptic terminal. found a striatal pathology characterized by a loss of striatal
Image analysis can be visual (Fig. 17.9) or semiquantita- postsynaptic D2 receptors in Parkinson-plus syndromes,
tive (similar to DaT SPECT), by calculating striata occipital postsynaptic dopaminergic imaging has been used for the
Fig. 17.9 Brain PET with 18F-DOPA. Left panel: transaxial slice in a slice in a patient with a degenerative parkinsonism, showing severely
patient without nigrostriatal degeneration, showing normal and sym- reduced uptake in both putamina
metrical uptake within the caudate and putamen. Right panel: transaxial
differential diagnosis between these disorders and diagnosing DLB has been reported. Nevertheless, possible
PD. Several studies assessing the postsynaptic D2 receptors causes of false-negative and false-positive findings should be
density with benzamide derivative ligands, such as [11C] considered when interpreting the [123I]MIBG scintigraphy
raclopride and 123I-iodobenzamide (123I-IBZM), have shown findings.
increased binding in untreated patients with early PD, espe-
cially in the putamen contralateral to the more affected 17.2.7.2 [18F]FDG PET/CT
body side. This phenomenon is due to upregulation of post- The measurement of regional glucose metabolism with [18F]
synaptic dopamine receptors (and to vacancy of such recep- FDG allows assessment of the effect of nigrostriatal degen-
tors) caused by degeneration of the presynaptic afferent eration on brain regions functionally related to the dopami-
neuronal axons. At the same time, reduced striatal binding nergic system, thus providing information about the different
in patients with Parkinson-plus syndromes has been brain systems involved in the neurodegenerative process and
reported, thus indicating reduced D2 receptor density due their modulation with therapy. Therefore, this “network anal-
to neuronal cell death within the basal ganglia [24]. ysis” allows use of [18F]FDG PET to characterize parkinso-
However, decreased D2 receptor binding has been observed nian disorders by revealing characteristic regional patterns
in both PSP and MSA and is also impaired in patients tak- associated with the motor and cognitive features of PD and
ing dopaminergic drugs. Moreover, false-negative findings identifying specific and highly stable metabolic patterns for
have been described. For these reasons, different tracer different Parkinson-plus syndromes. MSA is characterized
approaches have been proposed to better differentiate PD by metabolic reductions in the putamen and cerebellum,
and Parkinson-plus syndromes. whereas PSP is characterized by metabolic declines predom-
inantly in the upper brain stem and medial prefrontal cortex
17.2.7.1 [123I]MIBG Scintigraphy as well as in the medial thalamus, the caudate nuclei, the
Orthostatic hypotension associated with parkinsonism is anterior cingulate area, and the superior frontal cortex.
considered a characteristic trait of MSA, in particular when Patients with CBD exhibit an asymmetrical decrease in the
the orthostatic hypotension is severe or associated with brain [18F]FDG metabolism in the cortical (frontal and pari-
other signs of autonomic failure. However, many patients etal) and subcortical regions, whereas DLBD, which has
with PD may present orthostatic hypotension, caused by the visual hallucinations as the most characteristic neuropsychi-
disease itself or by pharmacologic treatment with L-dopa. atric feature, is characterized by reduced metabolism in the
Although the pattern of dysfunction of the sympathetic sys- bilateral occipital cortex.
tem of PD remains yet to be fully elucidated, the presence of
Lewy bodies in sympathetic ganglia of patients with PD,
with or without autonomic deficit, but not in patients with
MSA, is considered a hallmark of PD and Lewy body dis-
ease compared to MSA [25]. The demonstration in vivo of Key Learning Points
degeneration of the adrenergic post-ganglionic pathways is • Degenerative parkinsonisms include PD and the so-
feasible by highlighting the reduced accumulation of spe- called atypical parkinsonian (or Parkinson-plus
cific radiopharmaceuticals within the adrenergic termina- syndromes) such as MSA, PSP, CBD, and DLB.
tions in the heart. • The main biomarker for molecular imaging of PD
Several PET studies reported a reduced myocardial uptake and all Parkinson-plus syndromes consists in the
of [11C]meta-hydroxyephedrine or 18F-fluorodopamine in nigrostriatal dopaminergic pathway depletion, related
patients with PD and dysautonomia compared to normal to the degeneration of the substantia nigra neurons.
subjects and to patients with MSA. Similar results have been • Cocaine analogs that bind selectively and with high
reported in studies that have used the radiopharmaceutical affinity to DAT have been proposed for both PET
[123I]MIBG. Furthermore, when evaluating patients with and SPECT, providing an accurate approach to
MSA and PD, with and without dysautonomia, Courbon assess dopamine nigrostriatal pathway degeneration
et al. emphasized reduced cardiac uptake of [123I]MIBG as an occurring in degenerative parkinsonisms.123I-FP-
expression of sympathetic dysfunction in patients with PD CIT has been increasingly used in clinical practice,
without clinical signs and symptoms of dysautonomia, con- mainly because of its commercial availability in
cluding that the adrenergic post-ganglionic neurodegenera- both the European Union and the United States.
tion in PD is present even when dysautonomia is not clinically • SPECT with 123I-FP-CIT SPECT accurately differ-
apparent [26]. [123I]MIBG scintigraphy (Fig. 17.10) has been entiates patients with PD and Parkinson-plus syn-
reported to be a sensitive indicator of the presence of PD or dromes from those with essential tremor or other
of neurodegenerative diseases and in particular a marker of neurological diseases that do not involve the nigros-
disease with Lewy bodies. The potential of [123I]MIBG for
Fig. 17.10 Neuroadrenergic myocardial scintigraphy with [123I] heart/mediastinum uptake ratio. Right panel: planar scintigraphy and
MIBG. Left panel: planar scintigraphy and SPECT/CT (3 h postinjec- SPECT/CT (3 h postinjection) in a patient with Parkinson’s disease,
tion) in a patient with dysautonomia and parkinsonism (MSA), showing showing very low myocardial uptake of [123I]MIBG
normal myocardial tracer uptake. ROIs are positioned to calculate the
triatal dopaminergic. It is also clinically useful in the basal ganglia is a biomarker of Parkinson-plus
differentiating DLB from AD. Visual and semi- syndromes.
quantitative analysis can be used for interpretation. • [123I]MIBG scintigraphy has been reported to be a
• 18F-DOPA PET is a biomarker for in vivo measure- sensitive indicator of the presence of PD and in par-
ment of striatal dopamine levels and AADC activ- ticular a marker of disease with Lewy bodies. The
ity, but potentially less sensitive in reflecting potential of [123I]MIBG for diagnosing DLB has
dopaminergic neuronal loss, particularly in early been reported. The reduced uptake of [123I]MIBG
PD. 18F-DOPA PET has the same clinical indica- related to the degeneration of the adrenergic post-
tions of 123I-FP-CIT SPECT. ganglionic pathways is considered a hallmark of PD
• Postsynaptic dopaminergic imaging (D2 receptor and Lewy body disease compared to MSA.
imaging) with benzamide derivative ligands, such as • [18F]FDG PET can characterize parkinsonian disor-
[11C]raclopride and 123I-iodobenzamide (123I-IBZM), ders by revealing characteristic regional patterns
has been used for the differential diagnosis between identifying specific and highly stable metabolic pat-
Parkinson-plus syndromes and PD. the reduced D2 terns for different Parkinson-plus syndromes.
receptor density due to neuronal cell death within
-38 -35 -16 -14 +0 +2
-18 -15 +4 +6 +20 +22
12 4 +24 +26 +40 +42
-22 -24 +44 +46 +60 +62
Fig. 17.11 Statistical parametric mapping analysis showing the sig- FDG uptake in a patient with PSP (brain stem, medial and prefrontal
nificantly hypometabolic areas involved in patients with three different cortex, anterior cingulate, caudate nuclei). Right panel: patient with
atypical parkinsonisms. Left panel: patient with MSA, showing a statis- CBD, showing statistically significant impairment of [18F]FDG uptake
tically significant hypometabolism in both putamen and in the cerebel- in the right frontal and parietal cortex
lum. Middle panel: areas with statistically significant reduction of [18F]
17.2.8 SPECT Imaging of Brain Perfusion 17.2.8.1 Technical Issues

rCBF SPECT studies have peculiar features that require
Brain perfusion imaging with 99mTc-HMPAO and 99mTc- particular attention both in the pre-acquisition phase and in
ECD is useful for evaluating patients with several types of the acquisition phase, as well as during post-acquisition
neurological diseases. Both radiopharmaceuticals are processing.
small, lipophilic, and neutral molecules capable of cross- During the pre- and the acquisition phase, it is necessary
ing the intact BBB; within the brain, they distribute pro- to search for the highest possible resolution, given the com-
portionally to regional cerebral blood flow (rCBF). Once plexity of the brain anatomy and the small size of the various
inside the brain, they are trapped within the brain neurons structures, to limit acquisition times within 30–45 min to
for a time sufficient for the acquisition, and their uptake is obtain maximum patient immobility, and to standardize the
sufficient to perform imaging with standard SPECT patient’s preparation procedures. The optimal instrumenta-
instrumentation. These tracers are characterized by fast tion for a SPECT brain scan is a multi-detector gamma cam-
washout from extra-cerebral tissues (soft tissue, salivary era that, equipped with LEHR or LEUHR or fan-beam
glands, vessels) and by a high ratio gray/white matter collimators, allows imaging with the highest possible spatial
(Fig. 17.11). resolution for single-photon imaging. The gamma camera
Fig. 17.12 Patient positioning on the gamma camera imaging table for a brain perfusion SPECT scan (reproduced with permission from Volterrani
D, Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
should minimize the distance between the detectors and the The recommended activity for administration of 99mTc-
patient’s head (rotation radius), in order to optimize spatial HMPAO and 99mTc-ECD is about 740 MBq in an adult indi-
resolution. It is helpful to use comfortable head restraints for vidual. Although the uptake phase is rapid (<2 min), SPECT
immobilization during the acquisition. The patient is placed acquisition usually starts about 20 min after radiotracer
supine on the imaging table with his/her arms along the body administration, the minimum time needed for sufficient
and with the head inserted in a comfortable head-holder clearance of the radiopharmaceutical from the extracerebral
using, only when necessary, some bands as means of tissues and for improved target-to-background ratios.
restraint. The head is extended so as to have the orbito-meatal SPECT acquisition is performed on a circular 360°
line to coincide with the vertical line (Fig. 17.12). orbit with the smallest possible radius of rotation (ideal
Patient preparation is particularly important: a few hours <16 cm). The energy window is centered on the photoelec-
before the examination, the intake of substances that may tric peak of 99mTc (140 keV ± 10%). The views are acquired
alter perfusion and metabolism (coffee, smoke, alcohol) in “step-and-shoot” mode with angular sampling of about
must be discontinued. Various drugs may also reduce brain 3° (120–128 views). The acquisition time for each projec-
metabolism and perfusion; for instance, benzodiazepines tion (128 × 128, pixel size 2–4 mm) varies depending on
cause global reductions (although more pronounced in the the type of gamma camera range and collimator used (20–
frontal areas), while dilantin may result in cerebellar hypo- 40 s/view). The average acquisition time is about 30 min
perfusion. If sedation is necessary, it should be induced after overall. Data reconstruction is performed using FBP or
administration of the radiopharmaceutical. It is also i mportant iterative algorithms (OSEM) applying a low-pass filter
that the patient does not undergo excessive external sensory (e.g., a Butterworth filter 0.5 cycles/cm, order 6–8). Since
stimuli (visual and acoustic stimuli) to maintain brain activ- rCBF is evaluated both in the cerebral cortex and in deep
ity under “basal” conditions. To this purpose, the patient is structures such as the thalamus and striatum, it is advisable
prepared with a venous access and remains in a silent and to correct the emission data for attenuation, using either
low-light environment, for at least 10 min before and 5 min the transmission CT data (when using a hybrid SPECT/CT
after tracer injection. gamma camera) or the Chang approach.
Fig. 17.13 Brain perfusion

SPECT in epilepsy. Left
panel: transaxial slice of an
inter-ictal scan. Right panel:
the same section acquired in
the same patient during an
epileptic seizure (ictal scan).
Focally increased perfusion
within the left temporal lobe
can be observed in the ictal
study (reproduced with
permission from Volterrani D,
Applicazioni. Milan: Springer
Italy; 2010)
17.2.9 rCBF SPECT in Epilepsy zure, at the onset of which the perfusion tracer should be
promptly administered.
Epilepsy is a chronic CNS disease characterized by seizure Another issue concerns the use of 99mTc-HMPAO versus
attacks with variable presentation, depending on the origin 99m
Tc-ECD. Both these agents are cerebral perfusion tracers,
of the epileptogenic signal. Treatment is primarily pharma- although with some subtle differences; in particular, 99mTc-
cological, often multi-pharmacological, but in some subjects HMPAO seems to be more adequate to identify areas with
(15%) is not enough to prevent the onset of attacks. The increased rCBF, as its uptake/retention in the brain is more
inability to control epileptic seizures (drug-resistant epi- closely correlated with high blood flow than 99mTc-
lepsy) can lead over time to severe impairment of ECD. However, the in vitro instability of 99mTc-HMPAO
SNC. Approximately half of patients with drug-resistant makes it necessary to reconstitute it exactly at the onset of
focal epilepsy can benefit from a surgical approach to remove the crisis, thus increasing the risk of delaying the injection—
the epileptogenic focus. MR imaging is certainly the primary therefore to perform an early post-critical, rather than a criti-
investigation in the anatomical definition of the epilepto- cal, study. On the contrary, in vitro stability of 99mTc-ECD for
genic lesion. PET with [18F]FDG and SPECT with brain per- 6–8 h after its preparation facilitates timely administration of
fusion imaging agents are also increasingly being used in the this tracer.
clinical routine. The brain perfusion investigation is performed in two
Interictal brain perfusion SPECT (i.e., a scan acquired separate occasions, e.g., 1 day the interictal SPECT and on
when the patient does not have an epileptic crisis) has very another day the ictal SPECT study, under video-EEG moni-
low intrinsic sensitivity, frequently demonstrating mildly toring. Acquisitions and elaborations are performed in the
reduced perfusion of the epileptogenic lesion, whereas same way as described above. Comparative analysis of the
ictal brain perfusion SPECT (so defined when injection is two (critical and intercritical) SPECT scans provides infor-
performed within 45 s of the seizure onset) is successfully mation to localize the epileptogenic focus site (Fig. 17.3).
employed in several epileptic care centers to identify the Although the analysis can be simply visual, it is currently
epileptogenic focus, which appears as an area of hyperper- preferable to use softwares such as ISAS (ictal-interictal
fusion (Fig. 17.13). The inherent limitations of performing SPECT analyzed by SPM) and SISCOM (subtraction of ictal
a critical brain perfusion SPECT are related to the logisti- SPECT co-registered to MRI), developed with the Mayo
cal issues, since tracer injection must be performed as Foundation Analyze software (Fig. 17.14).
early as possible after the onset of the seizure—not an easy
task considering that timing of epileptic seizures is in gen-
eral highly unpredictable. The earlier the perfusion tracer Key Learning Points
is injected, the greater the sensitivity of the procedure. • Brain perfusion imaging with 99mTc-HMPAO and
Another consequence of the logistical difficulties associ- 99m
Tc-ECD has long been used and is still used for
ated with critical SPECT is the high cost associated with evaluating patients with several types of neurologi-
the availability of medical, nursing, and technical person- cal diseases.
nel dedicated to patient surveillance pending epileptic sei-
logically increases in individuals with intact vascular sys-

tem, while it remains unchanged or decreases in patients
with vascular disease and impaired vascular reserve. Brain
perfusion SPECT with acetazolamide test is a clinically
accessible and effective method used mainly for the assess-
ment of cerebral perfusion reserve in patients with carotid
occlusions (Fig. 17.15). It helps to define the risk of stroke or
TIA and to select patients who should undergo revasculariza-
tion (endoarterectomy or bypass).
The test consists of two SPECT scans (usually on sepa-
rate days), one in baseline conditions and one under acet-
azolamide stimulation. Similarly as described above for
comparative evaluation of the paired ictal and interictal
SPECT scans, the study acquired during acetazolamide infu-
sion is compared to the baseline scan either visually or, pref-
erably, using SPM softwares.
Key Learning Points

• Acetazolamide is a carbonic anhydrase inhibitor
determining a transient increase in CO2 that exerts a
high vasodilating power and increases cerebral
blood flow.
• Brain perfusion SPECT with acetazolamide test is a
method used mainly for the assessment of cerebral
Fig. 17.14 SISCOM analysis: area of increased perfusion in the right
prefrontal cortex, obtained after subtracting the intercritical SPECT perfusion reserve in patients with carotid
scan from the critical scan and co-registering the resulting image to the occlusions.
patient’s MRI (reproduced with permission from Volterrani D, Erba PA,
Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
Applicazioni. Milan: Springer Italy; 2010)
17.2.11 Brain Death Scintigraphy
• Tc-HMPAO and 99mTc-ECD can be used in criti-

99m The clinical diagnosis of brain death is defined as the irre-
cal SPECT thanks to their rapid uptake phase. versible loss of all functions of the brain, including the brain
• The analysis of critical and intercritical rCBF stem. The three essential components in patients with brain
SPECT can help in the anatomical definition of the death are (1) coma (lack of all evidences of responsiveness),
epileptogenic lesion in patients with drug-resistant (2) absence of brain stem reflexes (pupillary response to a
focal epilepsy. bright light, ocular movements using oculocephalic testing
and oculovestibular reflex testing, corneal reflex, etc.), and
(3) apnea (absence of a breathing drive).
No other tests are required if the full clinical examina-
17.2.10 rCBF SPECT with Acetazolamide Test tion, including either of two assessments of brain stem
reflexes and a single apnea test, is performed with unequiv-
Although recommended by guidelines of the Society of ocal results. Nevertheless, in some conditions such as skull
Nuclear Medicine since 1998, the use of acetazolamide or cervical injuries, cardiovascular instability, coma caused
(Diamox®) for diagnostic purposes is not yet approved in by barbiturates, drug intoxication, poisoning, and severe
Europe. hypothermia, it is impossible to complete the clinical
As a carbonic anhydrase inhibitor, acetazolamide deter- assessment with a high degree of reliability and/or cer-
mines in the brain a transient increase in CO2 at a level that tainty. In these circumstances, confirmatory (or ancillary)
exerts a high vasodilating power, thus increasing blood flow. tests verifying brain death are necessary. Moreover, ancil-
The vasodilation reached about 15–20 min after administra- lary tests are used to shorten the observation period—an
tion of the drug (1 g i.v.) allows assessment of the cerebral important factor when organ explanation/donation is being
vascular reserve; in particular, cerebral blood flow physio- considered.
Fig. 17.15 Acetazolamide test in a patient with critical stenosis of the left parietal cortex—thus indicating reduced vascular reserve (repro-
left internal carotid. Left panel: baseline brain perfusion SPECT, show- duced with permission from Volterrani D, Erba PA, Mariani G, Eds.
ing no obvious perfusion defects. Right panel: scan acquired after acet- Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
azolamide infusion, showing relative hypoperfusion of the frontal and Springer Italy; 2010)
In clinical practice, EEG tracing, brain angiography, CT the overlap of intracranial and extracranial activity concen-
angiography, MR angiography, transcranial Doppler ultra- tration sites that affects planar imaging. A blush of activity
sound (TCD), and/or brain death scintigraphy are currently apparently in the region of the nose on the anterior view (the
used as ancillary tests—especially in adults. “hot nose” sign) could conversely represent a rerouted blood
For the scintigraphic assessment of patients with sus- flow to the region of the brain stem or cervical region of the
pected brain death, brain-specific agents such as 99mTc- spinal cord [27]. Thus, whenever feasible SPECT images
HMPAO and 99mTc-ECD have replaced 99mTc-DTPA, since should be acquired in addition to flow and planar images
they allow the evaluation of both brain blood flow and brain [28].
cell viability. Early dynamic flow images should be acquired, The absence of demonstrable radionuclide activity within
because they can help to confirm the lack of brain blood flow the brain is consistent with the diagnosis of brain death. In
when the brain is not visualized on delayed images. Flow fact, in case of brain death, intracranial blood flow is com-
images (anterior or anterior and posterior views when using pletely absent, and delayed planar or SPECT images should
a dual-head gamma camera) are usually recorded with a demonstrate no tracer uptake in the brain. Final report of the
dynamic acquisition at 1 s per frame for at least 60 s, starting examination should specify the tracer used, injected activity,
before or at the time of i.v. injection, to ensure imaging and basic imaging information such as whether flow, planar,
begins before the bolus reaches the carotid arteries and ends or SPECT images were obtained. The extent and severity of
well after the venous phase. Late planar images are obtained brain perfusion deficits should also be described. Both cere-
at 20–30 min postinjection, usually acquiring anterior, right bral hemispheres and the posterior fossa (cerebellum) should
lateral, left lateral, and posterior views. Although SPECT be evaluated for a complete study, and specific mention of
acquisition is often not feasible in unstable patients main- the perfusion findings in the posterior fossa and brain stem
tained on life support equipment, it allows better visualiza- should be made (Fig. 17.16). If there is a small amount of
tion of the posterior fossa and brain stem structures avoiding remaining perfusion, a repeat study might be recommended.
Fig. 17.16 99mTc-ECD SPECT in a patient submitted to brain death diagnosis of brain death. Extracranial activity is present within the scalp
evaluation after head injury. Transaxial images show complete absence and nose (reproduced with permission from Volterrani D, Erba PA,
of activity within both hemispheres, the striatum, thalamus, and mid- Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
brain. However, there is a clear evidence of residual activity within the Applicazioni. Milan: Springer Italy; 2010)
brain stem and cerebellum that cannot fully confirm the final clinical
Key Learning Points uation of both brain blood flow and brain cell
• Brain death scintigraphy is one of the current ancil- viability.
lary tests used to support brain death diagnosis. • The absence of demonstrable radionuclide activity
• 99m
Tc-HMPAO and 99mTc-ECD have replaced the within the brain on SPECT or planar images is con-
prior use of 99mTc-DTPA, since they allow the eval- sistent with the diagnosis of brain death.
17.2.12 CSF Imaging ing the presence of activity within the subarachnoid space
along the spinal cord. In addition to planar imaging, SPECT/
Cerebrospinal fluid (CSF) is produced by the choroid plexus CT imaging can provide useful information on CSF distribu-
of the cerebral ventricles at a rate of 500 mL/day and is pres- tion, by combining functional and anatomic images to be
ent in volumes ranging from 60 to 200 mL in a normal adult used in conjunction with clinical findings [30].
person. After exiting through the fourth ventricle foramina, In a patient without a communicating hydrocephalus,
the CSF flows in the subarachnoid spaces of the superior sag- activity goes up through the spinal canal, floods the basal
ittal and all cerebral convexities, where reabsorption occurs cisterns after about 4 h, and then ascends through the Sylvian
at the arachnoid granulations. However, in pathologic condi- and interhemispheric cisterns to all peri-encephalic sub-
tions causing blockage of normal absorption, the main path- arachnoid spaces of the cerebral convexities. Thus, images
way of reabsorption is through the ependymal lining of the acquired at 6 h demonstrate the presence of activity within
ventricular system. the basal cisterns and an initial activity in the subarachnoid
Following the advent of CT and MR imaging of the brain, spaces. At 24–48 h activity is present within the subarach-
radionuclide imaging is used sparingly for the evaluation of noid spaces over the parasagittal space and the convexities,
CSF flow dynamics. The main clinical indication of CSF with a relative washout from the basal cisterns.
radionuclide imaging, also called cisternography, is a sus- The typical pattern of cisternography in a patient with
pected communicating hydrocephalus (normal pressure communicating hydrocephalus is characterized by the early
hydrocephalus). Although there are different possible causes (6 h) fill-in of activity into the lateral ventricles, which usu-
of communicating hydrocephalus, this condition is most fre- ally persists at 24–48 h, associated with delayed rise along
quently due to reduced absorption through the arachnoid gran- the subarachnoid spaces over the cerebral convexities. The
ulations of the subarachnoid space. There are several basis of this pattern of time-related activity distribution relies
radiographic features on CT and MRI that facilitate its diagno- on the reduced CSF flow toward the subarachnoid spaces
sis; nonetheless, it can sometimes be challenging to distin- occurring in hydrocephalus, which allows the reflux of radio-
guish communicating hydrocephalus from ventricular activity into the ventricles. Thus, the presence of radioactiv-
enlargement, as observed in patients with ex-vacuo dilatation ity in the lateral ventricles, at any time, should be considered
of the ventricles due to parenchymal volume loss (with com- an abnormal finding consistent with communicating hydro-
pensatory enlargement of the CSF spaces). Furthermore, MRI cephalus (Fig. 17.17).
can in some patients be contraindicated. When correctly diag- Since obstruction is one of the main complications
nosed, communicating hydrocephalus usually can be treated observed for the ventriculo-peritoneal shunt device implanted
with CSF shunting, most commonly with a device used to for treating these patients, radionuclide techniques can be
shunt cerebrospinal fluid to the peritoneal cavity (ventriculo- employed to evaluate shunt patency. The radiopharmaceuti-
peritoneal shunt), which provides prompt relief of clinical cal, usually 99mTc-DTPA, can be injected in sterile conditions
symptoms. The device consists of a catheter placed in the ven- into the reservoir or the catheter of the device tunneled under
tricles connected through a valve/reservoir to a second cathe- the skin. In case of shunt patency, serial images can demon-
ter tunneled under the skin and reaching the peritoneal cavity. strate the rapid transit of the radioactivity into the peritoneal
The principle of CSF imaging with cisternography concavity within 30–60 min, visualizing a widespread activity in
sists in the intrathecal administration of a diffusible radio- the abdomen. On the contrary, when an obstruction of the
pharmaceutical that remains in the CSF spaces until it is distal end of the catheter occurs, activity focally concentrates
definitely absorbed. The most commonly employed radio- at the tip of the catheter without significant visualization of
pharmaceutical is 111In-DTPA, which offers a sufficient long activity in the peritoneal cavity. In order to investigate a
half-life for its use in the functional study of CSF dynamics. proximal obstruction, a manual occlusion of the distal part of
However, 99mTc-DTPA has been used as an alternate radio- the device must be performed. In case of proximal obstruc-
pharmaceutical, based on intrathecal administration of an tion, the activity, which should diffuse back into the ventri-
activity of 18.5–37 MBq (0.5–1 mCi) for CSF flow cister- cles, fails to reach the ventricles, or it is not cleared from the
nography [29]. ventricles even after several hours.
The first step of the procedure is a lumbar puncture with a CSF imaging can be useful also in the assessment of
22-gauge needle. After withdrawing 1 mL of CSF, the same patients with suspected CSF leakage, such as rhinorrhea and/
volume of 111In-DTPA (18.5 MBq) is administered into the or otorrhea. Planar images of the skull can prove the pres-
subarachnoid space. ence and can localize a CSF leak, which is most frequently
Planar images of the skull (anterior, posterior, and lateral due to fistulae in the cribriform plate and ethmoid sinuses.
views) are acquired with a gamma camera at 6, 24, and 48 h Moreover, CSF rhinorrhea can be quantified by measuring in
after the intrathecal administration of the radiopharmaceuti- a well counter the activity absorbed by two pledgets placed
cal. An early image at 1–2 h over the lumbar region can also for 4–6 h in each nostril after the intrathecal administration
be acquired in order to verify the success of injection, reveal- of the radiopharmaceutical.
Fig. 17.17 111In-DTPA

cisternography in a patient
with communicating
hydrocephalus. Images at 6 h 6h
show activity mainly in the
basilar cisterns and lateral
ventricles. At 24 and 48 h, a
persistent activity in lateral
ventricles with minimal
activity over the Sylvian
space (anterior view) and
parasagittal space (lateral
views) can be seen 24 h
48 h
Anterior Posterior L Lateral R lateral
Finally, different forms of long-term intrathecal delivery References

of drugs are being employed for treating a variety of either
oncologic or non-oncologic conditions [31–36]. This thera- 1. Salloway S. Clinical and pathological examples of Alzheimer’s
disease, dementia with Lewy bodies, and frontotemporal dementia.
peutic approach is based on implantation of a subcutaneous Med Health R I. 2012;95:207–9.
reservoir, usually an Onmaya-type device connected to the 2. Gauthier S, Reisberg B, Zaudig M, Petersen RC, Ritchie K, Broich
subarachnoid space. It is frequently required to ascertain K, et al. Mild cognitive impairment. Lancet. 2006;367:1262–70.
patency of the system immediately after implantation and 3. Laforce R Jr, Rabinovici GD. Amyloid imaging in the differential
diagnosis of dementia: review and potential clinical applications.
occasionally in case of malfunction later on during chronic Alzheimers Res Ther. 2011;3:31. https://doi.org/10.1186/alzrt93.
treatment. This can be achieved by injecting the radiophar- 4. Silverman DH, Small GW, Phelps ME. Clinical value of neuro-
maceutical directly into the reservoir or through the lumbar imaging in the diagnosis of dementia. Sensitivity and specific-
subarachnoid space—depending on the clinical indication. ity of regional cerebral metabolic and other parameters for early
identification of Alzheimer’s disease. Clin Positron Imaging.
1999;2:119–30.
5. Foster NL, Heidebrink JL, Clark CM, Jagust WJ, Arnold SE,
Barbas NR, et al. FDG-PET improves accuracy in distinguish-
Key Learning Points ing frontotemporal dementia and Alzheimer’s disease. Brain.
• Cisternography consists in the intrathecal adminis- 2007;130:2616–35.
tration of a diffusible radiopharmaceutical that 6. Ishii K, Soma T, Kono AK, Sofue K, Miyamoto N, Yoshikawa T,
et al. Comparison of regional brain volume and glucose metabolism
remains in the CSF spaces until it is definitely
between patients with mild dementia with Lewy bodies and those
absorbed. with mild Alzheimer’s disease. J Nucl Med. 2007;48:704–11.
• 111In-DTPA offers a sufficient long half-life for its 7. Rowe CC, Villemagne VL. Brain amyloid imaging. J Nucl Med
use in the functional study of CSF dynamics. A Technol. 2013;41:11–8.
8. Johnson KA, Minoshima S, Bohnen NI, Donohoe KJ, Foster NL,
lumbar puncture is needed to administer 111In-
Herscovitch P, et al. Appropriate use criteria for amyloid PET: a
DTPA into the subarachnoid space. report of the Amyloid Imaging Task Force, the Society of Nuclear
• The main clinical indication of cisternography is a Medicine and Molecular Imaging, and the Alzheimer’s Association.
suspected communicating hydrocephalus. Alzheimers Dement. 2013;9:e106–9.
9. Rabinovici GD, Rosen HJ, Alkalay A, Kornak J, Furst AJ, Agarwal
• The typical pattern of cisternography in a patient
N, et al. Amyloid vs FDG-PET in the differential diagnosis of AD
with communicating hydrocephalus is character- and FTLD. Neurology. 2011;77:2034–42.
ized by the early (6 h) fill-in of activity into the lat- 10. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s dis-
eral ventricles, which usually persists at 24–48 h. ease: progress and problems on the road to therapeutics. Science.
2002;297:353–6.
11. Jack CR Jr, Knopman DS, Jagust WJ, Shaw LM, Aisen PS, Weiner 24. Van Laere K, Clerinx K, D’Hondt E, de GT, Vandenberghe

MW, et al. Hypothetical model of dynamic biomarkers of the W. Combined striatal binding and cerebral influx analysis of
Alzheimer’s pathological cascade. Lancet Neurol. 2010;9:119–28. dynamic 11C-raclopride PET improves early differentiation
12. Pontecorvo MJ, Mintun MA. PET amyloid imaging as a tool for between multiple-system atrophy and Parkinson disease. J Nucl
early diagnosis and identifying patients at risk for progression to Med. 2010;51:588–95.
Alzheimer’s disease. Alzheimers Res Ther. 2012;3:11. https://doi. 25. Orimo S, Amino T, ItohY, Takahashi A, Kojo T, Uchihara T, et al. Cardiac
org/10.1186/alzrt70. sympathetic denervation precedes neuronal loss in the sympathetic gan-
13. Förster S, Yousefi BH, Wester HJ, Klupp E, Rominger A, Forstl glia in Lewy body disease. Acta Neuropathol. 2005;109:583–8.
H, et al. Quantitative longitudinal interrelationships between brain 26. Courbon F, Brefel-Courbon C, Thalamas C, Alibelli MJ, Berry I,
metabolism and amyloid deposition during a 2-year follow-up Montastruc JL, et al. Cardiac MIBG scintigraphy is a sensitive tool
in patients with early Alzheimer’s disease. Eur J Nucl Med Mol for detecting cardiac sympathetic denervation in Parkinson’s dis-
Imaging. 2012;39:1927–36. ease. Mov Disord. 2003;18:890–7.
14. Mosconi L. Glucose metabolism in normal aging and Alzheimer’s 27. Appelt EA, Song WS, Phillips WT, Metter DF, Salman UA, et al.
disease: methodological and physiological considerations for PET The “hot nose” sign on brain death scintigraphy: where does the
studies. Clin Transl Imaging. 2013;1:217–33. flow really go? Clin Nucl Med. 2008;33:55–7.
15. Shimada H, Shinotoh H, Hirano S, Miyoshi M, Sato K, Tanaka N, 28. Donohoe KJ, Agrawal G, Frey KA, Gerbaudo VH, Mariani G, et al.
et al. Beta-amyloid in Lewy body disease is related to Alzheimer’s SNM practice guideline for brain death scintigraphy 2.0. J Nucl
disease-like atrophy. Mov Disord. 2013;28:169–75. Med Technol. 2012;40:198–203.
16. Fujishiro H, Iseki E, Higashi S, Kasanuki K, Murayama N, Togo 29. Khan SH, Rather TA. Detection of spontaneous CSF rhinorrhea in
T, et al. Distribution of cerebral amyloid deposition and its rele- an obese diabetic woman by Tc-99m DTPA cisternography: a case
vance to clinical phenotype in Lewy body dementia. Neurosci Lett. report. World J Nucl Med. 2009;8:182–3.
2010;486:19–23. 30. Thut DP, Kreychman A, Obando JA. 111In-DTPA cisternography
17. Catafau AM. Brain SPECT in clinical practice. Part I: Perfusion. J with SPECT/CT for the evaluation of normal pressure hydrocepha-
Nucl Med. 2001;42:259–71. lus. J Nucl Med Technol. 2014;42:70–4.
18. Herholz K, Schopphoff H, Schmidt M, Mielke R, Eschner W, 31. Adler JA, Lotz NM. Intrathecal pain management: a team-based
Scheidhauer K, et al. Direct comparison of spatially normal- approach. J Pain Res. 2017;10:2565–75.
ized PET and SPECT scans in Alzheimer’s disease. J Nucl Med. 32. Cohen-Pfeffer JL, Gururangan S, Lester T, Lim DA, Shaywitz AJ,
2002;43:21–6. Westphal M, et al. Intracerebroventricular delivery as a safe, long-
19. Asenbaum S, Pirker W, Angelberger P, Bencsits G, Pruckmayer term route of drug administration. Pediatr Neurol. 2017;67:23–35.
M, Brücke T. [123I]beta-CIT and SPECT in essential tremor and 33. Hamad M, Holland R, Kamal N, Luceri R, Mammis A. Potential
Parkinson’s disease. J Neural Transm. 1998;105:1213–28. for intrathecal Baclofen in treatment of essential tremor. World
20. Benamer TS, Patterson J, Grosset DG, Booij J, de Bruin K, van Neurosurg. 2017;105:170–5.
Royen E, et al. Accurate differentiation of parkinsonism and essen- 34. Lee YC, Hsieh CC, Chuang JP, Li CY. The necessity of intrathecal
tial tremor using visual assessment of [123I]-FP-CIT SPECT imag- chemotherapy for the treatment of breast cancer patients with lep-
ing: the [123I]-FP-CIT study group. Mov Disord. 2000;15:503–10. tomeningeal metastasis: a systematic review and pooled analysis.
21. Brooks DJ. Imaging approaches to Parkinson disease. J Nucl Med. Curr Probl Cancer. 2017;41:355–70.
2010;51:596–609. 35. Olmos-Jiménez R, Espuny-Miró A, Cárceles Rodríguez C, Díaz-
22. Jokinen P, Helenius H, Rauhala E, Bruck A, Eskola O, Rinne Carrasco MS. Practical aspects of the use of intrathecal chemother-
JO. Simple ratio analysis of 18F-fluorodopa uptake in striatal subre- apy. Farm Hosp. 2017;41:105–29.
gions separates patients with early Parkinson disease from healthy 36. Woolf SM, Baum CR. Baclofen pumps: uses and complications.
controls. J Nucl Med. 2009;50:893–9. Pediatr Emerg Care. 2017;33:271–5.
23. Ishikawa T, Dhawan V, Chaly T, Margouleff C, Robeson W, Dahl
JR, et al. Clinical significance of striatal DOPA decarboxylase
activity in Parkinson’s disease. J Nucl Med. 1996;37:216–22.
Hybrid Imaging for Tumors of the Brain
18
Giampiero Giovacchini, Mattia Riondato, Patrizia Lazzeri,
Elisa Borsò, Valerio Duce, Rossella Leoncini,
Elisabetta Giovannini, and Andrea Ciarmiello
Contents
18.2 Radiotracers 414
18.2.1 2-Deoxy-2-[18F]Fluoro-D-Glucose ([18F]FDG) 415
18.2.2 DNA Synthesis 421
18.2.3 Hypoxia Agents 421
18.3 Indications to PET/CT or PET/MR 422
18.4 PET/MR 426
References 427
Learning Objectives 18.1 Introductory Background

• Understand epidemiology, classification, and clinical pre-
sentation of brain tumors. There are many histotypes of brain tumors, the most frequent
• Understand the basic pathophysiology of brain tumors as being gliomas, i.e., tumors deriving from glia cells. Tumors
it relates to nuclear medicine imaging. deriving from neurons, typically the medulloblastoma, are
• Learn advantages and disadvantages of PET/CT and PET/ much less frequent. Most of the information presented in this
MR for imaging brain tumors. chapter has been obtained from studies in glioma patients,
• Learn the properties of radiotracers used for brain imag- but the principles of brain tumor imaging apply similarly to
ing and the rationale to their use in different clinical set- medulloblastoma. The term glioma will be used in this
tings (diagnostic approach, prognosis, evaluation of chapter.
response to treatment, differential between tumor recur- In addition to gliomas, malignancies of the central ner-
rence and scar, detection of recurrence). vous system (CNS) include meningiomas, primary brain
• Learn most common protocols for imaging brain tumors, lymphomas, and metastases from tumors outside the brain.
with special emphasis on measuring the main metabolic The latter two types of tumors (lymphomas and brain metas-
parameters of brain tumor metabolism. tases from primary non-CNS tumors) will be dealt with in
• Understand the principles of image interpretation for other chapters of this book. Meningiomas are mostly asymp-
nuclear medicine imaging of brain tumors. tomatic and easily diagnosed through computed tomogra-
• Learn the use of PET for the definition of volume delinea- phy. Even though many PET radiotracers can be used to
tion in the planning of radiotherapy and for the assess- support the diagnosis of meningioma, nowadays the diagno-
ment of the response to therapy. sis of meningiomas is mainly radiological, and PET radio-
tracers are almost exclusively used to investigate gliomas
G. Giovacchini (*) · M. Riondato · P. Lazzeri · E. Borsò · V. Duce rather than meningiomas. Therefore, this chapter will be
R. Leoncini · E. Giovannini · A. Ciarmiello devoted only to the discussion of gliomas. For more informa-
Nuclear Medicine Department, “S. Andrea” Hospital, tion on meningiomas, we refer the reader to the Further
La Spezia, Italy
Reading.
e-mail: giampiero.giovacchini@asl5.liguria.it

414 G. Giovacchini et al.
Histological features divide gliomas into low-grade and Generally, symptoms are more frequent in high-grade tumors
high-grade gliomas. Grading is based on several cellular fea- that are fast growing. In many cases, gliomas are an inciden-
tures, including the degree of mitosis, nuclear atypia, necro- tal finding at CT or MR.
sis, and microvascular proliferation. Histological features of Therapy of brain tumor is based on surgery and radio-
gliomas may vary substantially among different parts of the therapy [4]. Surgery may be performed with curative or
tumor; therefore, the most malignant area determines the debulking intent. Radiation therapy can prolong survival
final grade of the tumor. Grade I and grade II gliomas are and, together with gross surgical excision, it represents the
defined as low-grade gliomas, while grade III and IV glio- standard of care of brain tumors [5]. The overall efficacy of
mas are defined as high-grade gliomas. Higher grade is chemotherapy in gliomas is limited because the BBB limits
significantly related with a poorer prognosis. Other factors the access to the brain to molecules with low molecular
that may affect prognosis include older age, extent of tumor weight and high lipophilicity and the multidrug resistance
resection, anatomic location, and functional neurologic sta- system limits the efficacy of a wide range of drugs.
tus (traditionally quantified through the Karnofsky Temozolomide is now the most effective chemotherapy
Performance Scale). Tumor grade is identified through sev- drug, and it is approved for refractory anaplastic astrocy-
eral histological samplings of glioma tissue performed toma or for GBM [6]. Antiangiogenic drugs were devel-
through an invasive surgical procedure. In patients treated oped to target vascular endothelial growth factor (VEGF)
with surgery or radiotherapy prior to obtaining tissue for his- and VEGF receptor. Emerging literature suggests efficacy
topathology, accurate identification of tumor grade is more of antiangiogenic therapy, particularly bevacizumab, for
difficult because blood-brain barrier (BBB) damage and recurrent high-grade gliomas [7, 8]. Locoregional radioim-
tumor necrosis can occur as a consequence of therapy and munotherapy using monoclonal antibodies conjugated with
tumor sampling may be biased [1]. high-energy β-emitters (90Y or 177Lu) has been used with
Gliomas are more frequent in young adults, and they are some evidence of clinical benefit, but it is no longer used
slightly more common in men. In the last few years, nowadays.
O6-methylguanine-DNA methyltransferase promoter meth-
ylation and mutations in the cytosolic isocitrate dehydroge-
nase 1 or 2 genes have been identified as genetic factors with Key Learning Points
a favorable prognosis in astrocytic high-grade gliomas, • Gliomas are classified as high-grade and low-grade
whereas loss of heterozygosis on chromosomes 1p and 19q gliomas.
is highly correlated with low-grade gliomas [2]. However, • Therapy of gliomas consists in surgery combined to
other risk factors are still largely unknown. radiotherapy.
All gliomas have the tendency to recur locally. When they
recur, low-grade gliomas often display the histological and
imaging characteristics of high-grade gliomas, as a result of
genetic and phenotypic transformation. Grade III astrocyto- 18.2 Radiotracers
mas are commonly referred to as anaplastic astrocytomas.
Glioblastoma, a grade IV glioma, is the most common and the The radiotracers that can be used for imaging of gliomas
most malignant form of gliomas, accounting for about 50% of and factors affecting their uptake in the brain are listed in
brain tumors. Glioblastomas are exceptionally aggressive Tables 18.1 and 18.2, respectively.
tumors that are generally resistant to treatment. Average sur-
vival is few months to a general maximum of 1–2 years from
Table 18.1 Most common radiotracers currently available for brain
diagnosis. Glioblastomas are characterized by the abundance tumor imaging
of hypercellular anaplastic glioma cells with marked mitotic
Biological process Radiotracer
activity as well as necrosis and endothelial proliferation. Glucose transport across 2-Deoxy-2-[18F]fluoro-d-glucose ([18F]
Neoplastic cells frequently reveal pleomorphism showing dif- BBB and metabolism FDG)
ferent histologic features such as small homogeneous cells Amino acid transport and [11C]methionine
with scant cytoplasm, fibrillary-shaped cells, multinucleated protein synthesis 18
F-fluoroethyltyrosine (18F-FET)
giant cells, and cells with pleomorphic nuclei and cytoplasm Amino acid transport and 18
F-fluoro-L-3,4-
dopamine metabolism dihydroxyphenylalanine (18F-DOPA)
[3]. Because of such cellular pleomorphism, the tumor is often
DNA replication 18
F-fluorothymidine (18F-FLT)
referred to as glioblastoma multiforme (GBM) [3].
Lipid metabolism [11C]choline/18F-fluorocholine
Brain tumors remain asymptomatic for many years espe- Hypoxia 18
F-fluoromisonidazole (18F-MISO)
cially if they are located in mute cerebral areas, such as the 18
F-azomycin arabinoside (18F-FAZA)
frontal lobe. The most common symptom is headache. 64
Cu-methylthiosemicarbazone
Nausea, vomiting, and seizures occur only in the late phase. (64Cu-ATSM)
18 Hybrid Imaging for Tumors of the Brain 415
Table 18.2 Factors affecting uptake of radiopharmaceuticals in Table 18.3 Clinical indications to PET/CT and PET/MR imaging in
gliomas gliomas
1. Perfusion and blood-brain barrier integrity Before therapy
2. Histological grade 1. Confirm the diagnostic suspect
3. Glucose metabolic rate 2. In vivo tumor grading
4. DNA proliferation rate 3. Guide to stereotactic biopsy to
5. Protein synthesis rate 4. Presurgical planning
6. Membrane (phospholipid) proliferation rate 5. Identification of metabolically active tumor (or biological
7. Expression of membrane transporters target volume) for radiotherapy planning
8. Oxygen tissue concentration (hypoxia) 6. Prediction of survival
9. Dimension of the lesion (partial volume effect) After therapy
10. Necrotic areas 1. Identification of residual tumor after surgery
11. Radiotherapy treatment 2. Differentiate between tumor recurrence and radiation necrosis
3. Evaluation of response to therapy
4. Prediction of survival
18.2.1 2-Deoxy-2-[18F]Fluoro-d-Glucose
([18F]FDG) PET/CT other than pregnancy. Acquisition is performed in
3-D mode, which allows enhanced signal-to-noise ratio and
[18F]FDG allows measurement of glucose metabolism. [18F] better sensitivity, for an acquisition time of about 15 min.
FDG is the most commonly used tracer for brain tumor Between 125 and 250 MBq, [18F]FDG is injected according
imaging. The tracer was described in the 1970’s to study to body weight. The patient must rest in a quiet dimly lit
brain metabolism. In the 1980’s, the tracer was used for brain room with eyes closed in a condition of minimal audiosen-
tumor imaging, since brain tumor growth implies an increase sory stimulation starting 10–15 min before injection and dur-
in glucose consumption. ing the uptake phase. During acquisition counts are corrected
The 2-deoxyglucose model for quantifying the cerebral for attenuation, scatter, and random coincidences. Subsequent
metabolic rate for glucose (CMRglc) was defined by Sokoloff image reconstruction is performed using iterative reconstruc-
following autoradiographic experiments in rats and subse- tions. Scans are acquired using a 512 × 512 matrix. Filtering
quently adapted for human studies with PET by different is used to increase the signal-to-noise ratio during image pro-
authors. [18F]FDG enters the cell through a carrier-mediated cessing. Image reconstruction is very similar also for tracers
system, and it is phosphorylated within the cytoplasm by the that will be subsequently described.
enzyme hexokinase. Phosphorylated [18F]FDG accumulates Glucose metabolism is 2–3 times higher in gray matter
into the cytoplasm as a function of neural activity, it has no than in white matter. In conditions of sensory deprivation
other metabolic side pathways, and therefore [18F]FDG accu- and in normal subjects, glucose metabolism is uniform
mulation rate can be easily modeled and quantified. Initially, across gray matter.
dynamic scanning and arterial line-derived input function Images are interpreted by looking for areas of activity
were adopted for clinical research with the aim of obtaining higher than gray matter (typical of grade IV gliomas) and
absolute quantitative (mg/gram/min) values of CMRglc. This areas of activity lower than white matter (typical of grade I
method was based on the estimation of the four rate con- gliomas or of benign tumors). Values of tumor glucose
stants of a two-tissue compartment model (K1, k2, k3, k4). metabolism between these two extremes are frequent, and in
Adaptions of the PET method based on the use of population- these cases the high-physiological brain glucose metabolism
based (i.e., predefined) rate constants allowed calculation of makes the image interpretation more difficult. Asymmetry of
absolute values of CMRglc without the need of dynamic scan- glucose metabolism is observed on all transverse slices.
ning (however without obviating the need for an arterial Coronal and sagittal slices are also reviewed. In PET/CT, CT
line). Routine clinical studies are often performed using only is used for anatomic localization, and an MR, generally
semiquantitative (maximum or mean standardized uptake already performed, is reviewed for further anatomical refine-
values, SUVmax and SUVmean, and other derived indices) and ments. If PET/MR is available, PET, MR, and fused PET/
visual analysis, obtained with a single static acquisition MR images are analyzed. The indications to brain PET/CT
45–60 min after injection (uptake period). Fasting for at least or PET/MR are listed in Table 18.3.
6 h before tracer injection is needed to avoid competition
between dietary carbohydrates and [18F]FDG. Preparation 18.2.1.1 Amino Acid Tracers
may be occasionally demanding in diabetic patients experi- The most frequently used amino acid tracers are [11C]methi-
encing hyperglycemia. Motion artifacts should be minimized onine, O-(2-18F-fluoroethyl)-l-tyrosine (18F-FET), and
especially in patients with less compliance by using a head 3,4-dihydroxy-6-18F-fluorophenylalanine (18F-fluorodopa
holder. There are generally no major contraindications to ( F-DOPA)).
18
Neutral l-amino acids are essential amino acids; they can- play a much less important role in the inflammatory response
not be synthesized in the brain, and, consequently, they must and therefore have higher specificity for distinguishing
be supplied either by the diet or by breakdown of endogenous tumor recurrence from radiation necrosis.
proteins. Since a single type of transport carrier mediates the [11C]methionine is the prototype of amino acid tracers. It
endothelial transport of all amino acids, radiolabeled amino is the first amino acid that was consistently shown to be use-
acids compete with dietary amino acids for cerebral uptake; ful for brain tumor imaging, especially for differentiating
therefore fasting is needed before a PET scan. tumor recurrence from radiation necrosis. An activity of
Amino acids are the building blocks of proteins. Gliomas about 6 MBq/kg body weight is administered. Brain uptake
show increased rate of protein synthesis, making measure- of the tracer is quick, and images are acquired between 20
ment of amino acid uptake an attractive approach from the and 30 min postinjection (p.i.) (Figs. 18.1 and 18.2).
point of view of imaging. This measurement, however, has For 18F-FET (Fig. 18.3), an activity of about 200 MBq is
some methodological restrictions. Intracellular metabolism administered. Uptake increases and reaches a plateau
is different across the various amino acids. In addition to pro- 20 min p.i.; thereafter, the uptake values remain stable for
tein synthesis, either amino acids can be metabolized by about 2 h [13].
transamination (rapidly followed by decarboxylation for There is growing interest in the clinical applications of
CO2 production), or they can enter other neural metabolic 18
F-DOPA, which has an intracellular important pathway, i.e.,
pathways. For example, [11C]methionine has a 15% incorpo- transformation to dopamine that makes the overall uptake of
ration rate, while 18F-FET is not significantly incorporated the tracer less sensitive to the transport across the BBB. Uptake
into proteins and uptake is thought to reflect primarily trans- of the tracer is low in all brain regions, with the notable excep-
port across the BBB [9]. Regardless of the specific metabolic tion of the striatum. Contrary to studies performed for the
pathway, amino acid tracers were all shown to be useful for evaluation of movement disorders, where a dynamic scan or a
imaging brain tumors [10]. late scan acquisition at about 70 min is performed, for tumor
At variance with [18F]FDG, labeled amino acid uptake in imaging, an early scan at 20 min postinjection is preferred. At
normal cortex is very low (with no distinction between gray that time physiological uptake in the striatum is still limited,
and white matter), and their uptake is not significantly influ- and the vascular phase is definitely over [14]. At subsequent
enced by visual and auditory stimulation. Therefore, dim times postinjection, the specific uptake in the striatum
light is not necessary before or after tracer injection. increases, and it could mask the visualization of tumor near
Moreover, in comparison to [18F]FDG, amino acid tracers the basal ganglia or other tumors that have low 18F-DOPA
a b
5.5
0.0
Fig. 18.1 A 32-year-old woman with anaplastic oligodendroglioma: MR image; (d) MR FLAIR image shows a remarkable increase in the
(a) [11C]methionine PET image shows slightly elevated [11C]methio- FLAIR signal. The tumor was partially resected. The patient was alive
nine uptake; (b) T1-weighted MR image with gadolinium contrast at the time of this report 47.4 months after the operation (reprinted with
medium shows no enhancement; (c) fused [11C]methionine PET and permission from [10])
c d
Fig. 18.2 Combined [18F]FDG (upper row) a

and [11C]methionine (lower row) PET
studies. Patient #1 displays tumor
recurrence as the primary glioblastoma
grade IV, which is easily detectable with
[11C]MET (b) but not on [18F]FDG (a).
Patient #2 is affected by a primary
oligoastrocytoma grade II, which is hot on
[11C]methionine PET (d) and cold on [18F]
FDG PET (c) (reprinted with permission b
from [11])
d
a b
SUV
2.7
0.7
Fig. 18.3 18F-FET PET and early postoperative T1-weighted gross total resected. (b) This patient was evaluated as partially resected
MR. Early (48–72 h) postoperative post-contrast T1-weighted MR based on a smaller contrast-enhancing region (blue arrow) in the depth
alone and fused to 18F-FET PET scanning performed 2 weeks later for of the resection cavity in the left temporal region. Lack of metabolic
radiation treatment planning showing the superiority of PET with 18F- activity on 18F-FET PET, however, suggests reactive changes. Areas of
FET over MR for definition of residual tumor volume after surgery. (a) increased 18F-FET uptake are found in subcortical white matter anterior
Residual metabolically active non-enhancing tumor remnant anterior to and posterior to the cavity (red arrow), presumably infiltrating gliobas-
the resection cavity identified in the left frontal lobe (red arrow), pre- toma (reprinted with permission from [12])
sumably infiltrating glioblastoma. On MR, the tumor was evaluated as
uptake. Acquisition of early images at this time point allows ence. Static images are obtained at different times postinjec-
distinction of tracer activity in tumors that involve the stria- tion, depending upon the local center protocol and the tracer
tum from physiological activity in the striatum [15] (Figs. 18.4 used, with a 5- to 10-min-long acquisition.
and 18.5).
Regarding acquisition protocols, as previously alluded to, 18.2.1.2 Radiolabeled Choline
PET scans are mostly performed with a single static acquisi- Radiolabeled choline (either [11C]choline or 18F-fluorocholine)
tion some time after tracer injection with minor differences is a phospholipid precursor. The uptake of the tracer is
among centers. In selected instances, a dynamic acquisition thought to reflect membrane proliferation, particularly
is indicated. This is the case when one wishes to (1) generate choline-derived membrane phospholipids. In vivo studies
parametric images and calculate quantitative values of with proton magnetic resonance spectroscopy have shown
parameters of physiological interest, as described above for increased concentration of choline-containing phospholipids
CMRglc and [18F]FDG, and (2) estimate the value of the uni- (i.e., phosphorylcholine and glycerophosphocholine) in
directional (blood/brain) incorporation coefficient (e.g., brain tumors. In physiological conditions the tracer has no
KFLT). For 18F-FET PET, a dynamic acquisition was shown to significant uptake in the brain, and therefore it is suitable for
provide additional advantages in comparison to the standard imaging gliomas with high sensitivity. Either [11C]choline or
static acquisition. The shape of the 18F-FET time-activity 18
F-fluorocholine is available to most PET centers since the
curve may be helpful to better identify glioma grading, to tracer is successfully used for imaging prostate cancer. This
differentiate tumor recurrence from radiation-induced availability facilitated the application of radiolabeled choline
changes, and to predict patients’ prognosis [17]. for detection of brain tumors, yielding promising results.
Image interpretation is much easier than for [18F]FDG, False-positive findings may occur owing to tracer uptake by
since there is no physiological uptake of amino acid tracers inflammatory lesions, in brain metastases, and in
and all spots of tracer uptake are related to gliomas. As dis- meningiomas.
cussed above, an important exception is represented by Images are acquired using a static scan 2 min after injec-
18
F-FDOPA that accumulates in the striatum at late times. An tion. Image interpretation is similar to what is described for
early acquisition 20 min after injection avoids this interfer- amino acid tracers.
a b c
d e f
g h i
Fig. 18.4 Positive 18F-FDOPA PET/CT and 18F-FDOPA PET/MR tral and dorsal striatum. The 18F-FDOPA PET image (e) and fused PET/
images. (a–c) Glioblastoma multiforme. The axial FLAIR image (a) MR image (f) show increased uptake and a modified outline of the left
shows a lesion in the left insula and temporal lobe involving the ipsilat- striatum. (g–i) Glioblastoma multiforme. The axial FLAIR image (g)
eral ventral striatum (arrow). The 18F-FDOPA PET image (b) and fused shows an extensive infiltrative lesion involving the left temporal-
PET/MR image (c) clearly show asymmetrical uptake in the ventral parietal lobe, the ipsilateral thalamus, and the dorsal striatum (puta-
striatum, greater on the left, matching the lesion on MRI. (d–f) men). The 18F-FDOPA PET image (h) and fused PET/MR image (i)
Glioblastoma multiforme. The axial FLAIR image (d) shows an exten- show increased uptake and a modified outline of the left putamen,
sive infiltrative lesion involving the left frontal and temporal lobe matching the lesion on MRI (reprinted with permission from [15])
extending to the genu of the corpus callosum and to the ipsilateral ven-
a b
c d
Fig. 18.5 Combined 18F-DOPA (upper row) and [18F]FDG (lower row) lesion (arrows) suggestive of recurrence. Transaxial [18F]FDG PET (c)
PET studies. A 27-year-old man with right frontal grade II oligoastro- and PET/CT (d) images show no abnormal focus of tracer uptake and
cytoma treated primarily with surgery and radiotherapy, presented with are negative for recurrence. The patient underwent reoperation and was
clinical suspicion of recurrence. Transaxial 18F-DOPA PET (a) and found to have recurrent grade III glioma (anaplastic astrocytoma) on
PET/CT (b) images show tracer accumulation in a right frontal lobe histopathology (reprinted with permission from [16])
18.2.2 DNA Synthesis conditions, uptake of the tracer in the brain is negligible, so
that images resemble amino acid images (Fig. 18.6).
DNA synthesis is a fundamental step for cell proliferation.
Thymidine is a pyrimidine used for DNA but not for RNA
synthesis. Thymidine is incorporated into DNA through the 18.2.3 Hypoxia Agents
salvage pathway for pyrimidines. The amount of DNA syn-
thetized is lower than through the de novo pathway; however, The interest in imaging tumor hypoxia is due to the fact that
because of the limited number of compartments involved and hypoxia is associated with resistance to radiotherapy and to some
modeling assumptions, DNA synthesis can be more easily chemotherapy regimens and therefore with tumor progression.
quantified. Tracers designed to quantify DNA replication are Several radiopharmaceuticals have been evaluated as hypoxia
also named proliferation tracers. tracers. These tracers include nitroimidazole compounds, e.g.,
Initial studies with PET and [11C]thymidine showed 18
F-fluoromisonidazole (18F-MISO), 18F-azomycin arabinoside
rapid metabolism of the compound, which implied com- ( F-FAZA), and 64Cu-labeled methylthiosemicarbazone
18
plex corrections of the input function for recirculating (64Cu-ATSM). The common property of these tracers is that they
radiolabeled metabolites. Subsequently, 3-deoxy-3-18F-flu- all are bioreductive agents and their tissue binding is dependent
orothymidine (18F-FLT, a thymidine analog) was synthe- on tissue oxygen concentration. Specifically, as tissue hypoxia
tized. 18F-FLT has the advantages of 18F-labeling and of a increases, trapping also increases. Depending on the data analy-
more favorable metabolism. However, the BBB limits cel- sis method, different indices are being used to quantify tissue
lular uptake of 18F-FLT, and kinetic analysis is needed to binding, including SUVmax, a percent threshold of SUVmax, or a
accurately quantify DNA synthesis. However, studies pro- fixed tissue-to-blood ratio greater than a predetermined value
vided conflicting results on the correlation between 18F- (hypoxic volume) and distribution volume (derived from plasma
FLT kinetic parameters and Ki-67, an index of cellular input function). The correlation of these indices with fine needle
replication [18]. oxygen electrodes (the gold standard for measuring hypoxia) is
Images were traditionally obtained through a two-tissue variable because of several methodological issues [20]. Even
compartment, four-rate constant kinetic model, which yields though hypoxia tracers were expected to play an important role
parametric images of the blood-to-tissue transport (K1) and in imaging brain tumors since their description in 1990, they do
net 18F-FLT transport into the brain (KFLT). However, very not play an important role in clinical decision making.
recently it was shown that a late acquisition 1-h postinjection PET studies of brain tumor hypoxia are mostly limited
can be reasonably used in the clinical setting eliminating the to the use of 18F-MISO. 18F-MISO freely crosses the blood-
need of the cumbersome kinetic analysis. In physiological brain barrier and rapidly equilibrates within tissues inde-
T2-MR 18F-FDG 18F-FLT
Fig. 18.6 T2-weighted MR, [18F]FDG, and 18F-FLT PET images of a 33-year-old woman with a grade II astrocytoma. There is only very subtle
decrease in glucose metabolism in the right anterior cingulus in comparison to the left size and no tumor 18F-FLT uptake (arrow) [19]
Table 18.4 Most important favorable and unfavorable characteristics

of common current radiopharmaceuticals for brain tumor imaging Key Learning Points
Advantages Disadvantages • Several radiotracers are available for brain tumor
[18F]FDG – Easily available, overall – High-physiological imaging, amino acid tracers (most commonly,
good accuracy, good uptake in normal brain [11C]methionine, 18F-fluoroethyltyrosine, and
quality images – False positive by
– Allows tumor grading radiation necrosis and
18
F-fluorodopa) and [18F]FDG being the most fre-
– Correlates with prognosis infection quently used.
Amino – Negligible uptake in – Tracer uptake is • Images are acquired through static acquisition,
acid normal brain poorly affected by while image interpretation is based on visual
tracers – Optimal tumor delineation tumor grade
inspection of transversal, coronal, and sagittal PET
and contrast
– Correlates with prognosis and fused PET/MR slices.
– Very high negative • With the noteworthy exceptions of [18F]FDG and
predictive value for 18
F-FMISO, all remaining tracers have only negli-
distinguishing recurrence
gible uptake in the cortex, which makes detection of
from necrosis
– Useful for definition of gliomas more straightforward.
radiotherapy plan
18
F-FLT – Negligible uptake in – Uptake heavily
normal brain affected by BBB
– Optimal tumor delineation integrity
and contrast – Kinetic analysis with 18.3 Indications to PET/CT or PET/MR
– Superior to amino acid metabolite correction
tracers for grading is required to The clinical indications to brain tumor imaging together with
– Allows differential distinguish BBB
some examples taken by the scientific literature are listed in
diagnosis between tumor disruption from
recurrence and necrosis increased metabolism (Table 18.3).
1. Confirmation of diagnostic suspicion of cancer (meta-

pendent of perfusion. Tracer uptake in glioblastoma bolic characterization). In the majority of cases, MR is
multiforme is heterogeneous. Increased 18F-MISO tumor the initial study used for the diagnosis of brain tumors,
uptake is generally found in the periphery but not in the while PET is used as a second-line examination to sup-
necrotic center of glioblastomas multiforme. The necrotic port the diagnosis and define the extent of the lesion.
center is photopenic. The latter finding is expected, because Early studies in patients with brain tumor stressed the rel-
only viable cells are able to accumulate 18F-MISO and evance of the functional information provided by PET in
delivery to necrotic tissue is low. This pattern is also con- comparison to the morphologic information provided by
sistent with the spatial distribution of the hypoxia-inducible CT and MR. Much of this early work was performed by
factor 1 that is expressed in tumor areas adjacent to necro- the group of Di Chiro at the National Institutes of Health
sis [21]. No correlation was found between delayed [23]. At that time, quantification was considered a must.
18
F-MISO uptake and [15O]water activity in brain tumors, Di Chiro et al. first used PET with [18F]FDG in 23 glioma
thus indicating that hypoxia may occur also in cerebral patients. All ten high-grade gliomas demonstrated focal
areas with normal CBF. tracer uptake that was easily visible. Mean CMRglc in
Static images of 18F-FMISO are acquired at 2 h postin- high-grade gliomas (7.4 ± 3.5 mg/100 g/min) was signifi-
jection. Hypoxia tracers also have a physiological signal cantly higher than in low-grade gliomas
in the brain possibly related to their lipophilic nature and (4.0 ± 1.8 mg/100 g/min) that displayed no visible hot
consequent associated nonspecific uptake. In brain spot. This finding was considered a major achievement
tumors, however, hypoxia definitely increases and so and led to the recommendation that [18F]FDG PET is use-
does the signal. In an attempt to simplify the data analy- ful for tumor grading in vivo. Using receiver operating
sis method, it was shown that an image-derived region, curves (ROC) analysis, sensitivity and specificity were,
such as the cerebellum, can be used as a surrogate for respectively, 94% and 77% [23]. Despite significant tech-
blood sampling in quantification of hypoxia. Thus, static nological improvements, results very similar to these
imaging of 18F-MISO captures both maximal intensity were reported in a much larger study 13 years later [24].
and the spatial extent (hypoxic volume) of brain tumor PET with [11C]methionine has high sensitivity (about
hypoxia [22]. 75–95%) for distinguishing gliomas from nonmalignant
The main characteristics of most common radiotracers tumors, with lower values in low-grade gliomas, where
used for imaging brain tumors are outlined in Table 18.4. uptake may occasionally be negligible.
Numerous studies assessed the value of 18F-FET PET in Table 18.5 Sensitivity and specificity for diagnosis of brain gliomas/
glioma recurrence for most common PET tracers
the diagnosis of primary brain tumors. 18F-FET PET was
shown to be more accurate than [18F]FDG PET, 18F-FLT Sensitivity Specificity
Tracer (95% CI) (95% CI) Indication
PET, and 18F-fluorocholine PET to detect brain tumors
[18F]FDG 94% 77% Diagnosis
[25]. A recent meta-analysis including 13 studies with 18F-
[18F]FDG 75% 81% Necrosis vs. tumor
FET PET in 462 glioma patients showed sensitivity and recurrence
specificity of 82% (with 95% CI, 74–88%) and 76% (95% [18F]FDG 0.77 0.78 Necrosis vs. tumor
CI, 44–92%), respectively [25]. Using ROC analysis mean (0.66–0.85) (0.54–0.91) recurrence
and maximum tumor-to-background ratios (TBRmean and [11C] 0.70 0.93 Necrosis vs.
methionine (0.50–0.84) (0.44–1.0) recurrence
TBRmax) higher than 1.6 and 2.1, respectively, provided the
[11C] 75–95% 87%–100% Diagnosis
best diagnostic value for differentiating primary brain methionine
tumors from non-tumor lesions. Moreover, TBRmean and 18
F-FET 0.82 0.76 Diagnosis
TBRmax were significantly lower in grade I–II gliomas as (0.74–0.88) (0.44–0.92)
compared to grade III–IV gliomas. Although on a group 18
F-FLT 79% 63% Diagnosis
basis SUV and TBR allow identification of differences in Values among parenthesis represent 95% confidence interval. Numbers
18
F-FET uptake and metabolism between gliomas and non- separated by a line represent range
malignant lesions, on a single subject basis, this distinction
is more difficult because overlap exists [25]. The analysis ratio and tumor grade in astrocytic and oligodendroglial
of time-activity curves allows improving the differentiation tumor, [11C]methionine enabled the most straightforward
between low- and high-grade gliomas. Early (<15 min) visual localization of hot lesions.
maximal uptake followed by a declining retention curve A brief summary of sensitivity and specificity of main
has been related to high-grade glioma, and late (>15 min) radiotracers for the diagnosis of gliomas is presented in
maximal uptake followed by a cumulative curve has been Table 18.5.
related to low-grade tumor [17]. 2 . In vivo tumor grading. If biopsy is not possible or refused
Among other amino acid tracers, 18F-DOPA (reflecting by the patient, it would be desirable to be able to predict
dopamine metabolism) has greater diagnostic power tumor grade using PET, or to use an imaging technique to
compared to MR for the differential diagnosis between guide biopsy to the areas that are more likely to be higher
tumor recurrence and radiation necrosis. Specificity, sen- grade. Sampling errors on stereotactic biopsy based on
sitivity, and accuracy were 92.3%, 44.4%, and 80% for CT or MR imaging may occur, and therefore these tech-
MR and 100%, 88.9%, and 97.1% for 18F-DOPA [16]. niques are unreliable for prediction of tumor grade.
A study compared 18F-DOPA versus [11C]methionine, [18F]FDG is generally considered the gold standard
which is the most widely used amino acid tracer [14]. tracer for predicting tumor grade, because of the parallel
[11C]Methionine and 18F-DOPA images matched in all relationship between tumor grade and glucose (energy)
patients and showed all lesions as hot spots with higher consumption. On the other hand, the predictive value of
uptake than in the contralateral brain. SUVmax and tumor/ [11C]methionine PET for grading is more limited
contralateral cortex ratios did not statistically differ compared to [18F]FDG. Studies showed that there is size-
between the two tracers. The authors concluded that able overlap in [11C]methionine uptake values between
18
F-DOPA is accurate as a surrogate for [11C]methionine high-grade and low-grade gliomas. Semiquantitative
in imaging amino acid transport in malignant cerebral analysis allows some improvement in grading [26].
lesions for the purpose of visualization of vital tumor tis- Different methods for data analysis have been
sue. It combines the good physical properties of 18F with employed for 18F-FLT, and kinetic analysis has improved
the pharmacological properties of [11C]methionine, and it the relationship between PET signal and the biology of
might therefore be a valuable PET radiopharmaceutical gliomas, primarily tumor grade. In a study that was per-
that can be more widely used for brain tumor imaging in formed about a decade ago, Muzi et al. applied a two-tis-
centers without a cyclotron [14]. sue compartment, four-rate constant model to estimate
In a large study enrolling 94 patients with suspected the blood-to-tissue transport constant K1 and the net flux
brain tumor, [11C]choline had greater accuracy than [18F] constant KFLT. Cumbersome arterial blood scanning,
FDG (84% versus 71%). 18F-FLT has high sensitivity together with 90-min dynamic scanning, was needed
(83%) and very high specificity (approaching 100%). [18]. Results showed that K1 was higher for MR contrast-
A multitracer study compared [11C]methionine, [11C] enhancing gliomas than for non-contrast-enhancing glio-
choline, and [18F]FDG: whereas all three tracers showed a mas. KFLT was higher for high- grade tumors than for
similar correlation between the tumor-to-normal cortex low-grade tumors. The flux in non-contrast-enhancing
gliomas was indistinguishable from contralateral normal edges of the tumor is the method chosen to distinguish
brain. Most importantly, for contrast-enhancing tumors, significant from negligible tracer uptake. Many methods
Ki (i.e., blood flow) was higher than KFLT. Taken together, are currently available. Generally a fixed percent thresh-
these results indicated that in gliomas that show break- old value of tumor uptake is the most frequently adopted.
down of the BBB, flow-related transport (K1) dominates However, boundaries may be slightly redefined on the
18
F-FLT uptake (KFLT). Transport across the BBB and basis of visual impression and on the basis of the experi-
modest rates of 18F-FLT phosphorylation might limit the ence of the operator. Accurate delineation of tumor extent
assessment of cellular proliferation using 18F-FLT in is critical for a complete resection. Therefore, double
high-grade gliomas with significant BBB breakdown evaluation with PET and MR, either in integrated PET/
[18]. However, a different study reported that uptake of MR hybrid devices or after fusion of PET and MR
18
F-FLT correlated with tumor grade [27]. Uptake ratios acquired independently, is routinely performed
have similar power as kinetic analysis to predict long- nowadays.
term survival [28] and are nowadays increasingly used for An early study indicated that patients with glioblas-
clinical purposes. 18F-FLT SUVmax was significantly toma multiforme can be treated with standard conformal
higher in high-grade gliomas compared to low-grade gli- fractionated radiotherapy if there is no [18F]FDG uptake,
omas. 18F-FLT was slightly superior to [11C]methionine while an additional boost is indicated in patients with
for tumor grading. increased [18F]FDG uptake.
Higher [11C]choline uptake was detected in high-grade However, amino acid tracers, when available, are more
gliomas compared to low-grade gliomas [29]. However, commonly used in this application [31]. [11C]methionine
there are discrepant findings on the issue whether [11C] is the single amino acid tracer that is most frequently used
choline tumor metabolism, as measured by PET/CT, pre- nowadays to fuse functional PET images with morpho-
dicts or does not predict tumor grade [29]. logical images for defining the radiotherapy plan [32].
18
F-MISO uptake was greater in high-grade gliomas Integrating [11C]methionine PET with MR images is use-
than in low-grade gliomas. In a recent prospective study ful for planning surgery, with ensuing clinical impact in
of glioma patients who underwent both 18F-MISO and about 80% of the procedures. [11C]methionine PET has
[18F]FDG PET examinations, 18F-MISO PET showed an been used to direct biopsy before surgery [33].
improved ability to distinguish glioblastoma from lower [11C]choline PET/CT better delineated the boundaries
grades as compared to [18F]FDG [30]. between lesions and surrounding normal brain tissues
3 . Radiotherapy and surgical planning. Tumor delineation compared to MR, and the radiotherapy target volume was
is important to define the tumor volume that must be irra- changed in 31% of patients. However, until now, there
diated for therapeutic purposes. Adjuvant radiation ther- have been few applications of radiolabeled choline PET/
apy improves tumor control, and it is therefore carried out CT for radiation therapy planning.
in all tumor patients. Inaccurate tumor delineation with 4. Differentiation of tumor recurrence versus radiation
persistence of viable tumor cells beyond the target area injury. This differential diagnosis represents the most
(i.e., geographic misses) is a major cause of treatment common indication to a PET study. Radiotherapy is asso-
failure. In the past, radiotherapy planning volume was ciated with a risk of radiation necrosis. Radiotherapy-
based on MR. The target volume defined on the basis of induced contrast enhancement may be evident both on
morphological imaging is referred to as gross tumor vol- CT and on MR. contrast enhancement occurs as a
ume (GTV). One of the most important limitations in the consequence of BBB damage, while uptake of PET trac-
application of image-guided radiotherapy is the failure to ers, typically of amino acid tracers, is less affected by
define the tumor extension precisely, which can be attrib- BBB nonspecific changes. [18F]FDG PET has a some-
uted to nonspecific changes of MR after surgery or radia- what limited accuracy for distinguishing tumor recur-
tion therapy. Thus, PET has been used to identify the rence from radiation necrosis. Increased [18F]FDG
metabolically active brain areas. The metabolically active accumulation may occur following radiotherapy owing
volume, as identified by PET, is referred to as biological to inflammation with a time course that is difficult to
target volume (BTV). Logical combination of GTV and predict.
BTV provides the ultimate planning target volume (PTV). [11C]Methionine PET has greater accuracy than [18F]
PET with [18F]FDG showed promises for improving the FDG PET for the differentiation between tumor recur-
accuracy of defining the target volume and for radiation rence and radiation injury. Specificity of [11C]methionine
dose escalation. Use of PET in this context is considered PET ranged between 87% and 100% [26]. Semiquantitative
a standard procedure. The critical factor in defining the analysis improves this differential diagnosis [34].
[11C]choline PET/CT could be used for the differential bolic reduction after therapy predicted longer survival.
diagnosis between tumor recurrence and radiation necro- PET with [18F]FDG changed the intended management in
sis with higher sensitivities and specificity than MR [29]. 38% of patients with primary brain tumors [38]. [18F]FDG
18
F-DOPA PET/CT is highly sensitive and specific for PET volumes were predictive of survival and time to
detection of glioma recurrence, it is superior to [18F]FDG, tumor progression [39].
and it is especially advantageous in patients with low- Prolonged survival was shown in glioma patients with no
grade gliomas [16, 35]. Tripathi et al. directly compared [11C]methionine uptake after surgical resection as com-
the performance of 18F-DOPA, [18F]FDG, and 18F-FLT in pared to those with residual [11C]methionine uptake [40].
evaluating primary and recurrent low-grade gliomas: In gliomas, higher tumor-to-contralateral [11C]methionine
SUVmax for 18F-DOPA was 2–3 times higher than for 18F- count ratios are associated with reduced survival. The
FLT, but it was lower than for [18F]FDG; however, image prognostic value of [11C]methionine uptake has been con-
contrast was better in 18F-DOPA than in [18F]FDG images. sistently replicated in various studies [41].
The authors concluded that 18F-DOPA is superior to 18F- 18
F-FET can be used to identify residual tumor after sur-
FLT and [18F]FDG for low-grade gliomas [36]. gery, and postsurgical tumor metabolic volumes are pre-
5 . Prediction of prognosis. PET was shown to predict progno- dictive of progression-free survival and overall survival
sis in brain tumor patients with any tested radiotracer. Such [42]. Patients with greater decrease in 18F-FET uptake
prognostic information can be obtained at initial staging, by after antiangiogenic therapy with irinotecan-bevacizumab
assessing the response to therapy, and in the assessment of displayed longer survival [43].
tumor recurrence. Evaluation of response to treatment is a Harris et al. created parametric response maps of
general cornerstone of oncological imaging. Such applica- 18
F-DOPA and 18F-FLT before and after bevacizumab
tion had never reached widespread diffusion for central ner- treatment of recurrent gliomas. They reported that an
vous system (CNS) tumors because their poor response to increase in 18F-DOPA or 18F-FLT uptake on parametric
chemotherapy limited the use of such drugs in the clinical response maps after bevacizumab treatment predicted
practice. However, in the last years, more effective CNS shorter progression-free survival and that 18F-DOPA addi-
chemotherapy agents, such as temozolomide, and antian- tionally predicted overall survival [44].
giogenic drugs, such as bevacizumab, were developed 18
F-DOPA was used at baseline, 2 and 6 weeks after start of
rekindling interest in the assessment of the response to ther- bevacizumab therapy for assessing response to therapy.
apy through PET. Moreover, the capability of CT and MR to 18
F-DOPA metabolic tumor volume was a significant pre-
assess changes in tumor size or contrast enhancement in dictor of progression-free survival and overall survival [45].
response to therapy is limited because the BBB undergoes In posttreatment patients with high-grade gliomas, higher
reactive transient changes of permeability. This phenome- [11C]choline uptake predicted shorter survival [46].
non is called “pseudoprogression,” it can be observed in 18
F-fluorocholine metabolic volume derived from PET/
20–50% of cases, and it can induce overtreatment. Uptake CT predicted response in glioma patients treated with
of radiological contrast media in the brain is limited only by radiotherapy and temozolomide [47].
the molecular weight (i.e., size), and therefore such changes Hypoxic volume measured with 18F-MISO and the area of
in permeability often allow the contrast media to enhance in contrast enhancement in T1-weighted MR are significant
the tumor or in the tumor surroundings. On the other hand, prognostic factors for survival.
uptake of PET tracers in the brain always has a carrier-
mediated component, and it is therefore less influenced by
changes in BBB permeability that favors passive diffusion. Key Learning Points
The first demonstration of the prognostic information pro- • There are several indications to PET/CT or PET/
vided by [18F]FDG goes back to the 1980s, and since then MR for brain tumor imaging, the most common
it has been confirmed consistently [37]. Some authors being the differential diagnosis between tumor
have shown that semiquantitative evaluation might be recurrence and radiation necrosis and through the
superior to the visual evaluation and that the predictive metabolic characterization of the brain mass, to
power of SUVmax was more significant than that of other support the MR-based suspect of glioma.
traditional prognostic factors [37]. [18F]FDG PET was • Sensitivity and specificity are both higher than 80%
useful for monitoring the response to chemotherapy. with most tracers for both indications in most stud-
Following treatment with temozolomide, percent decline ies, with higher specificity for amino acid tracers
of CMRglc from baseline was greater in clinical respond- for the differential diagnosis.
ers than in clinical non-responders, and a greater meta-
MR spectroscopy and [11C]methionine PET to classify

• PET has an increasingly important role in defining inconclusive MR findings as either low-grade gliomas or
the brain tumor edges before surgery and radiother- high-grade gliomas and to detect the most malignant parts
apy, to increase the accuracy of both procedures. within tumors before surgery. They found that areas with
• Many studies have been performed in glioma maximum [11C]methionine uptake provided satisfactory
patients with virtually any tracer and consistently tumor grading. The spatial overlap between cerebral areas of
showed that PET can be successfully used for any high PET uptake and highest metabolite concentrations
clinical indication and that, therefore, PET should determined by MR spectroscopy was limited, which sug-
be used in the work-up of all glioma patients. gests that the diagnostic gain provided by MR spectroscopy
• There are some differences among tracers: amino is limited [49].
acid tracers are best for differentiation between A whole-body hybrid PET/MR system was used in
tumor recurrence and radiation necrosis; moreover, another study by Preuss et al. to obtain [11C]methionine PET
they are often used for defining the radiation ther- and MR imaging data necessary for biopsy planning and
apy plan. neuronavigation in a small series of pediatric patients with
• All tracers carry prognostic information. brain tumor. Satisfactory histopathologic specimens were
• [18F]FDG is more advantageous for tumor grading, obtained in all patients. Additional anesthesia as well as
but it is increasingly less used nowadays as amino CT-related radiation exposure could also be avoided [50]. In
acid tracers find broader acceptance. patients with high-grade gliomas studied with PET/MR and
• Hypoxia tracers, proliferation tracers, and radiola- 18
F-FET, both modalities provided sufficient diagnostic qual-
beled choline are at this moment of second choice ity, and the 18F-FET PET findings were in accordance with
PET tracers. those of conventional MR.
All combined PET/MR imaging studies discussed above
agree on the opportunity of obtaining all necessary imaging
parameters in one session with excellent spatial coregistra-
18.4 PET/MR tion of MR and PET data. Moreover, Dunet et al. demon-
strated the superiority of the combination of dynamic
The development of magnetic field-insensitive PET detector 18
F-FET PET and diffusion-weighted MR over the single
technology paved the way from PET/CT to hybrid PET/MR modality approach for glioma grading and argued in favor of
systems, which have been commercially available since combined 18F-FET PET/MR in the initial assessment of pri-
2011. Due to the higher contrast between gray and white mary brain tumors [51].
matter, MR provides substantial advantages over CT for
diagnosing CNS pathologies. For scanning of neurological
patients, the combination of both PET and MR has often Key Learning Points
proved to be superior to the single modality approach, • PET/CT is the current state-of-the-art imaging tech-
improving workflow and patient comfort. Moreover, regard- nique for brain tumors; it is well established and
ing its research potential, PET/MR improves the ability to widely available for all nuclear medicine centers
characterize different pathophysiological aspects of diseases throughout the world.
and allows optimal image fusion and correction for motion • PET/MR represents an innovative technique, as it
artifacts and for partial volume effect [48]. combines two techniques that are mandatory in the
Studies of PET or PET/CT versus MR had shown that, diagnostic process and avoids exposure to
irrespective of the tracer, PET provides more accurate infor- CT-radiation.
mation than morphological imaging. PET/MR studies with • However, PET/MR is less accessible and so far its
dedicated PET/MR scanners are now providing evidence of clinical use is restricted to academic research
the synergistic effect of these two techniques. centers.
At this time, there are few studies with PET/MR scanners
in brain tumor patients.
Boss et al. were first to use a brain-dedicated hybrid PET/
MR system for imaging brain tumors. They reported that MR 18.5 Conclusions
and PET data acquired in ten patients was of high diagnostic
quality and that there was a high level of concordance Many tracers are nowadays available for studying brain
between tumors to background ratios derived from PET/CT tumors, including [18F]FDG, amino acid (primarily [11C]
and those derived from PET/MR. The same group later used methionine, 18F-FET, and 18F-DOPA), 18F-FLT, radiolabeled
choline, DNA replication, and hypoxia tracers. Amino acid vant temozolomide versus radiotherapy alone on survival in glio-
blastoma in a randomised phase III study: 5-year analysis of the
tracers are sensitive to transport across the BBB and, to EORTC-NCIC trial. Lancet Oncol. 2009;10:459–66.
some extent, protein synthesis; they constitute the gold stan- 6. van den Bent MJ, Taphoorn MJ, Brandes AA, Menten J, Stupp
dard for differentiating tumor recurrence vs. radiation R, Frenay M, et al. Phase II study of first-line chemotherapy with
necrosis. Among amino acid available tracers, [11C]methio- temozolomide in recurrent oligodendroglial tumors: the European
Organization for Research and Treatment of Cancer Brain Tumor
nine is currently more validated in the clinical setting and Group Study 26971. J Clin Oncol. 2003;21:2525–8.
more widely used. Other promising tracers include 7. Chamberlain MC, Raizer J. Antiangiogenic therapy for high-grade
18
F-DOPA and 18F-FET. [18F]FDG is much less used nowa- gliomas. CNS Neurol Disord Drug Targets. 2009;8:184–94.
days than in the past, but it appears to be more reliable than 8. Olson JJ, Nayak L, Ormond DR, Wen PY, Kalkanis SN,
Committee ACJG. The role of cytotoxic chemotherapy in the
amino acid tracers for prediction of grading. All radiotracers management of progressive glioblastoma: a systematic review
yield prognostic information either at baseline or in response and evidence-based clinical practice guideline. J Neurooncol.
to therapy. 2014;118:501–55.
The differential diagnosis between tumor recurrence and 9. Herholz K. Brain tumors: an update on clinical PET research in
gliomas. Semin Nucl Med. 2017;47:5–17.
radiation injury is the most frequent clinical scenario in 10. Kobayashi K, Hirata K, Yamaguchi S, Manabe O, Terasaka S,
which PET/CT and PET/MR are being used. MR alone has Kobayashi H, et al. Prognostic value of volume-based measure-
both limited sensitivity and specificity. PET, in combination ments on 11C-methionine PET in glioma patients. Eur J Nucl Med
with MR, is increasingly used to define the gross tumor vol- Mol Imaging. 2015;42:1071–80.
11. Van Laere K, Ceyssens S, Van Calenbergh F, de Groot T, Menten J,
ume to be irradiated. A unique capability of PET is to predict Flamen P, et al. Direct comparison of 18F-FDG and 11C-methionine
in vivo tumor grading better than any other technique. High- PET in suspected recurrence of glioma: sensitivity, inter-observer
metabolic activity predicts higher tumor grade and vice variability and prognostic value. Eur J Nucl Med Mol Imaging.
versa. Changes in cerebral tracer uptake of different tracers 2005;32:39–51.
12. Poulsen SH, Urup T, Grunnet K, Christensen IJ, Larsen VA, Jensen
can be used to monitor the response to therapy. Lower meta- ML, et al. The prognostic value of FET PET at radiotherapy
bolic reductions after therapy as well as higher baseline planning in newly diagnosed glioblastoma. Eur J Nucl Med Mol
uptake values have unfavorable prognosis. Imaging. 2017;44:373–81.
The broad radiopharmaceutical availability combined to 13. Pauleit D, Floeth F, Herzog H, Hamacher K, Tellmann L, Muller
HW, et al. Whole-body distribution and dosimetry of O-(2-
the development of PET/CT and, more recently, PET/MR [18F]fluoroethyl)-L-tyrosine. Eur J Nucl Med Mol Imaging.
hybrid devices has enabled a great improvement in the 2003;30:519–24.
knowledge of biological properties of brain tumors and to 14. Becherer A, Karanikas G, Szabo M, Zettinig G, Asenbaum S,
some extent has facilitated the difficult diagnostic task. Most Marosi C, et al. Brain tumour imaging with PET: a comparison
between [18F]fluorodopa and [11C]methionine. Eur J Nucl Med Mol
centers are equipped with PET/CT scanners, while only few Imaging. 2003;30:1561–7.
academic research centers are also equipped with PET/MR 15. Morana G, Puntoni M, Garre ML, Massollo M, Lopci E, Naseri M,
scanners. PET/MR scanners have logistic and technical et al. Ability of 18F-DOPA PET/CT and fused 18F-DOPA PET/MRI
advantages: with one single imaging session, both PET and to assess striatal involvement in paediatric glioma. Eur J Nucl Med
Mol Imaging. 2016;43:1664–72.
MR, which are both needed for diagnostic purposes, can be 16.
Karunanithi S, Sharma P, Kumar A, Khangembam BC,
performed (“one-stop-shop”), and image processing (accu- Bandopadhyaya GP, Kumar R, et al. 18F-FDOPA PET/CT for
rate coregistration, possibly motion and partial volume cor- detection of recurrence in patients with glioma: prospective com-
rection) can be achieved. parison with 18F-FDG PET/CT. Eur J Nucl Med Mol Imaging.
2013;40:1025–35.
17. Calcagni ML, Galli G, Giordano A, Taralli S, Anile C, Niesen A,
et al. Dynamic O-(2-[18F]fluoroethyl)-L-tyrosine (F-18 FET) PET
References for glioma grading: assessment of individual probability of malig-
nancy. Clin Nucl Med. 2011;36:841–7.
1. Herholz K. Amino acid PET and clinical management of glioma 18. Muzi M, Spence AM, O’Sullivan F, Mankoff DA, Wells JM, Grierson
patients. Eur J Nucl Med Mol Imaging. 2010;37:683–4. JR, et al. Kinetic analysis of 3′-deoxy-3′-18F-fluorothymidine in
2. Jansen NL, Suchorska B, Wenter V, Schmid-Tannwald C, Todica patients with gliomas. J Nucl Med. 2006;47:1612–21.
A, Eigenbrod S, et al. Prognostic significance of dynamic 18F-FET 19. Choi SJ, Kim JS, Kim JH, Oh SJ, Lee JG, Kim CJ, et al. [18F]3′-
PET in newly diagnosed astrocytic high-grade glioma. J Nucl Med. deoxy-3′-fluorothymidine PET for the diagnosis and grading of
2015;56:9–15. brain tumors. Eur J Nucl Med Mol Imaging. 2005;32:653–9.
3. Kraus JA, Lamszus K, Glesmann N, Beck M, Wolter M, Sabel M, 20. Kisely M, Emri M, Lengyel Z, Kalvin B, Horvath G, Tron L,
et al. Molecular genetic alterations in glioblastomas with oligoden- et al. Changes in brain activation caused by caloric stimulation in
droglial component. Acta Neuropathol. 2001;101:311–20. the case of cochleovestibular denervation – PET study. Nucl Med
4. Fink JR, Muzi M, Peck M, Krohn KA. Multimodality brain tumor Commun. 2002;23:967–73.
imaging: MR imaging, PET, and PET/MR imaging. J Nucl Med. 21.
Bruehlmeier M, Roelcke U, Schubiger PA, Ametamey
2015;56:1554–61. SM. Assessment of hypoxia and perfusion in human brain tumors
5. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ, using PET with 18F-fluoromisonidazole and 15O-H2O. J Nucl Med.
Janzer RC, et al. Effects of radiotherapy with concomitant and adju- 2004;45:1851–9.
22. Muzi M, Peterson LM, O’Sullivan JN, Fink JR, Rajendran JG, 38. Hillner BE, Siegel BA, Shields AF, Duan F, Gareen IF, Hanna L,
McLaughlin LJ, et al. 18F-Fluoromisonidazole quantification of et al. Impact of dedicated brain PET on intended patient manage-
hypoxia in human cancer patients using image-derived blood sur- ment in participants of the national oncologic PET Registry. Mol
rogate tissue reference regions. J Nucl Med. 2015;56:1223–8. Imaging Biol. 2011;13:161–5.
23. Di Chiro G, DeLaPaz RL, Brooks RA, Sokoloff L, Kornblith PL, 39. Tralins KS, Douglas JG, Stelzer KJ, Mankoff DA, Silbergeld DL,
Smith BH, et al. Glucose utilization of cerebral gliomas measured Rostomily RC, et al. Volumetric analysis of 18F-FDG PET in glio-
by [18F] fluorodeoxyglucose and positron emission tomography. blastoma multiforme: prognostic information and possible role in
Neurology. 1982;32:1323–9. definition of target volumes in radiation dose escalation. J Nucl
24. Delbeke D, Meyerowitz C, Lapidus RL, Maciunas RJ, Jennings Med. 2002;43:1667–73.
MT, Moots PL, et al. Optimal cutoff levels of F-18 fluorodeoxy- 40. Nariai T, Tanaka Y, Wakimoto H, Aoyagi M, Tamaki M, Ishiwata
glucose uptake in the differentiation of low-grade from high-grade K, et al. Usefulness of L-[methyl-11C] methionine-positron emis-
brain tumors with PET. Radiology. 1995;195:47–52. sion tomography as a biological monitoring tool in the treatment of
25. Dunet V, Rossier C, Buck A, Stupp R, Prior JO. Performance glioma. J Neurosurg. 2005;103:498–507.
of 18F-fluoro-ethyl-tyrosine (18F-FET) PET for the differential 41. Kim S, Chung JK, Im SH, Jeong JM, Lee DS, Kim DG, et al.
diagnosis of primary brain tumor: a systematic review and meta- 11
C-methionine PET as a prognostic marker in patients with glioma:
analysis. J Nucl Med. 2012;53:207–14. comparison with 18F-FDG PET. Eur J Nucl Med Mol Imaging.
26. Kubota R, Kubota K, Yamada S, Tada M, Takahashi T, Iwata R, 2005;32:52–9.
et al. Methionine uptake by tumor tissue: a microautoradiographic 42. Piroth MD, Holy R, Pinkawa M, Stoffels G, Kaiser HJ, Galldiks
comparison with FDG. J Nucl Med. 1995;36:484–92. N, et al. Prognostic impact of postoperative, pre-irradiation
27. Collet S, Valable S, Constans JM, Lechapt-Zalcman E, Roussel 18
F-fluoroethyl-l-tyrosine uptake in glioblastoma patients treated
S, Delcroix N, et al. [18F]-fluoro-L-thymidine PET and advanced with radiochemotherapy. Radiother Oncol. 2011;99:218–24.
MRI for preoperative grading of gliomas. Neuroimage Clin. 43. Hutterer M, Nowosielski M, Putzer D, Waitz D, Tinkhauser G,
2015;8:448–54. Kostron H, et al. O-(2-18F-fluoroethyl)-L-tyrosine PET predicts
28. Chen W, Delaloye S, Silverman DH, Geist C, Czernin J, Sayre J, failure of antiangiogenic treatment in patients with recurrent high-
et al. Predicting treatment response of malignant gliomas to bevaci- grade glioma. J Nucl Med. 2011;52:856–64.
zumab and irinotecan by imaging proliferation with [18F]fluorothy- 44. Harris RJ, Cloughesy TF, Pope WB, Nghiemphu PL, Lai A, Zaw T,
midine positron emission tomography: a pilot study. J Clin Oncol. et al. 18F-FDOPA and 18F-FLT positron emission tomography para-
2007;25:4714–21. metric response maps predict response in recurrent malignant glio-
29. Giovannini E, Lazzeri P, Milano A, Gaeta MC, Ciarmiello
mas treated with bevacizumab. Neuro Oncol. 2012;14:1079–89.
A. Clinical applications of choline PET/CT in brain tumors. Curr 45. Schwarzenberg J, Czernin J, Cloughesy TF, Ellingson BM,

Pharm Des. 2015;21:121–7. Pope WB, Grogan T, et al. Treatment response evaluation using
30. Hirata K, Terasaka S, Shiga T, Hattori N, Magota K, Kobayashi H, 18
F-FDOPA PET in patients with recurrent malignant glioma on
et al. 18F-Fluoromisonidazole positron emission tomography may bevacizumab therapy. Clin Cancer Res. 2014;20:3550–9.
differentiate glioblastoma multiforme from less malignant gliomas. 46. Li W, Ma L, Wang X, Sun J, Wang S, Hu X. 11C-choline PET/
Eur J Nucl Med Mol Imaging. 2012;39:760–70. CT tumor recurrence detection and survival prediction in post-
31. Douglas JG, Stelzer KJ, Mankoff DA, Tralins KS, Krohn KA, Muzi treatment patients with high-grade gliomas. Tumour Biol.
M, et al. [F-18]-fluorodeoxyglucose positron emission tomography 2014;35:12353–60.
for targeting radiation dose escalation for patients with glioblas- 47. Bolcaen J, Acou M, Boterberg T, Vanhove C, De Vos F, Van den
toma multiforme: clinical outcomes and patterns of failure. Int J Broecke C, et al. 18F-FCho PET and MRI for the prediction of
Radiat Oncol Biol Phys. 2006;64:886–91. response in glioblastoma patients according to the RANO criteria.
32. Grosu AL, Lachner R, Wiedenmann N, Stark S, Thamm R,
Nucl Med Commun. 2017;38:242–9.
Kneschaurek P, et al. Validation of a method for automatic image 48. Catana C, Drzezga A, Heiss WD, Rosen BR. PET/MRI for neuro-
fusion (BrainLAB System) of CT data and 11C-methionine-PET logic applications. J Nucl Med. 2012;53:1916–25.
data for stereotactic radiotherapy using a LINAC: first clinical 49. Bisdas S, Ritz R, Bender B, Braun C, Pfannenberg C, Reimold M,
experience. Int J Radiat Oncol Biol Phys. 2003;56:1450–63. et al. Metabolic mapping of gliomas using hybrid MR-PET imag-
33. Pirotte B, Goldman S, Van Bogaert P, David P, Wikler D, Rorive ing: feasibility of the method and spatial distribution of metabolic
S, et al. Integration of [11C]methionine-positron emission tomo- changes. Investig Radiol. 2013;48:295–301.
graphic and magnetic resonance imaging for image-guided sur- 50. Preuss M, Werner P, Barthel H, Nestler U, Christiansen H, Hirsch
gical resection of infiltrative low-grade brain tumors in children. FW, et al. Integrated PET/MRI for planning navigated biopsies in
Neurosurgery. 2005;57:128–39.. discussion 39 pediatric brain tumors. Childs Nerv Syst. 2014;30:1399–403.
34. Terakawa Y, Tsuyuguchi N, Iwai Y, Yamanaka K, Higashiyama S, 51. Dunet V, Maeder P, Nicod-Lalonde M, Lhermitte B, Pollo C, Bloch
Takami T, et al. Diagnostic accuracy of 11C-methionine PET for dif- J, et al. Combination of MRI and dynamic FET PET for initial gli-
ferentiation of recurrent brain tumors from radiation necrosis after oma grading. Nuklearmedizin. 2014;53:155–61.
radiotherapy. J Nucl Med. 2008;49:694–9.
35. Talbot JN, Kerrou K, Montravers F, Nataf V, Chevalme Y. FDOPA
PET has clinical utility in brain tumour imaging: a proposal for
a revision of the recent EANM guidelines. Eur J Nucl Med Mol
Imaging. 2007;34:1131–2.
Further Reading
36. Tripathi M, Sharma R, D’Souza M, Jaimini A, Panwar P, Varshney
R, et al. Comparative evaluation of F-18 FDOPA, F-18 FDG, and Albert NL, Weller M, Suchorska B, Galldiks N, Soffietti R, Kim MM,
F-18 FLT-PET/CT for metabolic imaging of low grade gliomas. et al. Response Assessment in Neuro-Oncology working group
Clin Nucl Med. 2009;34:878–83. and European Association for Neuro-Oncology recommenda-
37. Colavolpe C, Chinot O, Metellus P, Mancini J, Barrie M, Bequet- tions for the clinical use of PET imaging in gliomas. Neuro Oncol.
Boucard C, et al. FDG-PET predicts survival in recurrent high- 2016;18:1199–208.
grade gliomas treated with bevacizumab and irinotecan. Neuro Cornelius JF, Langen KJ, Stoffels G, Hänggi D, Sabel M, Jakob Steiger
Oncol. 2012;14:649–57. H. Positron emission tomography imaging of meningioma in clini-
cal practice: review of literature and future directions. Neurosurgery. Galldiks N, Law I, Pope WB, Arbizu J, Langen KJ. The use of amino
2012;70:1033–41. acid PET and conventional MRI for monitoring of brain tumor ther-
Ferda J, Ferdová E, Hes O, Mraček J, Kreuzberg B, Baxa J. PET/ apy. Neuroimage Clin. 2016;13:386–94.
MRI: multiparametric imaging of brain tumors. Eur J Radiol. Herholz K. Brain tumors: an update on clinical PET research in glio-
2017;94:A14–25. mas. Semin Nucl Med. 2017;47:5–17.
Hybrid Imaging of the Head
and Neck Region 19
Alejandro Fernández and Valle Camacho
Contents
19.2 Indications to Perform the Diagnostic or Therapeutic Procedure 433
19.2.1 Pretreatment Staging 433
19.2.2 Radiotherapy Planning 433
19.2.3 Monitoring of Response to Therapy 434
19.2.4 Follow-Up 435
19.2.5 Cervical Carcinoma of Unknown Origin 436
19.3 Contraindications (Relative) 438
19.3.1 Pregnancy 438
19.3.2 Breast-Feeding 438
19.3.3 Lack of Cooperation 438
19.4 Acquisition Procedures and Guidelines 438
19.4.1 Patient Preparation 438
19.4.2 Radiopharmaceutical 439
19.4.3 Acquisition 439
19.5 ata Processing
D 440
19.5.1 PET Image Reconstruction 440
19.5.2 CT Image Reconstruction 440
19.6 Data Interpretation and Reporting 440
19.6.1 Display 440
19.6.2 Physiological [18F]FDG Distribution and Interpretation Criteria 441
19.6.3 Documentation and Reporting 441
19.6.4 Description of the Findings 444
19.7 Future Perspectives 445
19.7.1 New Oncological Tracers 445
19.7.2 PET/MRI 445
19.7.3 Digital PET/CT 445
References 446
Learning Objectives • Understand the rationale underlying the common clin-

• Summarize the main features of histology, aetiology ical uses of [18F]FDG PET/CT in patients with head
and clinical presentation for head and neck cancers. and neck cancers.
• Summarize the main principles of treatment according • Describe the established indications for [18F]FDG
to stage of the disease. PET/CT in patients with head and neck cancers,
including initial staging, follow-up, radiotherapy plan-
ning and assessment of response to treatments.
A. Fernández (*) · V. Camacho • Describe the use of [18F]FDG PET/CT imaging in
Nuclear Medicine Department, Hospital de la Santa Creu i Sant patients with metastasis from an unknown primary
Pau, Barcelona, Spain
e-mail: AFernandezLe@santpau.cat tumour in the head and neck region.

432 A. Fernández and V. Camacho
• Describe the methodology for performing and inter- extent of disease involvement, which are unique in each of
preting [18F]FDG PET/CT in patients with head and the regions and subregions of the head and neck. The N cat-
neck cancers. egory describes the absence or presence of metastasis in the
• Summarize the clinical evidence supporting the use of regional lymph nodes. Lymph node imaging characteristics
tracers other than [18F]FDG for PET/CT imaging in using size criteria and morphologic findings on CT can assist
patients with head and neck cancers. in determining metastatic disease involvement. A variety of
• Summarize the clinical evidence supporting the use of size criteria is used for the assessment of metastatic nodes.
[18F]FDG PET/MR imaging in patients with head and The M category classifies metastatic disease involvement in
neck cancers. distant soft tissue, osseous lesions or abnormal distant lymph
nodes. In the eighth AJCC staging manual, the most signifi-
cant update creates a separate staging algorithm for high-risk
19.1 Introductory Background human papillomavirus-associated cancer of the oropharynx,
distinguishing it from oropharyngeal cancer with other
The term head and neck cancer (HNC) generally encom- causes. Other modifications include the reorganizing of skin
passes malignant neoplasms of soft tissue origin of the oral cancer (other than melanoma and Merkel cell carcinoma)
cavity, lips, nasal cavity, paranasal sinuses, pharynx, larynx from a general chapter for the entire body to a head and
and salivary glands, as well as sarcomas arising in this region. neck-specific cutaneous malignancies chapter; the division
The skin is sometimes included as well. About 95% are squa- of cancer of the pharynx into three separate chapters; changes
mous cell carcinomas (or variants) arising from the mucosa or to the tumour (T) categories for oral cavity, skin and naso-
adenocarcinomas from the associated secretory glands [1]. pharynx; and the addition of extranodal cancer extension to
The major risk factors for squamous cell carcinoma of the lymph node category (N) in all but the viral-related cancers
head and neck (HNSCC) are a history of tobacco and alcohol and mucosal melanoma [8].
use; risk is increased when either of these factors is present The factors that determine the initial treatment are those
individually, but that risk is multiplied when both factors are related to tumour characteristics (size, location, histology
present in combination [2]. Over the past decade, the impor- and nodal and metastatic involvement), the patient (age and
tance of infection by the human papillomavirus (HPV) in the general condition) and the factors depending on availability.
pathogenesis of oropharyngeal cancers has been recognized Most patients present with complicated locally advanced
[3], and Epstein-Barr virus has emerged as an important pre- disease requiring multidisciplinary treatment plans based on
disposing condition for the development of non-keratinizing variable combinations of surgery, radiation therapy and
squamous cell carcinoma of the nasopharynx [4]. chemotherapy.
Positron emission tomography (PET) imaging with the This chapter reviews the role of [18F]FDG PET/CT in the
radiolabelled glucose analogue 2-deoxy-2-[ F]fluoro-d- imaging of the head and neck cancer, whereas surgical stag-
18
glucose ([18F]FDG) plays an increasingly important role in ing of the disease with radioguided sentinel lymph node
the pretreatment staging, radiotherapy planning, treatment biopsy is discussed in Chap. 16 of this book (“Image-Guided/
response assessment and post-therapy follow-up in various Radioguided Surgery”).
HNSCCs. [18F]FDG PET/CT is superior to contrast-enhanced
computed tomography (CT) and magnetic resonance imag-
ing (MRI) in the detection of carcinomas of unknown pri-
mary (CUP), of cervical lymph node involvement and in the Key Learning Points
identification of distant metastases [5, 6]. • When excluding thyroid/parathyroid tumours and
The pattern of [18F]FDG accumulation within the neck is paragangliomas, about 95% of primary cancers
subject to a wide range of physiologic activity which may be arising in the head and neck region are squamous
difficult to distinguish from malignancy. For example, meta- cell carcinomas.
bolically active brown fat, physiologic activity in neck mus- • Infection by the human papillomavirus is recog-
cles and atherosclerotic disease can have [18F]FDG activity nized as an important predisposing condition for the
similar to that of metastatic adenopathy. Correlative CT development of squamous cell carcinomas of the
imaging provides anatomic localization of these findings and head and neck.
is useful to distinguish physiologic activity from metastatic • [18F]FDG PET/CT has an important role for pre-
disease. treatment staging, radiotherapy planning, assess-
Squamous cell carcinoma of the head and neck is classi- ment of response to therapy and post-treatment
fied using TNM staging criteria according to the American follow-up of squamous cell carcinomas of the head
Joint Committee on Cancer (AJCC) staging manual [7]. The and neck.
T category is based on the size of the primary tumour and the
19 Hybrid Imaging of the Head and Neck Region 433
(MRI, CT and ultrasounds) were 75% and 79%, respectively

• Special attention is required when interpreting [18F] [10, 11].
FDG PET/CT images of the head and neck, due to Approximately 4–15.4% of patients with HNSCC have
physiologic accumulation of [18F]FDG in some distant metastases at initial presentation. The most common
anatomic structures of this region. sites of metastasis include the lung, bone and abdomen.
• TNM staging is important for selecting the most Whole-body [18F]FDG PET/CT is more accurate than con-
appropriate treatment strategy in patients with head ventional imaging for detection of metastatic foci [12]. In
and neck cancers. addition, [18F]FDG PET/CT detects distant metastases or a
second primary tumour in up to 15% patients with squamous
cell carcinoma, which can significantly alter treatment plans
[9]. In a multicentre prospective study to evaluate the impact
19.2 I ndications to Perform the Diagnostic of PET/CT in HNSCC, Lonneux et al. [13] found that PET/
or Therapeutic Procedure CT improved the TNM classification of the disease and
altered the management of 13.7% of the patients with
19.2.1 Pretreatment Staging HNSCC.
In 2017, the National Comprehensive Center Network
The conventional staging of HNSCC typically consists of updated the clinical practice guidelines for PET/CT imaging
detailed physical examination, endoscopy and CT scan of the of head and neck cancer and suggested the use of PET/CT
neck and chest. MRI is often added when assessing oral cavity for initial staging of oral cavity, oropharyngeal, hypopharyn-
and paranasal sinus cancers, particularly in those tumours with geal, glottic and supraglottic cancers for stage III–IV dis-
a known predilection for perineural invasion or where second- ease, as well as mucosal melanoma and nasopharyngeal
ary changes in sinus air cells can be difficult to differentiate carcinoma (World Health Organization class 2–3 and N2–3
from local disease; however, numerous reports on initial stag- diseases) [14].
ing indicate that the sensitivity of PET/CT is equivalent to or
superior to that of MRI and CT [9]. In comparison to morpho-
logical imaging methods (CT or MRI), PET/CT is particularly 19.2.2 Radiotherapy Planning
advantageous in allowing assessment of neck lymph nodes,
potential distant metastases and synchronous second prima- The importance of PET/CT in the planning of radiotherapy
ries in a single examination (see Figs. 19.1 and 19.2). Different in HNSSC patients has been extensively demonstrated (see
meta-analysis studies have been published reporting a sensi- Fig. 19.3). In a prospective study by Ciernik et al. [15], [18F]
tivity and specificity of [18F]FDG PET/CT in staging nodal FDG PET/CT defined tumour volume more definitively than
disease of about 79–85% and 84–86%, respectively, whereas diagnostic CT without contrast medium [16, 17]. [18F]FDG
sensitivity and specificity of conventional diagnostic tests PET/CT modified staging and radiotherapeutic planning in
Fig. 19.1 A 61-year-old man recently diagnosed with epiglottis carcinoma with bilateral adenopathy. Transaxial slices of CT, [18F]FDG and fused
PET/CT images show hypermetabolism in the right epiglottis (thin arrow) and in bilateral lymph nodes
Fig. 19.2 A 75-year-old man with oropharynx carcinoma with bilat- show primary tumour and right lymph node. The study also demon-
eral metastatic right lymph node in CT. Volumetric display and trans- strates multiple bone metastasis and subcentimetric pulmonary nodes
axial, coronal and sagittal TC, [18F]FDG PET/CT and fused images (arrow)
up to one-third of untreated HNSSC patients [18–20]. accurate identification of viable tumour within residual
Abramyuk et al. [20] compared the staging modifications masses, identification of small tumour deposits and the char-
determined by [18F]FDG PET/CT in 102 patients with acterization of secondary enlarged inflammatory lymph nodes.
untreated primary HNSSC. [18F]FDG PET/CT imaging led PET/CT has significant advantages for treatment response
to modification in RT planning in 14 of 102 patients (13.7%). assessment as it is a functional imaging approach, and does
As such, [18F]FDG PET/CT improved selection of candi- not rely on morphological changes [21] (see Fig. 19.4).
dates for curative and palliative RT. In general, focal and asymmetric [18F]FDG uptake with
intensity greater than in surrounding normal tissues (in par-
ticular, muscle) and blood vessels should be considered sug-
19.2.3 Monitoring of Response to Therapy gestive of residual disease. On the other hand, diffuse
(nonfocal) [18F]FDG uptake within the radiation field is usu-
Chemoradiation has been used for the treatment of locoregion- ally an indicator of post-radiation inflammation. In a num-
ally advanced HNSCC. It has been accepted as a part of the ber of retrospective and prospective studies, PET/CT has
definitive treatment after surgery. Structural imaging tech- shown a high sensitivity and specificity to detect residual/
niques (CT/MRI) have been used for treatment response recurrent lesions after chemotherapy and radiotherapy
assessment with limited success due to their limitations in treatment [22–24]. In a recent meta-analysis, the authors
Fig. 19.3 In radiotherapy planning, [18F]FDG PET/CT results in significant reduction of metabolic gross tumour volume (GTV) with respect to
GTV planning with CT
support the use of PET/CT 12 weeks post-treatment for the tion of disease. However, early detection of locoregional
assessment of residual or recurrent disease [23], whereas the recurrence may potentially improve survival by facilitat-
potential clinical utility of PET for early response assessing timely salvage treatment. Lee et al. [25], in a meta-
ment (PET interim) during chemoradiotherapy has not been analysis study, described that overall sensitivity and
explored systematically. negative predictive value (NPV) for locoregional recur-
rence were higher using PET/CT (92.5% and 94.8%,
respectively) than conventional imaging (55% and 76.9%,
19.2.4 Follow-Up respectively). They concluded that, for routine surveil-
lance, the initial PET scan should be performed within
The need and frequency of post-treatment imaging assess- 6 months after completion of treatment and the proper
ment for patients treated for HNSCC are still highly con- timing of next routine PET scan for subclinical patient
troversial. It is unclear whether patients with distant with initial negative PET result might be 1 year after ini-
relapse, but without symptoms, benefit from early detec- tial PET scan.
Fig. 19.4 A 63-year-old man with right tonsil carcinoma (T3N0M0) [18F]FDG PET and fused PET/CT images 3 months after treatment
treated with neoadjuvant chemoradiotherapy. Transaxial slices of CT, (bottom row) show disappearance of tumour lesion suggesting com-
[18F]FDG PET and fused PET/CT images prior to treatment (top row) plete response to treatment
show intense hypermetabolism in right tonsil. Transaxial slices of CT,
19.2.5 Cervical Carcinoma of Unknown Origin can improve the chance of survival as well as can lower the
morbidity by better targeting the treatment options, includ-
An unknown primary tumour in the neck is diagnosed when ing surgery, and decreasing the field of irradiation. According
a patient presents with a neck metastasis, but no primary to several studies, [18F]FDG PET/CT is very helpful in local-
tumour is found. In this context, PET/TC was added to the ization of primary tumour. Lee et al. [27] have demonstrated
diagnostic workup at the beginning of its clinical use (see that [18F]FDG PET/CT showed higher sensitivity (69%) for
Fig. 19.5). The treatment of an unknown primary tumour can detection of occult primary tumours than did CT (16%)
consist of neck dissections, tonsillectomies and radiation (P < 0.001) or combined CT and MRI (41%, P = 0.039) in 56
therapy for all mucosal sites and both sides of the neck [26]. patients with cervical metastasis from an unknown primary
Detection of a primary tumour has implications for ther- tumour.
apy in that targeted treatment of a known primary tumour
Fig. 19.5 A 70-year-old man with metastatic neck lymph node of lymph node left neck area (thin arrow) and hypermetabolism in left
unknown primary tumour. Coronal, sagittal and transaxial slices of CT, tonsil (thick arrow). Surgical histopathology demonstrated a primary
[18F]FDG PET and fused PET/CT images show hypermetabolism in tonsil tumour
in the milk [29]. However, as the lactating breast accumulates

Key Learning Points [18F]FDG, it is suggested that contact between mother and
• The performance of [18F]FDG PET/CT is equiva- child be limited for 12 h after injection of [18F]FDG in order to
lent to or superior to that of CT and MRI for initial reduce the radiation dose that the infant receives from external
pretreatment staging of patients with head and neck exposure to radiation emitted by the mother [28].
cancers.
• Distant metastatic sites in patients with head and
neck cancers are detected by [18F]FDG PET/CT 19.3.3 Lack of Cooperation
with greater sensitivity and specificity than conven-
tional imaging with either CT or MRI. The lack of cooperation or the inability to cooperate with the
• NCCN guidelines recommend [18F]FDG PET/CT procedure may be relative contraindication.
for initial staging of oral cavity, oropharyngeal,
hypopharyngeal, glottic and supraglottic cancers
for stage III–IV disease. Key Learning Point
• In a substantial fraction of patients with head and • Relative contraindications to perform an [18F]FDG
neck cancers, integration of [18F]FDG PET/CT in scan should be considered during pregnancy and
the pretreatment imaging protocol planning results breast-feeding.
in modification of radiotherapy planning.
• [18F]FDG PET/CT has greater sensitivity and speci-
ficity than either CT or MRI for assessing response
to therapy in patients with head and neck cancers 19.4 Acquisition Procedures
submitted to chemotherapy and/or radiotherapy. and Guidelines
• During post-therapy follow-up of patients treated
for head and neck cancer, [18F]FDG PET/CT has PET/CT imaging combines both anatomic and functional
greater sensitivity and negative predictive value imaging, thus potentially providing more accurate diagnosis
than conventional imaging. and improved patient management [28, 30]. Intravenous
• [18F]FDG PET/CT has greater sensitivity than either contrast is important in CT evaluation of the neck for defin-
CT or combined CT and MRI for detecting occult ing vascular structures and characterizing lymph nodes.
primary tumours in the head and neck region. Newer PET/CT scanners equipped with multidetector CT
can provide images that are equivalent in quality to dedicated
neck CT [31].
19.3 Contraindications (Relative)

19.4.1 Patient Preparation
19.3.1 Pregnancy
The main purpose of patient preparation is to reduce tracer
For any diagnostic procedure in a female patient known or uptake in normal tissue (kidneys, bladder, skeletal muscle,
suspected to be pregnant, a clinical decision is necessary in myocardium, brown fat) while maintaining and optimizing
which the benefits are weighed against the possible harm. tracer uptake in the target structures (tumour tissue) while at
The International Commission on Radiological Protection the same time keeping patient radiation exposure levels as
(ICRP) reports that for an adult patient, the administration of low as reasonably possible (ALARA) [28].
259 MBq (7 mCi) of [18F]FDG results in an absorbed radia-
tion dose of 4.7 mGy to the nongravid uterus (i.e. 1. Before reporting to the Nuclear Medicine Centre, nondia-
1.8 × 10–2 mGy/MBq). A pregnancy test may help with the betic patients should not consume any food, simple carbo-
decision, provided the 10-day post-ovulation blackout is hydrates or liquids other than plain (unflavoured) water for
understood [28, 29]. at least 4 h prior to the start of the [18F]FDG PET/CT
study. Type I and insulin-dependent type II diabetic
patients should not have insulin injections for at least 4 h
19.3.2 Breast-Feeding before [18F]FDG injection, and they should be made to
achieve normal glycaemic values prior to the study. Type
The ICRP does not recommend interruption of breast-feeding II non-insulin-dependent diabetic patients should continue
after [18F]FDG administration, since little [18F]FDG is excreted to take oral medication to control their blood sugar level.
2. The blood glucose level should be checked before [18F] ation correction of the PET images as well as for ana-
FDG administration. Tumour uptake of [18F]FDG is tomic localization. The head and neck region is typically
reduced in hyperglycaemic states. Most institutions included in routine skull base-to-proximal thigh PET
reschedule the patient if the blood glucose level is greater study; however, the evaluation of the head and neck in
than 150–200 mg/dL. Reducing the serum glucose level these studies is suboptimal.
by administering insulin can be considered, but the The strategy is:
administration of [18F]FDG should be delayed 4 h after • CT topogram: 120 kV; 10 mA
insulin administration. • Low-dose CT scan: 140 kV; 80 or 50 mA
3. When a diagnostic CT scan with intravenous contrast • PET acquisition from the skull base to proximal thigh
agent enhancement is to be performed as part of the [18F] (time per bed between 1.5 and 3 min)
FDG PET/CT study, indications, contraindications and 2. PET/CT imaging of the neck: the arms should be down to
restrictions have to be assessed by a qualified physician. eliminate streak artefacts from the humerus. To improve
4. Medication that interacts with intravenous contrast agent the spatial resolution, a smaller field of view can be used.
(e.g. metformin for the treatment of diabetes) and relevant The strategy is:
medical history (e.g. compromised renal function, adverse • CT topogram: 120 kV; 10 mA
reactions or claustrophobia) should be taken into • Intravenous contrast-enhanced diagnostic CT scan
consideration. • Head and neck PET acquisition (with 30 s of delay)
If the PET/CT data are used for radiotherapy planning,
the PET/CT imaging of the neck should be performed in
19.4.2 Radiopharmaceutical the position used for radiotherapy treatment, employing
the same dedicated positioning devices (e.g. the same
1 . Product: [18F]FDG. radiotherapy table top, laser alignment, immobilization
2. Nuclide: Fluorine-18. mask and measures).
3. Activity: The minimum recommended administered [18F]
FDG activity and PET acquisition duration for each bed
position must be adjusted. Therefore, one may decide to
apply a higher activity and reduce the duration of the Key Learning Points
study or, preferably, to use a reduced activity and increase • Special attention must be paid to control of serum
the study duration, thereby keeping ALARA principles in glucose level before administering [18F]FDG for a
mind as well. There are different methods for determining PET/CT scan, either in diabetic patients (adequate
the minimum [18F]FDG administered dose in adults. One metabolic control) or in nondiabetic patients (fast-
specification is 3.7 MBq/kg (0.1 mCi/kg), while other ing for at least 6 h).
specifications include the time per bed position and • Hyperglycaemia reduces tumour uptake of [18F]
patient weight. FDG.
4. Administration: Intravenous. • For contrast-enhanced CT as part of the PET/CT
5. Uptake period for [18F]FDG: The recommended interval scan, indications, contraindications, restrictions and
between [18F]FDG administration and the start of acqui- possible use of concomitant medications must be
sition is 60 min. Following the injection, it is important taken into due consideration before administering
for the patient to rest quietly during this period, as the iodinated contrast medium.
excessive motion may result in muscle uptake; talking • The amount of [18F]FDG administered for the scan
should be avoided to minimize vocal cord activity. must be selected on the basis of the individual
Patients may go to the toilet while waiting, preferably patient’s habitus (body weight—with adequate cor-
more than 30 min after injection. Patients should empty rection for paediatric age), according to ALARA
their bladder 5 min before the start of the [18F]FDG principles.
PET/CT study. • Acquisition of the PET/CT scan must follow estab-
lished protocols and guidelines regarding both the
PET components and the CT components of the
19.4.3 Acquisition scan.
• The head and neck region is best explored with the
1. PET/CT imaging from the skull base to proximal thigh: patient’s arms down to avoid streak artefacts from
the arms should be elevated over the head if the patient the humerus.
can tolerate this position. The CT scan is used for attenu-
19.5 Data Processing longitudinal filter in the z-dimension is often used to opti-
mize the resolution in the axial direction and to modify
Raw emission PET data consists of a large number of events the slice sensitivity profiles. For attenuation correction,
along lines of response between detector pairs that have to be only the standard kernels are used.
reconstructed to obtain both the volumetric images and 4. Additionally, post-processing such as volume rendering
sequenced slices in the three spatial [28, 30]. This emission or maximum intensity projections benefits from using the
data must be corrected for attenuation prior to fusion with high-quality reconstructed CT data.
the CT sinograms for quantification and final display.
Key Learning Points

19.5.1 PET Image Reconstruction • The PET emission data must be corrected for sev-
eral factors before further processing.
1. In the case of head and neck studies, the acquired PET • The reconstructed voxel size should be 3–4 mm in
emission data must be corrected for geometrical response, any direction through selection of adequate matrix
detector efficiency, system dead time, random coinci- size and zoom factors.
dences, scatter, attenuation and sampling nonuniformity. • The PET emission data should be reconstructed
Addition of time-of-flight information is recommended both with and without attenuation correction in
when available. order to identify artefacts in the CT attenuation-
2. Matrix sizes and zoom factors should be chosen such that corrected images due to highly attenuating
reconstructed voxel sizes are within 3.0–4.0 mm in any materials.
direction. • The PET emission data can be reconstructed in
3. It is good clinical practice to perform reconstructions either the 2D mode or the 3D mode, using suitable
with and without attenuation correction to resolve issues procedures.
arising from potential artefacts generated by the CT-based • Current reconstruction methods are based on itera-
attenuation correction procedure related to highly attenu- tive approaches.
ating materials, such as contrast agent or metal implants. • CT image reconstruction follows standard proto-
4. Datasets acquired in the 3D mode can either be rebinned cols commonly employed for routine clinical
into 2D data and reconstructed with a 2D algorithm or be imaging.
reconstructed with a fully 3D algorithm.
5. Iterative reconstruction approaches are now a standard for
both 2D and 3D acquisition modes.
6. The reconstructed image volume is displayed in transax- 19.6 Data Interpretation and Reporting
ial, coronal and sagittal planes and as a rotating maximum
intensity projection image. 19.6.1 Display
The software packages of an integrated PET/CT system typi-

19.5.2 CT Image Reconstruction cally provide for image analysis and SUV calculations, reg-
istered and aligned CT and [18F]FDG PET images [28, 30]
1. The CT sinograms acquired during the PET/CT study are displayed in monitors that should be approved for clinical
usually reconstructed using filtered back-projection algo- use in radiology and nuclear medicine [32, 33] or national
rithms. Recently introduced iterative reconstruction guidelines.
methods for CT data may be applied, if available in the
PET/CT system. 1. Fusion images must be evaluated using an SUV scale in
2. CT data are usually reconstructed separately at full field the axial, coronal and sagittal planes as well as maximum
of view for the attenuation correction of PET emission intensity projection (MIP) images for review in the 3D
data and for CT clinical interpretation. The reconstruc- cine mode.
tions will probably differ in their slice thickness, appro- 2. [18F]FDG PET images with and without attenuation cor-
priate zoom, slice overlap, filter, etc. rection should be available for review.
3. Spiral CT data are collected into axial or tilted planes or 3. Quantitative information with respect to size and [18F]
fully 3D. In addition to the reconstruction kernel that FDG uptake (SUV) can be retrieved.
modulates the image characteristics within the slices (i.e. 4. Image data should be stored on an approved PACS system
spatial resolution, edge enhancement and noise texture), a and in DICOM format [34].
19.6.2 Physiological [18F]FDG Distribution uptake for response assessment studies, when large
and Interpretation Criteria changes in body weight may occur during the course of
the treatment [38].
Head and neck anatomy is complex. The presence or absence 9. Although there are no conclusive data on the optimum
of abnormal [18F]FDG accumulation on the PET images, interval between chemotherapy and [18F]FDG PET/CT,
especially focal accumulation, in combination with intensity an interval of at least 10 days between the last treatment
of uptake and anatomical size should be evaluated. Both CT and the [18F]FDG PET/CT examination is generally con-
and PET attenuation-corrected images may need to be sidered adequate for measurement of response [39].
reviewed to identify artefacts caused by patient motion. This is because of the balance between any possible
effects on tumour metabolism (such as macrophage
1. Physiologic uptake of [18F]FDG can be seen to some impairment) and systemic effects (such as bone marrow
extent in every viable tissue, including the tonsils and at activation following bone marrow depression, which
the base of the tongue due to physiological accumulation may or may not be caused by growth factors). If an inter-
in the lymphatic tissue of Waldeyer’s ring. Non- val of 10 days is not possible, [18F]FDG PET/CT should
pathological hypermetabolism can be identified in sali- be delayed as long as possible after the previous chemo-
vary glands, muscles of the floor of the mouth and ocular therapy administration (i.e. until as close as possible to
extrinsic, cervical or masticatory muscles [35] (Fig. 19.6). the next treatment cycle).
2. Uptake in cervicothoracic brown fat is observed more 10. It is assumed that the side effects of radiotherapy are
often in young patients and when the ambient tempera- longer-lasting; investigations of patients with head and
ture is low. Brown fat uptake is often identified by neck carcinoma treated with radiation have shown that
matching regions of fat attenuation on CT with the PET/ radiation-induced inflammation can be seen on the [18F]
CT fused images (Fig. 19.7). FDG PET/CT images for 2–3 months after the end of
3. Bilateral uptake in vocal cords can be observed when the treatment [40, 41]. Waiting 2 or 3 months following
patient speaks during the uptake interval after [18F]FDG completion of radiation therapy before obtaining a PET/
administration, with greater uptake of the healthy cord as a CT scan is clinically appropriate, as patients rarely
compensation for a contralateral recurrent nerve paralysis. develop clinical problems in the first 3 months after
4. The lymph node anatomy of the head and neck is com- treatment (Fig. 19.8).
plex and must be well known to correctly interpret the 11. In patients who have undergone surgery, uptake depends
findings of PET/CT. on the extent of surgery, the presence of infection/
5. Increased uptake of [18F]FDG can be seen in granulation inflammation in the wound and how long after surgery
tissue (e.g. healing wounds), infections and other inflam- images are acquired. For example, there are few visible
matory processes as well as in benign salivary gland signs of a mediastinoscopy after 10 days, but sternotomy
tumours, such as the pleomorphic adenoma and signs will remain visible for months. Following surgery,
Warthin’s tumour, which may present high uptake of [18F]FDG PET/CT should be delayed for at least 6 weeks
[ F]FDG. A detailed description of pitfalls and situa-
18
due to postsurgical inflammation if the scan is primarily
tions that can lead to false-positive (benign processes being done to assess the surgical field.
that can show [ F]FDG uptake) or false-negative [ F] 12. Due to the high physiological [18F]FDG uptake in the
18 18
FDG PET/CT interpretation has been published [36]. brain, [18F]FDG PET/CT is of limited value for detec-
6. In neoplastic processes, tracer accumulation in anatomic tion of brain metastases. Therefore, [18F]FDG PET/CT
abnormalities seen on CT scan or other imaging may be is generally not used for the primary detection or exclu-
particularly significant. When appropriate, the report sion of brain metastases.
should correlate PET/CT findings with those of other
diagnostic tests. These tests should be reinterpreted in the
context of all available data imaging and clinical data. 19.6.3 Documentation and Reporting
7. For response assessment, the images should be viewed
over the same dynamic grey scale or colour scale range. The SNMMI has recently published reporting recommenda-
In clinical trials, criteria for visual analysis should be tions for oncological [18F]FDG PET/CT imaging [42].
defined a priori within the study protocol [37]. Abnormalities of immediate clinical importance should be
8. SUV (normalized to body weight) measurement is directly or verbally communicated to the appropriate health-
widely used in clinical studies in addition to visual care provider if a delay in treatment might result in signifi-
assessments. SUV normalized to lean body mass (LBM) cant morbidity. It is recommended to report the following
is a recommended quantitative measure of [18F]FDG contents:
Fig. 19.6 A 82-year-old man in surveillance after treatment of tabolism in vestibular/gingival mucosa due to inflammatory processes
T3N1M0 carcinoma of oral cavity. Coronal and transaxial slices of CT, caused by bad oral hygiene (thin arrow). There is also bilateral uptake
[18F]FDG PET and fused PET/CT images show an intense hyperme- in digastric musculature that must be considered as physiological but
could be confused with adenopathies (thick arrow)
1. Study identification: full name of the patient, medical 3. Procedure description: blood glucose level before [18F]
record number and date of birth. The protocol name of the FDG administration should be documented. Study-
examination should also be included, as well as the date specific information should include the imaging agent,
and time of its performance. the amount of injected activity in megabecquerels and/or
2. Clinical information: age, gender, weight, height, reason millicuries, the route of administration and the date and
for referral and the specific question to be answered. If time of administration. The anatomical site of administra-
known, the diagnosis and a brief treatment history should tion is optional but should be recorded. Registration of
be provided. The results of relevant diagnostic tests and Time interval between administration of [18F]FDG and
prior imaging findings should be summarized. the start time of the acquisition is optional too, but could be
Fig. 19.7 Volumetric display (MIP), coronal and transaxial slices of Hypermetabolic areas are located in low-density areas visualized in CT,
CT, [18F]FDG PET and fused CT/PET images showing a very intense corresponding to fat tissue
and symmetrical [18F]FDG uptake in cervicothoracic brown fat.
Fig. 19.8 A 65-year-old man with T4N2M0 carcinoma of the larynx area of heterogeneous hypermetabolism (arrow) in glottic region that
treated with conventional radiotherapy 2 months prior to the scan. can disturb an accurate assessment of residual tumour activity after
Sagittal slices of CT, [18F]FDG PET and fused images show an extended treatment
important in the evaluation of inflammatory processes. they are interpreted in the context of other imaging exam-
The part of the body that was covered should be described inations (CT, PET/CT, MRI, etc.) and clinical data.
from the start to the end point (head and neck and whole 5. Assessment of response to therapy: if an [18F]FDG PET/
body acquisition). CT study is performed for assessing response to therapy,
4. Comparisons with previous examinations and reports:
the extent and the intensity of the [18F]FDG uptake should
PET/CT studies are more valuable when correlated with be documented and compared to prior measurements, if
previous diagnostic CT, previous PET, previous PET/CT, available. Examples of criteria for therapy response with
previous MRI and all appropriate imaging studies and [18F]FDG as a metabolic biomarker have been suggested
clinical data that are relevant. When PET/CT is performed by the European Organisation for Research and Treatment
for monitoring therapy, a comparison of the extent and of Cancer [37]. In 2009, Wahl et al. suggested the so-
intensity of uptake may be summarized as metabolic pro- called PERCIST criteria for assessing solid tumour
gressive disease, metabolic stable disease, metabolic par- response [38].
tial response or metabolic complete response. 6. Summary and diagnosis/impression: clearly identify the
5. CT examination optimized for diagnosis: if the CT scan study as normal or abnormal. The question asked in the
was requested and performed as a diagnostic examina- study requisition should be directly addressed. If possi-
tion, then the CT component of the study could be ble, a definite diagnosis or, when appropriate, a differen-
reported separately, if necessary, to satisfy regulatory, tial diagnosis should be stated. Whenever possible this
administrative or reimbursement requirements. should provide a staging assessment (TNM or other) stat-
6. Dosimetric parameters: it should be included as required ing whether there are categories of uncertainty. If appro-
by regulations. The report should state whether CT with priate, repeat examinations and/or additional examinations
or without CT contrast agent was used for attenuation should be recommended to clarify or confirm findings.
correction. For therapy evaluation, serial studies should be compared,
using visual and/or semiquantitative assessment as
appropriate.
19.6.4 Description of the Findings
It is good practice to provide a structured report with concise Key Learning Points
concluding statements intended to answer the specific clini- • Display of images from an [18F]FDG PET/CT scan
cal question(s) posed, if possible. Nevertheless, there is a in patients with head and neck cancers follows the
great variation in the style of reporting. Quality of the study general format and modalities as for other regions
should be described, for example, limited because of motion, of the body.
muscular uptake or hyperglycaemia. • Particular attention should be paid to evaluation of
the non-CT attenuation-corrected images.
1. CT-related artefacts should be mentioned, such as metal- • Physiological uptake of [18F]FDG with variable
lic artefacts and other information on large patient body intensity can be observed in normal structures of
habitus, when the quality of the study is affected. the head and neck including the lymphatic
2. Description of the location, the extent and the intensity Waldeyer’s ring; major salivary glands; muscles of
(SUV and/or SUL) of pathological [18F]FDG accumula- the floor of the mouth; ocular extrinsic, cervical or
tion related to normal tissue. Description of relevant find- masticatory muscles; brown fat; and vocal cords.
ings on CT and their relationship to pathological [18F] • Good knowledge of the complex anatomy of lymph
FDG accumulation. [18F]FDG uptake may be reported as nodes of the head and neck is mandatory to cor-
mild, moderate or intense and compared to the back- rectly interpret the [18F]FDG PET/CT scan.
ground uptake in liver parenchyma. • The possibility of increased [18F]FDG uptake in
3. Limitations: confounding factors that might influence the inflamed/healing tissues (such as early after surgery
sensitivity or specificity of the [18F]FDG PET/CT study or radiotherapy) as well as in some benign tumours
may be mentioned, such as small lesions (partial volume must adequately be taken into account when inter-
effect), inflammatory changes, muscle activity, high preting an [18F]FDG PET/CT scan of the head and
blood glucose levels at the time of injection, para-vascular neck.
infiltration of [18F]FDG or tissue injections. • [18F]FDG accumulation at tumour sites must be
4. Complementary information: comparison with previous adequately correlated with anatomic abnormalities
examinations should be part of the [18F]FDG PET/CT observed in the CT component of the PET/CT scan.
report. [18F]FDG PET/CT studies are more valuable if
hypoxia, DNA damage repair and apoptosis, all treatment

• Standardized protocols must be implemented when resistance mechanisms. The anti-EGFR antibody cetuximab
re-evaluating a patient for assessment of response to has been labelled with various radionuclides and has been
therapy. tested as an imaging biomarker in several HNSCC models.
• Protocols for assessment of response to treatment These studies suggest that EGFR-targeting tracers can be
must take into account adequate time intervals after used to monitor EGFR receptor expression in HNSCC and
surgery, chemotherapy and/or radiotherapy. have the potential to noninvasively monitor cetuximab treat-
• Findings of the [18F]FDG PET/CT scan should be ment and steer individualized treatment regimens [46].
reported following recommended standard formats, The indications for PET imaging have recently been
with a final statement on diagnosis/impression extended in parallel with the development of new tracers to
addressing the specific clinical issues raised by the specifically image phaeochromocytomas and paraganglio-
referring physician/specialist. mas arising in the head and neck region. Because of unique
tumour characteristics (e.g. the presence of cell membrane
and intracellular vesicular norepinephrine transporters),
19.7 Future Perspectives these tumours are especially suited to be imaged by means of
specific functional imaging approaches. PET/CT imaging
[18F]FDG PET is now the gold standard for noninvasive with 18F-DOPA or 68Ga-DOTA peptides has become the first-
functional imaging in oncology. In head and neck tumours, line imaging modality in the evaluation of cervical paragan-
[18F]FDG PET is not recommended for detection of the pri- gliomas, regardless of the genetic background [47, 48].
mary tumour, and its value for metastatic lymph nodes is still
a matter of debate. With regard to staging of the primary
tumour, [18F]FDG PET may influence the treatment decision 19.7.2 PET/MRI
if distant metastases or second primary tumours are detected.
With the intention of solving the limitations of the technique, Integrated PET/MRI scanners combine anatomic with func-
as well as of implementing new indications, different meth- tional imaging and may have a specific impact on the staging
odologies are currently being developed. and treatment of head and neck cancer. Considering that
MRI is the diagnostic imaging modality of choice in these
tumour sites, integrated PET/MRI scanners may further
19.7.1 New Oncological Tracers improve the potential benefits of isolated PET or MRI diag-
nostic tools. Advantages include better identification of peri-
Hypoxia tracers for radiotherapy planning in head and neck neural tumour spread as well as assessment of infiltration in
cancer image biologic tumour characteristics reflecting radi- the epipharynx, tongue and larynx. In addition, dynamic
ation resistance mechanisms and thus offer potential for tai- MRI studies such as dynamic contrast-enhanced MRI and
lored radiation therapy. Hypoxia is a feature of many solid blood oxygen level-dependent MRI, as well as MR spectros-
tumours and in particular squamous cell carcinomas of the copy, may add complementary functional information. A
head and neck. 18F-fluoromisonidazole (18F-MISO), a nitro- trend that is currently emerging is the use of multiparametric
imidazole PET tracer that is reduced and bound to cell con- imaging with the PET-MRT to the more accurate noninva-
stituents under hypoxic conditions, is being extensively used sive tumour characterization [49].
for the detection of hypoxia in head and neck tumours. Apart Nevertheless, PET/MR imaging does not offer significant
from its prognostic value, Rischin et al. published data sup- additional information in initial staging of squamous cell
porting the predictive value of 18F-MISO PET [43]. carcinoma of the head and neck when compared with stan-
18
F-fluoroerythronitroimidazole, 18F-fluoroazomycin arabi- dard MR imaging. In patients with suspected tumour recur-
noside and 18F-2-(2-nitroimidazol-1-yl)-N-(3,3,3-trifluoro- rence, [18F]FDG PET/MR imaging has higher sensitivity
propyl)-acetamide (18F-EF3) are members of a new than MR imaging, although its accuracy is equivalent to the
generation of nitroimidazoles. These new tracers showed a accuracy of [18F]FDG PET/CT [50].
higher and more heterogeneously distributed tracer uptake in
tumours than in adjacent neck muscle and have proven fea-
sible and of sufficient quality for clinical use in patients with 19.7.3 Digital PET/CT
head and neck cancer [44, 45].
Another important imaging target for this application in The increase in effective sensitivity and improved spatial
HNSCC patients is the epithelial growth factor receptor resolution led to an improved visibility of small lesions,
(EGFR). It steers the pathways related to proliferation, which is not only important for detection of lesions and
metastases in [18F]FDG PET/CT. In case of the digital PET 4. Hettmann A, Demcsák A, Decsi G, Bach Á, Pálinkó D, Rovó L,
et al. Infectious agents associated with head and neck carcinomas.
scanner, the digital Silicon Photomultipliers (SiPMs) are Adv Exp Med Biol. 2016;897:63–80.
capable of detecting and processing single scintillation pho- 5. Cammaroto G, Quartuccio N, Sindoni A, Di Mauro F, Caobelli F,
tons because their elements match the size of the scintillator Young AIMN Working Group. The role of PET/CT in the manage-
crystal elements and they incorporate electronics to achieve ment of patients affected by head and neck tumors: a review of the
literature. Eur Arch Otorhinolaryngol. 2016;273:1961–73.
a one-to-one relation between the scintillator crystal ele- 6. Oyen WJG, Marres HAM, Kaanders JHAM. Progress in nuclear
ments and the digital photomultipliers. This design, espe- medicine procedures in head and neck oncology. Q J Nucl Med Mol
cially in regions with small structures as the head and neck, Imaging. 2011;55:485–6.
results in an improved spatial and timing resolution and rela- 7. Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE,
Brookland RK, et al. The Eighth Edition AJCC Cancer Staging
tively high maximum count rates, more than the block design Manual: continuing to build a bridge from a population-based to a
of analogue PMT-based scanners [51]. more “personalized” approach to cancer staging. CA Cancer J Clin.
2017;67:93–9.
8. Lydiatt WM, Patel SG, O’Sullivan B, Brandwein MS, Ridge JA,
Migliacci JC, et al. Head and neck cancers-major changes in the
Key Learning Points American Joint Committee on Cancer eighth edition cancer staging
• The clinical usefulness of new PET imaging agents manual. CA Cancer J Clin. 2017;67:122–37.
9. Al-Ibraheem A, Buck A, Krause BJ, Scheidhauer K, Schwaiger
is being explored in patients with head and neck M. Clinical applications of FDG PET and PET/CT in head and neck
cancers, with particular attention to hypoxia agents cancer. J Oncol. 2009;2009:1–13.
and to emerging targets aiming at tailoring treat- 10. Yongkui L, Jian L, Wanghan, Jingui L. 18FDG-PET/CT for the
ments on an individual patient’s basis. detection of regional nodal metastasis in patients with primary head
and neck cancer before treatment: a meta-analysis. Surg Oncol.
• Non-[18F]FDG tracers possibly employed in 2013;22:e11–6.
patients with head and neck cancers include 11. Kyzas PA, Evangelou E, Denaxa-Kyza D, Ioannidis JPA.

18
F-DOPA and 68Ga-DOTA peptides for 18
F-fluorodeoxyglucose positron emission tomography to evalu-
paragangliomas. ate cervical node metastases in patients with head and neck
squamous cell carcinoma: a meta-analysis. J Natl Cancer Inst.
• PET/MRI is emerging as an especially useful 2008;100:712–20.
approach to patients with head and neck cancers, 12. Yoo J, Henderson S, Walker-Dilks C. Evidence-based guideline
due to its possibility to combine the detailed ana- recommendations on the use of positron emission tomography
tomic information provided by MRI with functional imaging in head and neck cancer. Clin Oncol. 2013;25:e33–66.
13. Lonneux M, Hamoir M, Reychler H, Maingon P, Duvillard

molecular imaging provided by PET. C, Calais G, et al. Positron emission tomography with [18F]
• The diagnostic potential of MRI integrated with Fluorodeoxyglucose improves staging and patient management
PET is enhanced by the use of the multifaceted in patients with head and neck squamous cell carcinoma: a multi-
modalities of MR imaging (dynamic contrast center prospective study. J Clin Oncol. 2010;28:1190–5.
14. Adelstein D, Gillison ML, Pfister DG, Spencer S, Adkins D, Brizel
enhancement, evaluation of blood oxygen levels, DM, et al. NCCN guidelines insights: head and neck cancers, ver-
spectroscopy, etc.). sion 2.2017. J Natl Compr Cancer Netw. 2017;15:761–70.
• The most recent technological advances in PET 15. Ciernik IF, Dizendorf E, Baumert BG, Reiner B, Burger C, Davis
imaging that aim at improving resolution of clinical JB, et al. Radiation treatment planning with an integrated positron
emission and computer tomography (PET/CT): a feasibility study.
scanners are expected to be especially useful for Int J Radiat Oncol Biol Phys. 2003;57:853–63.
exploring small-size anatomic structures as those of 16. Moule RN, Kayani I, Moinuddin SA, Meer K, Lemon C, Goodchild
the head and neck. K, et al. The potential advantages of 18FDG PET/CT-based target
volume delineation in radiotherapy planning of head and neck can-
cer. Radiother Oncol. 2010;97:189–93.
17. Daisne J-F, Duprez T, Weynand B, Lonneux M, Hamoir M,

Reychler H, et al. Tumor volume in pharyngolaryngeal squa-
References mous cell carcinoma: comparison at CT, MR imaging, and
FDG PET and validation with surgical specimen. Radiology.
1. Escott EJ. Role of positron emission tomography/computed tomog- 2004;233:93–100.
raphy (PET/CT) in head and neck cancer. Radiol Clin North Am. 18. Koshy M, Paulino AC, Howell R, Schuster D, Halkar R, Davis
2013;51:881–93. LW. F-18 FDG PET-CT fusion in radiotherapy treatment planning
2. Blot WJ, McLaughlin JK, Winn DM, Austin DF, Greenberg RS, for head and neck cancer. Head Neck. 2005;27:494–502.
Preston-Martin S, et al. Smoking and drinking in relation to oral 19. Dietl B, Marienhagen J, Kühnel T, Schreyer A, Kölbl O. The impact
and pharyngeal cancer. Cancer Res. 1988;48:3282–7. of FDG-PET/CT on the management of head and neck tumours: the
3. Winquist E, Agbassi C, Meyers BM, Yoo J, Chan KKW. Head radiotherapist’s perspective. Oral Oncol. 2008;44:504–8.
and Neck Disease Site Group. Systemic therapy in the curative 20. Abramyuk A, Appold S, Zöphel K, Baumann M, Abolmaali

treatment of head and neck squamous cell cancer: a systematic N. Modification of staging and treatment of head and neck can-
review. J Otolaryngol Head Neck Surg. 2017;46(1):29. https://doi. cer by FDG-PET/CT prior to radiotherapy. Strahlenther Onkol.
org/10.1186/s40463-017-0199-x. 2013;189:197–201.
21. Mak D, Corry J, Lau E, Rischin D, Hicks RJ. Role of FDG-PET/ 37.

Quantitative Imaging Biomarkers Alliance. FDG-PET/CT
CT in staging and follow-up of head and neck squamous cell carci- Technical Committee. FDG-PET/CT as an imaging biomarker mea-
noma. Q J Nucl Med Mol Imaging. 2011;55:487–99. suring response to cancer therapy, version 1.05, Publicly Reviewed
22. Isles MG, McConkey C, Mehanna HM. A systematic review
Version. Quantitative Imaging Biomarkers Alliance. 2013. http://
and meta-analysis of the role of positron emission tomography www.rsna.org/uploadedfiles/rsna/content/science_and_education/
in the follow up of head and neck squamous cell carcinoma fol- qiba/qiba_fdg-pet_profile_v105_publicly_reviewed_version_
lowing radiotherapy or chemoradiotherapy. Clin Otolaryngol. final_11dec2013.pdf. Accessed 23 Nov 2014.
2008;33:210–22. 38. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to
23. Cheung PKF, Chin RY, Eslick GD. Detecting residual/recur-
PERCIST: evolving considerations for PET response criteria in
rent head neck squamous cell carcinomas using PET or PET/ solid tumors. J Nucl Med. 2009;50(Suppl 1):122S–50.
CT. Otolaryngol Neck Surg. 2016;154:421–32. 39. Boellaard R, O’Doherty MJ, Weber WA, et al. FDG PET and PET/
24. Moeller BJ, Rana V, Cannon BA, Williams MD, Sturgis
CT: EANM procedure guidelines for tumour PET imaging: version
EM, Ginsberg LE, et al. Prospective risk-adjusted [18F] 1.0. Eur J Nucl Med Mol Imaging. 2010;37:181–200.
Fluorodeoxyglucose positron emission tomography and computed 40. Andrade RS, Heron DE, Degirmenci B, et al. Posttreatment assess-
tomography assessment of radiation response in head and neck can- ment of response using FDG-PET/CT for patients treated with
cer. J Clin Oncol. 2009;27:2509–15. definitive radiation therapy for head and neck cancers. Int J Radiat
25. Lee JC, Kim JS, Lee JH, Nam SY, Choi S-H, Lee S-W, et al. Oncol Biol Phys. 2006;65:1315–22.
F-18 FDG-PET as a routine surveillance tool for the detection of 41. Kawabe J, Higashiyama S, Yoshida A, Kotani K, Shiomi S. The role
recurrent head and neck squamous cell carcinoma. Oral Oncol. of FDG PET-CT in the therapeutic evaluation for HNSCC patients.
2007;43:686–92. Jpn J Radiol. 2012;30:463–70.
26. Issing WJ, Taleban B, Tauber S. Diagnosis and management of 42. Niederkohr RD, Greenspan BS, Prior JO, et al. Reporting guid-
carcinoma of unknown primary in the head and neck. Eur Arch ance for oncologic 18F-FDG PET/CT imaging. J Nucl Med.
Otorhinolaryngol. 2003;260:436–43. 2013;54:756–61.
27. Lee JR, Kim JS, Roh J-L, Lee JH, Baek JH, Cho K-J, et al. 43. Rischin D, Hicks RJ, Fisher R, et al. Prognostic significance of
Detection of occult primary tumors in patients with cervical metas- [18F]-misonidazole positron emission tomography-detected tumor
tases of unknown primary tumors: comparison of 18F-FDG PET/ hypoxia in patients with advanced head and neck cancer randomly
CT with contrast-enhanced CT or CT/MR imaging — prospective assigned to chemoradiation with or without tirapazamine: a sub-
study. Radiology. 2015;274:764–71. study of Trans-Tasman Radiation Oncology Group Study 98.02. J
28. Boellaard R, Delgado-Bolton R, Oyen WJG, Giammarile F, Tatsch Clin Oncol. 2006;24:2098–104.
K, Eschner W, et al. FDG PET/CT: EANM procedure guidelines 44. Troost EGC, Schinagl DAX, Bussink J, Boerman OC, van der
for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging. Kogel AJ, Oyen WJG, Kaanders JHAM. Innovations in radiother-
2015;42:328–54. apy planning of head and neck cancers: role of PET. J Nucl Med.
29. Zanotti-Fregonara P, Laforest R, Wallis JW. Fetal radiation dose 2010;51:66–76.
from 18F-FDG in pregnant patients imaged with PET, PET/CT, and 45. Lehtiö K, Oikonen V, Grönroos T, Eskola O, Kalliokoski K,

PET/MR. J Nucl Med. 2015;56:1218–22. Bergman J, et al. Imaging of blood flow and hypoxia in head and
30. Delbeke D, Coleman RE, Guiberteau MJ, Brown ML, Royal HD, neck cancer: initial evaluation with [15O]H2O and [18F]fluoroery-
Siegel BA, et al. Procedure guideline for tumor imaging with 18F- thronitroimidazole PET. J Nucl Med. 2001;42:1643–52.
FDG PET/CT 1.0. J Nucl Med. 2006;47:885–95. 46. van Dijk LK, Boerman OC, Kaanders JH, Bussink J. PET imaging
31. Wong TZ, Fras IM. PET/CT protocols and practical issues for in head and neck cancer patients to monitor treatment response:
the evaluation of patients with head and neck cancer. PET Clin. a future role for EGFR-targeted imaging. Eur J Nucl Med Mol
2007;2:413–21. Imaging. 2017;44(Suppl 1):4–16.
32. The Medical Electrical Safety Standards, IEC 60601-1/EN
47. Havekes B, King K, Lai EW, Romijn JA, Corssmit EP, Pacak
60601-1. K. New imaging approaches to phaeochromocytomas and paragan-
33. The Medical ECM Standards, IEC 60601-1-2, EN 60601-1-2. gliomas. Clin Endocrinol. 2010;72:137–45.
34. Quantitative Imaging Biomarkers Alliance. Quantitative FDG-PET 48. Guichard JP, Fakhry N, Franc J, Herman P, Righini CA, Taieb
Technical Committee. UPICT oncology FDG-PET CT protocol. D. Morphological and functional imaging of neck paragangliomas.
http://qibawiki.rsna.org/index.php?title=FDG-PET_tech_ctte. Eur Ann Otorhinolaryngol Head Neck Dis. 2017;134:243–8.
Accessed 23 Nov 2014. 49. Judenhofer MS, Wehrl HF, Newport DF, et al. Simultaneous PET-
35. ACR Guidelines and Standards Committee. ACR-SPR practice
MRI: a new approach for functional and morphological imaging.
parameter for performing FDG-PET/CT in oncology. American Nat Med. 2008;14:459–65.
College of Radiology. 2014. http://www.acr.org/~/media/71B7467 50. Platzek I. 18F-Fluorodeoxyglucose PET/MR imaging in head and
80F934F6D8A1BA5CCA5167EDB.pdf. Accessed 23 Nov 2014. neck cancer. PET Clin. 2016;11:375–86.
36. Cook GJ, Fogelman I, Maisey MN. Normal physiological and 51. van der Vos CS, Koopman D, Rijnsdorp S, Arends AJ, Boellaard R,
benign pathological variants of 18-fluoro-2-deoxyglucose positron- van Dalen JA, et al. Quantification, improvement, and harmoniza-
emission tomography scanning: potential for error in interpretation. tion of small lesion detection with state-of-the-art PET. Eur J Nucl
Semin Nucl Med. 1996;26:308–14. Med Mol Imaging. 2017;44(Suppl 1):4–16.
Radionuclide Imaging of Cardiovascular
Disease 20
Matteo Bauckneht, Flavia Ticconi, Roberta Piva,
Riemer H. J. A. Slart, Alberto Nieri, Silvia Morbelli,
Paola Anna Erba, Cecilia Marini, H. William Strauss,
and Gianmario Sambuceti
Contents
20.2 Myocardial Perfusion Imaging of Coronary Artery Disease 450
20.2.1 Pathophysiologic Basis of Myocardial Ischemia 450
20.2.2 Myocardial Perfusion Imaging: Clinical Indications 452
20.2.3 MPI: Patient Preparation and General Information 454
20.2.4 MPI: Stressors Characteristics 455
20.2.5 SPECT-MPI: Radiopharmaceuticals 464
20.2.6 PET-MPI: Radiopharmaceuticals 466
20.2.7 MP Image Interpretation 468
20.2.8 Multimodality Imaging of Coronary Atherosclerotic Disease 470
20.3 Multimodality Imaging in Cardiovascular Infections 471
20.3.1 Clinical Background on Infections of the Cardiovascular System 472
20.3.2 Radionuclide Imaging Techniques in Cardiac Infections 474
20.3.3 Radionuclide Imaging of Prosthetic Valve Endocarditis 478
20.3.4 Radionuclide Imaging of Native Valve Endocarditis 478
20.3.5 Radionuclide Imaging of CIED Infections 480
20.3.6 Radionuclide Imaging of LVAD-Related Infections 481
20.3.7 Extracardiac Workup of Patients with Infective Endocarditis 481
20.3.8 Clinical Background on Vascular Prosthetic Infections 481
M. Bauckneht P. A. Erba
Nuclear Medicine, IRCCS Policlinico San Martino, Genoa, Italy Regional Center of Nuclear Medicine, Department of Translational
F. Ticconi · R. Piva · A. Nieri
Nuclear Medicine, Department of Health Sciences,
University of Genoa, Genoa, Italy C. Marini
Nuclear Medicine, Department of Health Sciences,
Riemer H. J. A. Slart
University of Genoa, Genoa, Italy
Biomedical Photonic Imaging Group, University of Twente,
Enschede, The Netherlands CNR Institute of Bioimages and Molecular Physiology, Milan,
Italy
S. Morbelli · G. Sambuceti (*)
Nuclear Medicine, IRCCS Policlinico San Martino, Genoa, Italy H. W. Strauss
Molecular Imaging and Therapy Service, Memorial Sloan
Nuclear Medicine, Department of Health Sciences, University of
Kettering Cancer Center, New York, NY, USA
Genoa, Genoa, Italy
e-mail: sambuceti@unige.it

450 M. Bauckneht et al.
20.4 C
linical Indications, Technique, and Principles of Noninfectious, Non-atherosclerotic
Cardiovascular Diseases 485
20.4.1 Clinical Background on Myocarditis 485
20.4.2 Clinical Background on Acute Pericarditis 486
20.4.3 Clinical Background on Cardiac Sarcoidosis 488
20.4.4 Clinical Background on Cardiac Amyloidosis 489
20.5 Heart Failure 491
20.6 Large Vessels Vasculitis 492
References 493
Learning Objectives
• Define the rationale underlying the use of radionuclide 20.2 M
yocardial Perfusion Imaging
methods in cardiovascular diseases. of Coronary Artery Disease
• Understand the clinical conditions in which nuclear medi-
cine can or should be used. 20.2.1 Pathophysiologic Basis of Myocardial
• Learn the clinical background and the clinical relevance Ischemia
of information obtained with radionuclide imaging.
• Learn the techniques and methodology for radionuclide The hallmark of coronary artery disease (CAD) is the pres-
imaging of the cardiovascular system. ence of focal obstructions of the major epicardial coronary
• Understand the criteria for image interpretation. arteries. The mechanism underlying the formation of coro-
• Understand the rationale for combining radionuclide nary plaques is, in virtually all patients, atherosclerosis. This
techniques with other imaging modalities. process begins in childhood and proceeds with discrete
phases characterized by a periodic shift between subintimal
inflammation leading to enlargement of the lesion and heal-
ing of the lesion [3]. This process is repeated numerous
20.1 Introduction times, often with rupture of the lesion, leading to formation
of a thrombus. Fortunately, most ruptures and thrombus for-
Multiple imaging modalities are clinically available for mation are clinically silent. The lesion heals, often with a
evaluation of global and regional cardiac function, determi- reduction in the size of the lumen. Depending on the patient
nation of regional myocardial perfusion, and evaluation of and his/her risk factors (e.g., hyperlipidemia, hypertension,
coronary anatomy, including SPECT with CT, PET with and diabetes), the lesion may recur, or if the thrombus is
CT, multi-detector computed tomography (MDCT), mag- large enough to suddenly occlude the vessel, resulting in a
netic resonance imaging (MRI), and ultrasound. clinical event.
Emblematic of this revolution is the capability of cardiac The thrombus is caused by the release of pro-thrombotic
MDCT to integrate evaluation of coronary artery anatomy cytokines from the inflamed plaque [4]. This sequence of
with noninvasive assessment of myocardial perfusion and pathophysiological processes suggests that a significant frac-
contractile function [1]. MDCT still needs to be technically tion of myocardial infarcts with ST-segment elevation
optimized, particularly to identify small changes in myo- (STEMI) occur in coronary arteries that were not severely
cardial perfusion, which can be done reproducibly with narrowed before the event. From an epidemiological point of
radionuclide techniques [2]. On the other hand, despite view, non-critical lesions are obviously more frequent than
proven clinical efficacy in the detection of CAD (primarily critically narrowed vessels and are more likely to contain the
based on the presence of calcium in the coronary arteries), “culprit” lesion. This hypothesis was confirmed by the angi-
MDCT imaging performs suboptimally in the noninvasive ographic evaluation of patients treated with systemic
characterization of multivessel CAD. On the other hand, thrombolysis (usually streptokinase or tissue plasminogen
the specificity of nuclear imaging is improved by attenua- activator) in the late 1980s [5], as well as by the lack of con-
tion correction. Accordingly, multimodality SPECT/CT or cordance between the site of ischemia (documented by myo-
PET/CT approach is the current standard for evaluating cardial perfusion scan) and subsequent myocardial
myocardial perfusion. infarction [6]. This model is relatively less adequate for the
20 Radionuclide Imaging of Cardiovascular Disease 451
pathophysiology of non-ST-segment elevation myocardial The ischemic cascade is triggered by an inadequate

infarction (non-STEMI), in which biological inflammation increase in perfusion in response to an increase in myocar-
involves plaques causing severe obstructions [4]. dial oxygen demand, and thus its presence, location, and
Stable angina pectoris, on the other hand, reflects the extent are best identified by the evaluation of myocardial
hemodynamic impact of a stable atheroma causing a critical perfusion. The physics of single-photon emission and the
stenosis (usually occupying 75–90% of the lumen). Gould decrease in myocardial extraction with increasing flow of
et al. [7] define coronary plaques (either a single stenosis or soluble tracers such as the 99mTc-labelled perfusion tracers
multiple stenoses) characterized by a measurable hemody- do not permit a reliable calculation of blood flow per unit
namic impact limiting the increase in blood flow through the mass of myocardium. Nevertheless, they do provide an accu-
vessel causing ischemia. The regulation of coronary vasomo- rate description of flow distribution throughout the left and
tor tone and thus of coronary blood flow and capillary pres- right ventricular myocardium. Ischemia is identified on myo-
sure occurs in each vessel. Restriction imposes a resistance cardial SPECT and PET images by comparison of regional
to flow and thus causes a pressure gradient that is propor- myocardial perfusion with tracer injection with the patient at
tional to flow through it and to the degree of lumen obstruc- rest to that with the patient at stress. A relative decrease of
tion. This phenomenon is particularly relevant for the left tracer concentration in an area of myocardium on the stress-
ventricle (LV) for two main reasons: (1) the left ventricular injected images, which improves on the rest-injected images,
myocardium performs fivefold more work than the right defines an area of stress-induced ischemia. Classical litera-
ventricle and (2) the muscle in the LV myocardium develops ture assumes that oxygen consumption and thus coronary
a pressure able to collapse endocardial microvessels during blood flow are homogeneously distributed throughout the
systole and thus to stop coronary blood flow. This phasic pat- different myocardial walls. Reversible perfusion defects
tern of flow results in near maximal myocardial oxygen induced by exercise or by inotropic stimulation (e.g., by
extraction through the left ventricular coronary circulation at dobutamine infusion) can be considered as markers of a tran-
baseline and asks that every increase in myocardial oxygen sient imbalance between flow demand and supply.
demand must be met by a parallel increase in coronary blood In the large number of patients in whom exercise cannot
flow. As a consequence, myocardial ischemia is mostly a left be performed (e.g., patients with orthopedic or neurological
ventricular phenomenon and occurs because of the following impairment), pharmacological vasodilation using dipyri-
sequence of events. damole, adenosine, or the A2 adenosine receptor agonist
Under resting conditions, the relatively low-flow need regadenoson is commonly employed. This procedure
(approximately 0.5–1 mL × g−1 × min−1) is preserved by the increases coronary blood flow without significantly increas-
autoregulatory vasodilation of microvessels triggered by ing myocardial oxygen consumption. Accordingly, an area
the pressure drop [8, 9]. During exercise, myocardial oxy- of decreased perfusion seen on a vasodilator stress test indi-
gen consumption increases. Since this increased demand cates that maximal perfusion is reduced in an area distal to a
cannot be met by increased oxygen extraction, it is met by stenosis or a region of inappropriate vasoconstriction com-
an increase in coronary blood flow. The flow increase pared to adjacent areas that increase flow due to the vasodila-
occurs by microcirculatory vasodilation. In the presence of tor. Accordingly, reversible perfusion defects under
a hemodynamically significant flow-limiting stenosis (typi- pharmacological vasodilation should be interpreted as a sign
cally a narrowing of 50–75% of the vessel lumen; the length of regional reduction in coronary flow reserve. However, a
of the narrowing is also a factor limiting flow) the post- profound vasodilation can actually induce ischemia, when
stenotic pressure is reduced, reducing oxygen delivery to post-stenotic coronary pressure is decreased leading to col-
the myocardium distal to the stenosis. This “hypotension” lapse of endocardial microvessels (transmural steal) or to
inevitably results in an inadequate flow increase and is fur- impair flow supply by collateral branches. This phenomenon
ther aggravated by the occurrence of a paradoxical vaso- usually identifies severe coronary obstructions and is associ-
constriction [10]. The resulting oxygen debt causes ated with ST-segment shift and/or wall motion abnormality,
contractile impairment, the possible loss of membrane but is not usually associated with chest pain.
polarization (causing ST-segment depression or elevation Regardless of the mechanism of stress, the location, sever-
on ECG), and, ultimately, the development of angina. ity, and extent of perfusion defects can be quantified by
Myocardial perfusion imaging identifies the location and inspection or by commercially available software. The extent
severity of ischemia and the resultant impairment of and severity of decreased perfusion is usually expressed as
contractile function [6]. the summed stress score. When applied to rest-injected
p erfusion images, the summed rest score represents an index

of irreversibly damaged myocardium. The difference between • The physics of single-photon emission and the
the summed rest score and the summed stress score provides kinetic features of 99mTc-labelled tracers provide an
an index of jeopardized myocardium, which has been used as accurate description of flow distribution in the left
a prognostic indicator [11]. However, a very severe stenosis ventricular myocardium.
can reduce resting flow to a level that impairs contractile • Ischemia causes a relative decrease of tracer con-
function. The myocardium responds to very low flow by centration in the area of decreased perfusion com-
reducing or eliminating contractile function, allowing the pared to normally perfused myocardium.
myocyte to remain viable but not functional, a condition • The region of decreased perfusion is not seen on the
called hibernating myocardium. Perfusion imaging alone rest-injected images.
cannot recognize this condition of resting hypoperfusion and • Gated-acquisition complements the perfusion marker
absent contractile function, duplicating the scintigraphic pat- by defining transient wall motion abnormality.
tern of myocardial scar. Accordingly, different approaches are
needed to detect this state, such as 201Tl redistribution or eval-
uation of myocardial metabolism using [18F]FDG [12]. 20.2.2 Myocardial Perfusion Imaging:
Using tracers that remain in the myocardium, myocardial Clinical Indications
perfusion imaging defines regional perfusion as it was at the
time of injection and contractile function as it is at the later The information provided by myocardial perfusion imaging
time of image acquisition. For rest imaging this temporal gap (MPI) [13–16] is of clinical value in the following
has a limited relevance since both perfusion and function are conditions:
assumed to remain stable over time. For stress acquisition the
delay is extremely relevant, since during the interval between • Detection of coronary heart disease: Diagnosis, location,
tracer injection and imaging there is time for recovery of con- extent, and severity of ischemia and/or necrosis.
tractile function, resulting in a scan that may show a large area • Evaluation of the functional significance of coronary
of ischemia with “normal” contractile function. The combined stenosis.
evaluation of post-stress and rest function complements the • Prognostic stratification after a myocardial infarction or
perfusion marker. The transient nature of contractile dysfunc- before high-risk surgery.
tion implies that a longer time gap between ischemia (injec- • Monitoring the effects of treatment (after revascularization,
tion) and acquisition of the post-stress images reduces the after correction of risk factors, after medical treatment).
degree of dysfunction and its prevalence. However, although • Assessment of myocardial viability.
early imaging can increase detection of contractile dysfunc-
tion, it also causes the challenge of separating myocardial 20.2.2.1 Appropriateness Criteria (AC)
radioactivity from tracer accumulation in the bowel. According to established appropriateness criteria for the use
of MPI [13–17], the nuclear physician has to consider both
symptoms and general conditions to identify both risk for
Key Learning Points future cardiovascular events and pretest probability to dis-
• The hallmark of coronary artery disease (CAD) is play CAD. In relation to symptoms, the following data need
the presence of focal obstructions of one or more to be evaluated:
major epicardial coronary arteries. Chest pain: the AC for SPECT-MPI defines a chest pain
• Clinically relevant conditions can be induced by syndrome as “any constellation of symptoms that the physi-
any coronary plaque with two major mechanisms cian believes may represent a complaint consistent with
related to sudden changes in its biological signaling obstructive coronary artery disease (CAD). Examples of
during the inflammatory period or to its effect on such symptoms include, but are not exclusive to, chest pain,
blood flow hemodynamics in its stable phase, chest tightness, burning, dyspnea, shoulder pain, and jaw
respectively. pain, and new ECG abnormalities.”
• The former mechanism reflects the so-called plaque Anginal equivalent: non-chest pain symptoms, such as
vulnerability or instability and is responsible for the dyspnea or worsening effort tolerance, that are felt to be con-
large majority of acute ischemic syndromes. sistent with CAD may also be considered to be an anginal
• The second mechanism reflects the hemodynamic equivalent.
impact of a critical stenosis, defined as any given In the absence of symptoms, MPI can be indicated when-
coronary plaque characterized by a measurable ever clinical conditions and present risk factors identify a
hemodynamic impact potentially leading to the measurable risk for future cardiovascular events (including
occurrence of ischemia. myocardial infarction or death over a given time period).
Risk for future events is usually classified as follows:
Low: defined by the age-specific risk level that is below • Intermediate pretest probability: 11–90% pretest proba-
average; in general, low risk will correlate with a 10-year bility of CAD.
absolute risk for coronary heart disease (CHD) less than 10%. • High pretest probability: >90% pretest probability of CAD.
Moderate: defined by the age-specific risk level that is
average or above average; in general, moderate risk will cor- MPI is considered appropriate in patients with intermedi-
relate with a 10-year absolute risk between 10% and 20%. ate and high probability of CAD, as well as in those subjects
High: a 10-year absolute risk greater than 20%. in whom CAD needs to be excluded and exercise ECG can-
The pretest probability of CAD for symptomatic (isch- not be performed due to inability to exercise or to an abnor-
emic equivalent) patients describes the pretest probability of mal baseline ECG. In patients previously submitted to X-ray
the presence of CAD, regardless of the risk for future events, evaluation of coronary calcification (CAC), MPI is consid-
and should be assessed in the presence of suspected symp- ered inappropriate when the Agatston score for coronary
toms. There are a number of algorithms dedicated to this pur- calcium is <100 and appropriate when >400, respectively.
pose and most often define this variable as: Patients in the intermediate range require a further defini-
tion of cardiovascular risk and thus can be considered
• Very low pretest probability: <5% pretest probability of CAD. candidates for MPI according to the clinical condition

• Low pretest probability: <10% pretest probability of CAD. (Figs. 20.1 and 20.2).
Fig. 20.1 Appropriateness
criteria for myocardial CHD risk
perfusion imaging according (Framingham risk)
to level of risk for coronary
heart disease (CHD) (modified
from [14])
Low risk Intermediate risk High risk
ECG
interpretable?
Yes No
Inappropriate Inappropriate Uncertain Appropriate
Ischemic equivalent
Acute Chronic
Definite PreTest
ACS Inappropriate probability
Low
Possible ECG Interpretable

and able to exercise?
Intermediate
High
Yes No
Appropriate Inappropriate Appropriate Appropriate
Fig. 20.2 Appropriateness criteria for myocardial perfusion imaging according to symptoms (modified from [14])
Prior test results
Normal Abnormal
Yes Yes
New New
symptoms? symptoms?
Uncertain No No Appropriate
Timing of last test Timing of last test
< 2 Years ≥ 2 Years < 2 Years ≥ 2 Years

ago ago ago ago
CHD Risk?
Low Int/High
Inappropriate Inappropriate Uncertain Inappropriate Uncertain
Fig. 20.3 Appropriateness criteria for myocardial perfusion imaging according to results of prior tests (modified from [14])
Overall, MPI is considered inappropriate as a gatekeeper

for preoperative risk assessment. However, it is indicated in yocardial infarction or before non-cardiac vascu-
m
the setting of intermediate risk or prior to high-risk surgery lar surgery.
when at least one risk factor is present and the patient has a • The procedure is also useful after revascularization
limited functional capacity (Fig. 20.3). to determine if the bypass or stents are still
Following an acute coronary syndrome (ACS), MPI is not functional.
generally indicated for 3 months after ACS, except in those • MPI is considered appropriate in patients with
patients where a prior coronary angiogram had not been per- intermediate and high probability of CAD, as well
formed. Following revascularization with percutaneous cor- as in those subjects in whom CAD needs to be
onary intervention (PCI) or coronary artery bypass grafting excluded and exercise ECG cannot be performed
(CABG) in a more chronic setting, recurrence of symptoms due to inability or to an abnormal baseline ECG.
or the presence of suspected incomplete revascularization is
usually needed to constitute appropriate indications. The
revascularization procedure and the time elapsed before con-
sidering MPI in a variety of appropriate use ratings are 20.2.3 MPI: Patient Preparation and General
described in several guidelines. Information
Before MPI, clinical history should be obtained, including the

Key Learning Points indication for the test, symptoms, risk factors, medication, and
• MPI is useful in the identification of ischemia and/ prior diagnostic or therapeutic procedures. In diabetic patients,
or myocardial scar and location, extent, and severity diet and insulin dosing should be optimized on the day of
of ischemia and/or scar. examination. In selected insulin-treated patients, it may be use-
• The procedure is useful to determine if a stenosis ful to check blood sugar concentration before an exercise stress
is causing ischemia and in stratification after a test. Furthermore, all patients should be hemodynamically and
clinically stable for a minimum of 48 h prior to the test.
In general, heavy meals should be avoided before a stress 20.2.4 MPI: Stressor Characteristics
test, while fasting should start at least 4 h before the stress test. (See Table 20.1)
Caffeine-containing beverages and food (coffee, tea, cola, choc-
olate, etc.) should be avoided for at least 12 h before the exam. In patients with suspected or known CAD, dynamic exercise
Nitrates, beta-blockers, and calcium antagonist may is the stress test of choice. However, patients must be able to
reduce the incidence of ischemia. However, this interference exercise to a workload that increases the heart rate to at least
mostly applies when MPI is used for diagnostic purposes. 85% of age-adjusted predicted maximum. The maximum
Whenever possible, these agents should be withheld for at predicted heart rate for age is calculated by subtracting the
least three half-lives of the drug. Methylxanthine-containing age in years from 220.
medications should be suspended for at least five half-lives. Dynamic exercise should not be performed in patients
By contrast, when MPI is performed to characterize isch- who cannot achieve an adequate hemodynamic response
emic threshold or to monitor therapy effectiveness, cardiac because of non-cardiac physical limitations such as lung
medications can be continued. diseases, peripheral vascular disease, musculoskeletal dis-
If pharmacological stress testing is planned, the nuclear eases, neurological diseases, or poor motivation. These
physician should inform the patients about possible adverse patients should undergo pharmacological stress testing.
reactions. The major drugs commonly used as substitutes for exercise
stress testing are the vasodilators (adenosine, regadenoson,
20.2.3.1 Pregnancy and Lactation and dipyridamole), which induce coronary hyperemia. The
All women of childbearing age should define the possible beta-receptor agonist (dobutamine), which increases myo-
presence of pregnancy or lactation. cardial oxygen demand, is rarely used, due to variable
Pregnant Patients response to the infusion and the side effects of atropine,
In general, MPI should not be performed in confirmed which is required if patients do not have an adequate rate
pregnancy or in women whose pregnancy is not excluded response to dobutamine.
(e.g., a missed period). If MPI is considered necessary for a All the stress procedures must be supervised by a
pregnant woman, special attention must be paid to the qualified and appropriately trained health-care profes-
absorbed dose to the unborn child. The fetal absorbed dose sional. The staff must be experienced in the selection of
differs depending on the stage of pregnancy, with higher the most appropriate stress type and should have the clin-
absorbed doses in the early stages. 99mTc-labelled radiophar- ical skills to recognize patients with an increased risk of
maceuticals according to a 2-day procedure are preferred. complications. Appropriate facilities for cardiopulmo-
The stress test should be performed first. If the stress images nary resuscitation must be available and the staff must
are normal, the rest images are not necessary. have up-to-date knowledge of advanced life support
Lactating Patients (ALS) techniques.
In general, elective diagnostic nuclear medicine proce-
dures should not be performed if the patient is breastfeeding 20.2.4.1 Exercise Stress Testing
[16, 17]. Although exercise testing is generally a safe procedure, both
myocardial infarction and death have been reported and can
be expected to occur at a rate of about 1 per 10,000 tests,
Key Learning Points depending on the local case mix.
• Before MPI all patients should be hemodynami- The ECG, heart rate, and blood pressure must be moni-
cally and clinically stable for a minimum of 48 h tored and recorded at each stage of exercise as well as dur-
prior to the test. ing ST-segment abnormalities and chest pain. The patient
• Caffeine-containing beverages and food (coffee, should be continuously monitored for transient rhythm dis-
tea, cola, chocolate, etc.) should be avoided in the turbances, ST-segment changes, and other electrocardio-
day preceding the exam. graphic manifestations and symptoms. A 12-lead ECG is
• Decision about withholding anti-ischemic treat- standard.
ment should be tailored to the clinical indication for Equipment and Protocols
the examination. Both treadmill and bicycle ergometers are used for exer-
• MPI should not be performed in confirmed preg- cise testing. Although the bicycle ergometer is generally
nancy and women in whom pregnancy is not smaller and less expensive and produces less motion of the
excluded. upper body, quadriceps fatigue in patients who are not
• Similarly, elective diagnostic nuclear medicine pro- experienced cyclists is a limitation, because subjects usu-
cedures should not be performed if the patient is ally stop before reaching their maximum oxygen uptake.
breastfeeding. Several treadmill exercise protocols are available, differing
in speed and inclination of the treadmill; the Bruce and
Table 20.1 Main characteristics of physical and pharmacological stressors
456
Absolute indications for early

Protocols Patient preparation Absolute contraindications Relative contraindications termination
Dynamic exercise – Fasting is recommended 3 h – Acute coronary syndrome – Patients with decompensated – ST-segment depression ≥3 mm
(Bruce and modified before not hemodynamically and or inadequately controlled – Ischemic ST-segment elevation of
Bruce protocols: the – Interruption of nitrates, clinically stable congestive heart failure >1 mm in leads without
tracer should be injected beta-blockers, calcium – Acute pulmonary embolism – Active deep vein pathological Q waves
close to peak exercise. antagonist according to – Severe pulmonary thrombophlebitis or deep – Sustained ventricular
The patient should be cardiologist’ indications hypertension (blood pressure vein thrombosis tachyarrhythmia
encouraged to continue ≥200/110 mmHg) – Left bundle branch block, – Supraventricular tachycardia or
exercise for at least – Acute aortic dissection ventricular paced rhythm atrial fibrillation with a high heart
1–2 min after tracer – Symptomatic severe aortic – Hypertension with resting rate response
injection) stenosis systolic or diastolic blood – A decrease in systolic blood
– Hypertrophic, obstructive pressures >200/110 mmHg pressure of >20 mmHg, despite
cardiomyopathy – Recent stroke or transient increasing workload
– Uncontrolled cardiac ischemic attack – Systolic blood pressure
arrhythmias causing – Moderate to severe aortic ≥250 mmHg or diastolic blood
symptoms or hemodynamic stenosis pressure ≥130 mmHg
instability – Angina sufficient to cause distress
– Acute myocarditis and to the patient
pericarditis – Central nervous system symptoms
– Active endocarditis (e.g., ataxia, dizziness, or
near-syncope)
– Peripheral hypoperfusion (cyanosis
or pallor)
Inability of the patient to continue
the test
Protocols Patient preparation Absolute contraindications Relative contraindications Indications for early termination Adverse effects
Adenosine – Fasting is recommended 4 h – Acute coronary syndrome – Mild to moderate asthma and – Systolic blood pressure <80 mmHg – Flushing
(continuous infusion at before not hemodynamically and COPD – Persistent second-degree or sign of (35–40%)
140 μg/kg/min over – No coffee, tea, cola, clinically stable – Severe sinus bradycardia third-degree AV or sinoatrial block – Chest pain
4–6 min with injection chocolate at least 12 h before – History of severe (heart rate <40/min) – Wheezing (25–30%)
of the tracer at 2–3 min. – Interruption of nitrates, bronchospasm – Mobitz Type I second-degree – Severe chest pain associated with – Dyspnea (20%)
The infusion should be beta-blockers, calcium – Greater than first-degree AV-block (Wenckebach) ST depression of 2 mm or greater – Dizziness (7%)
continued for 2–3 min antagonist according to heart block or sick sinus – Severe atherosclerotic lesions – Signs of poor perfusion (pallor, – Nausea (5%)
after injection of the cardiologist’ indications syndrome, without a of extracranial artery cyanosis, cold skin) – Symptomatic
tracer) pacemaker – Severe aortic stenosis hypotension (5%)
– Symptomatic aortic stenosis – The use of dipyridamole
and hypertrophic obstructive during the last 24 h (to avoid
cardiomyopathy possible enhancement of the
– Systolic blood pressure drug effect)
<90 mmHg
– Uncontrolled hypertension
(systolic BP >200 mmHg or
diastolic BP >110 mmHg)
– Cerebral ischemia
– Hypersensitivity to drug’s
component
M. Bauckneht et al.
Regadenoson See adenosine See adenosine – Severe sinus bradycardia See adenosine – Headache (29%)
(0.4 μg/5 mL (heart rate <40/min) – Dyspnea (25%)
intravenous injection – Mobitz Type I second-degree – Flushing (17%)
over 10–20 s followed AV-block (Wenckebach) – Chest discomfort
by a 10 s flush of – Severe atherosclerotic lesions (11%)
5–10 mL NaCl 0.9%. of extracranial artery – Chest pain (8%)
10–20 s later the tracer – Severe aortic stenosis – Angina (8%)
is injected) – The use of dipyridamole – Dizziness (7%)
during the last 24 h (to avoid – Nausea (6%)
possible enhancement of the – Abdominal
drug effect) discomfort (6%)
Dipyridamole See adenosine See adenosine – Mild to moderate asthma and See adenosine – Chest pain (20%)
(0.56 mg/kg/min over COPD – Headache (12%)
4 min. The – Severe sinus bradycardia – Dizziness (12%)
radiopharmaceutical is (heart rate <40/min) – Ventricular
injected 3–5 min after – Severe atherosclerotic lesions extrasystoles (5%)
the completion of the of extracranial artery – Nausea (5%)
infusion) – Severe aortic stenosis – Hypotension (5%)
20 Radionuclide Imaging of Cardiovascular Disease
– The use of dipyridamole – Flushing (3%)

during the last 24 h (to avoid N.B: Chest pain is
possible enhancement of the non-specific and is
drug effect) not necessarily
indicative of the
presence of CAD
Dobutamine – Fasting is recommended 3 h – Acute coronary syndrome – Beta-blockers – Achieving >85% of the age- – Palpitation (29%)
(starting at a dose of (5 before not hemodynamically and – Severe aortic stenosis predicted peak heart rate – Chest pain (31%)
to) 10 μg/kg/min and – Interruption of nitrates, clinically stable – Symptomatic or large aortic – Systolic BP <80 mmHg – Headache (14%)
increasing at 3–5 min beta-blockers, calcium – Atrial tachyarrhythmias with aneurysm – Systolic BP >230 mmHg or – Flushing (14%)
intervals to 20, 30, and antagonist according to uncontrolled ventricular – Left bundle branch block diastolic pressure >115 mmHg – Dyspnea (14%)
40 μg/kg/min. The cardiologist’ indications response – Paced ventricular rhythm – Significant cardiac arrhythmia – Significant
radiopharmaceutical – Prior history of ventricular – Severe chest pain associated with supraventricular or
should be injected when tachycardia ST depression of 2 mm or greater ventricular
the heart rate is ≥85% of – Uncontrolled hypertension – Pallor, cyanosis, cold skin arrhythmias
the age-predicted (systolic BP >200 mmHg or (8–10%)
maximum heart rate diastolic BP >110 mmHg)
(220 age). Dobutamine – Aortic dissection
infusion should be – Hypersensitivity to
continued for 2 min after dobutamine
the radiopharmaceutical
injection)
457
Table 20.2 Grade and speed (expressed both as MPH and as km/h) of • Left bundle branch block (pharmacologic test is recom-
each 3-min stage of the Bruce protocol for treadmill exercise testing mended to avoid false-positive studies due to decreased
(total duration 21 min)
uptake in the septum), ventricular paced rhythm.
Stage % Grade MPH km/h MET units
• Hypertension with resting systolic or diastolic blood pres-
1 10 1.7 2.7 4
sures ≥200/110 mmHg.
2 12 2.5 4.0 7
3 14 3.4 5.5 10
• Recent stroke or transient ischemic attack.
4 16 4.2 6.8 13 • Severe hyperthyroidism.
5 18 5.0 8.0 15 • Severe electrolyte imbalance.
6 20 5.5 8.9 18
7 22 6.0 9.7 21 A Maximal Exercise Test Should Comply with the
MPH: miles/h Following Steps
MET: metabolic equivalent of task, i.e., the resting volume oxygen con-
sumption per minute (VO2) for a 70-kg, 40-year-old man (1 MET cor- • Before exercise, an i.v. cannula should be inserted for
responds to 3.5 mL/min/kg of body weight)
good venous access, both of the radiopharmaceutical and,
if needed, for resuscitation.
modified Bruce protocols are the most widely used (see • The ECG must be monitored continuously during the
Table 20.2). exercise test and for at least 5 min of recovery. A 12-lead
The objective of exercise testing is to maximally increase ECG printout/record should be obtained at every stage of
myocardial oxygen consumption and hence myocardial per- exercise, at peak exercise, and during recovery (i.e., ide-
fusion. There are two determinants of myocardial oxygen ally every min and with the occurrence of special events
consumption: heart rate and systolic blood pressure. such as arrhythmias, etc.).
The rate pressure product (heart rate × systolic blood pres- • The blood pressure should be recorded at least every
sure, also called “double product”) can be applied as a surro- 2–3 min during exercise.
gate parameter of myocardial hyperemia. Rate pressure • Exercise should be symptom limited, with the goal for the
product values >25,000 mmHg/min indicate a good level of patient to achieve ≥85% of his/her age-predicted maxi-
exercise, and values >30,000 mmHg/min indicate excellent mum heart rate.
levels. In general the goal of stress test should be to achieve • The radiopharmaceutical should be injected close to the
≥85% of the age-predicted maximum heart rate. However, if peak exercise; the patients should be encouraged to con-
the patient has symptoms for which he/she was referred to tinue the exercise for at least 1–2 min after the injection of
during radiopharmaceutical administration, stopping the test the radiopharmaceutical to allow the tracer to clear from
before achieving 85% of predicted maximum heart rate should the blood.
be considered despite the suboptimal increase in heart rate.
Absolute Contraindications to Maximal, Dynamic Exercise Absolute Indications for Early Termination of Exercise
• Acute coronary syndrome, until the patient has been sta- • Diagnostic ST-segment depression (≥1.5 mm).
ble for at least 48 h and the risk is clinically assessed as • Ischemic ST-segment elevation of >1 mm in leads without
acceptable. pathological Q waves.
• Acute pulmonary embolism. • Appearance of sustained ventricular tachyarrhythmia.
• Severe pulmonary hypertension. • Occurrence of supraventricular tachycardia or atrial fibril-
• Acute aortic dissection. lation with a high heart rate response.
• Symptomatic severe aortic stenosis. • A decrease in systolic blood pressure of >20 mmHg,
• Hypertrophic, obstructive cardiomyopathy. despite increasing workload.
• Uncontrolled cardiac arrhythmias causing symptoms or • Markedly abnormal increase of blood pressure (systolic
hemodynamic instability. blood pressure ≥230 mmHg or diastolic blood pressure
• Acute myocarditis and pericarditis. ≥115 mmHg).
• Active endocarditis. • Angina.
• Central nervous system symptoms (e.g., ataxia, dizziness,
Relative Contraindications to Maximal, Dynamic Exercise or near-syncope).
• Peripheral hypoperfusion (cyanosis or pallor).
• Patients with decompensated or inadequately controlled • Sustained ventricular tachycardia or fibrillation.
congestive heart failure. • Inability of the patient to continue the test.
• Active deep vein thrombophlebitis or deep vein • Technical difficulties in monitoring ECG or blood
thrombosis. pressure.
Relative Indications for Early Termination of Exercise Relative Contraindications to Vasodilator Stress Tests
• ST-segment depression <1.5 mm horizontal or • For adenosine and dipyridamole: mild to moderate asthma
downsloping. and chronic obstructive pulmonary disease (COPD).
• Arrhythmias other than sustained ventricular tachycardia • Regadenoson may be used if the patient is not actively
(including multifocal premature ventricular contractions wheezing; in patients with severe COPD, an inhaled bron-
(PVCs), triplets of PVCs, heart block, or bradyarrhyth- chodilator should be administered immediately prior to
mias), especially if symptomatic. administration of regadenoson.
• Fatigue, dyspnea, cramp, or claudication. • Severe sinus bradycardia (heart rate <40/min).
• Development of bundle branch block or intraventricular • Severe atherosclerotic lesions of extracranial artery.
conduction defect that cannot be distinguished from ven- • The use of dipyridamole during the last 24 h (to avoid
tricular tachycardia [18]. possible enhancement of the drug effect).
20.2.4.2 Vasodilator Stress Testing Early Termination of a Vasodilator Stress Test

with Adenosine, Regadenoson, Vasodilators should be stopped early under the following
or Dipyridamole circumstances:
Adenosine, regadenoson, and dipyridamole induce myo-
cardial hyperemia mediated by adenosine receptors inde- • Severe hypotension (systolic blood pressure <80 mmHg).
pendent of myocardial oxygen demand. This results in a • Persistent second-degree or sign of third-degree atrioven-
modest increase in heart rate and a modest decrease in both tricular or sinoatrial block.
systolic and diastolic blood pressures. Myocardium sup- • Wheezing.
plied by a stenotic coronary artery has a reduced perfusion • Severe chest pain.
reserve and this leads to heterogeneity of perfusion during
vasodilation or even to myocardial ischemia caused by Combination with Low-Level Exercise
coronary steal. Because myocardial tracer uptake is pro- Low-level exercise can be performed in conjunction with
portional to perfusion, this results in heterogeneous uptake vasodilator tests. Low-level exercise significantly reduces
of tracer. the side effects (flushing, dizziness, nausea, headache,
Indications vasodilator-induced hypotension) and improves image qual-
The indications are the same as for exercise myocardial ity due to reduced bowel activity and higher target-
perfusion imaging but refer to patients not able or expected background ratio. Accordingly, if possible low-level exercise
to be unable to achieve ≥85% of maximal age-predicted is recommended in combination with vasodilator stress test-
heart rate during exercise. Vasodilators (without exercise) ing. Low-level exercise is not recommended for patients with
should be preferred to exercise in the presence of left bundle left bundle branch block or ventricular paced rhythm [18].
branch block.
Absolute Contraindications to Vasodilator Stress Tests: Adenosine
Adenosine induces direct coronary arteriolar vasodilation
• Acute coronary syndrome (vasodilator stress test may be through specific activation of the A2A receptor. This results in
considered when the patient has been stable for at least a 3.5- to 4-fold increase in myocardial blood flow in healthy
24 h and the clinically assessed risk is deemed coronary vessels. Activation of A1, A2B, and A3 receptors may
acceptable). cause undesirable side effects of adenosine infusion: AV
• Severe bronchospasm. block (A1 receptor), peripheral vasodilation (A2B receptor),
• Greater than first-degree heart block or sick sinus syn- and bronchospasm (A2B and A3 receptors). Peak vasodilation
drome, without a pacemaker. occurs within 1–2 min after the start of the adenosine infu-
• Symptomatic aortic stenosis and hypertrophic obstructive sion. The half-life of adenosine is approximately 10 s. It is
cardiomyopathy. either phosphorylated to adenosine monophosphate by
• Systolic blood pressure <90 mmHg. adenosine kinase or degraded to inosine by adenosine

• Recent cerebral ischemic episode. deaminase.
• Active treatment with xanthines. Side Effects of Adenosine
However, in acute coronary syndromes including unsta- • Minor side effects are common and occur in approxi-
ble angina, a vasodilator stress test may be considered when mately 80% of patients. The common side effects are
the patient has been stable for at least 48 h and the clinically flushing (35–40%), chest pain (25–30%), dyspnea (20%),
assessed risk is deemed acceptable. dizziness (7%), headache (5%), nausea (5%), and
symptomatic hypotension (5%). Chest pain is non-spe- Absolute Contraindications

cific and is not necessarily indicative of CAD.
• AV block occurs in approximately 8% of cases. However, • Patients with ongoing wheezing or a history of significant
the incidence of second-degree AV block is only 4% and reactive airway disease.
that of complete heart block is less than 1%. Most cases • Second- or third-degree AV block without a functioning
(>95%) of AV block are self- limiting and do not require pacemaker.
termination of the infusion. • Sinus node disease, such as sick sinus syndrome or symp-
• ST-segment depression of 1 mm or greater occurs in tomatic bradycardia, without a functioning pacemaker.
5–7% of cases. Compared to chest pain, ST changes may • Systolic BP less than 90 mmHg. The risk of serious hypo-
be indicative of true ischemia. tension may be higher in patients with autonomic dys-
• Fatal or nonfatal myocardial infarction is extremely rare function, hypovolemia, left main coronary artery stenosis,
but has been reported. stenotic valvular heart disease, pericarditis or pericardial
• Atrial fibrillation has been reported within several min- effusions, or stenotic carotid artery disease with cerebro-
utes of initiation of adenosine infusion. vascular insufficiency.
• New onset or recurrence of convulsive seizures has • Uncontrolled hypertension (systolic BP >200 mmHg or
occurred infrequently following adenosine administration. diastolic BP >110 mmHg).
• Hemorrhagic and ischemic cerebrovascular accidents • Recent (<48 h) use of dipyridamole or dipyridamole-
have occurred. containing medications (e.g., Aggrenox).
• Known hypersensitivity to adenosine.
Due to the exceedingly short half-life of adenosine • Unstable angina, acute coronary syndromes, or less than
(<10 s), most side effects resolve in a few seconds after dis- 2–4 days after an acute myocardial infarction.
continuation of the adenosine infusion. Theophylline or ami-
nophylline are only rarely required. Relative Contraindications
Hemodynamic Effects
Adenosine results in a modest increase in heart rate and a • Profound sinus bradycardia (heart rates <40/min).
modest decrease in both systolic and diastolic blood pressure • Mobitz Type 1 s-degree AV block (Wenckebach).
(BP). Systolic BP decreases by 10 ± 37 mmHg, and diastolic • Ingestion of caffeinated foods or beverages (e.g., coffee,
BP decreases by 8 ± 19 mmHg, while the heart rate increases tea, sodas) within the last 12 h.
by 14 ± 30 bpm. Maximum hemodynamic changes after • Seizure disorder. New onset or recurrence of convulsive
adenosine are as follows: increase in heart rate of more than seizures has been reported following adenosine adminis-
40 bpm in 3%, decrease in BP systolic of more than 35 mmHg tration. Methylxanthine (aminophylline) use is not
in 8%, and decrease in BP diastolic of more than 25 mmHg recommended in patients who experience seizures in

in 5% of patients. association with adenosine administration.
Procedure
An infusion or a syringe pump is necessary for adenosine Indications for early termination of adenosine infusion
administration at a constant infusion rate. Either two sepa- and for reversal of adenosine using intravenous aminophyl-
rate i.v. lines or one i.v. line with a dual-port Y-connector is line (50–250 mg intravenously at least 1 min after the tracer
required to allow injection of the radiopharmaceutical with- injection) include the following:
out interruption of the adenosine infusion. ECG and blood
pressure monitoring should be carried out as with exercise • Severe hypotension (systolic BP <80 mmHg).
stress testing. • Development of symptomatic, persistent second-degree
Adenosine Dose or complete AV block.
Adenosine should be given as a continuous infusion at • Other significant cardiac arrhythmia.
140 μg/kg/min over 4–6 min with injection of the radiophar- • Wheezing.
maceutical at 2–3 min. The infusion should be continued for • Severe chest pain associated with ST depression of 2 mm
2–3 min after the injection of the radiopharmaceutical. For or greater.
patients at risk of complications (recent ischemic event, bor- • Signs of poor perfusion (pallor, cyanosis, cold skin).
derline hypotension, inadequately controlled asthma), the
infusion can be started at a lower dose (50 μg/kg/min). If this Note: For signs or symptoms not significant enough to
dose is tolerated for 1 min, the dose rate can be increased to terminate the test, one can consider shortening the time of
75, 100, and 140 μg/kg/min at 1-min intervals and then con- infusion from 6 to 4 min. The adenosine infusion should be
tinued for 4 min. The radiopharmaceutical should be injected terminated early if there are pronounced hemodynamic
1 min after starting 140 μg/kg/min dose. A shorter duration changes or wheezing or other symptoms, which are not
of infusion may also be effective. enough to stop the test.
Regadenoson Maximum hemodynamic changes after regadenoson were as

Regadenoson is a high-affinity agonist for the A2A adenosine follows: increase in heart rate of more than 40 bpm in 5%,
receptor, with at least tenfold lower affinity for the A1 adenos- decrease in systolic BP of more than 35 mmHg in 7%, and
ine receptor, and weak, if any, affinity for the A2B and A3 ade- decrease in diastolic BP of more than 25 mmHg in 4% of
nosine receptors. Activation of the A2A adenosine receptor by patients.
regadenoson produces coronary vasodilation and increases Procedure
coronary blood flow by the same mechanism by which ade- Regadenoson can be administered intravenously by man-
nosine produces coronary vasodilation. The maximal plasma ual injection. A Y-connector is not needed. ECG and blood
concentration of regadenoson is achieved within 1–4 min pressure monitoring should be carried out as with exercise
after injection and parallels the onset of the pharmacody- stress testing.
namic response. The half-life of this initial phase is approxi- Regadenoson Dose
mately 2–4 min. An intermediate phase follows, with a The recommended intravenous dose of regadenoson is
half-life on average of 30 min coinciding with loss of the 0.4 mg, regardless of weight. It should be administered into
pharmacodynamic effect. The last phase consists of a decline a peripheral vein, using a 22-gauge or larger catheter or nee-
in plasma concentration with a half-life of approximately 2 h. dle, as a slow bolus over ~30 s followed by a 10-s flush of
Side Effects of Regadenoson 5–10 mL NaCl 0.9%; the radiopharmaceutical is injected
within 2 min of regadenoson administration using the same
• The initial studies with regadenoson were done with bolus intravenous line used for regadenoson.
injection of the agent. Bolus administration resulted in a Absolute Contraindications
much higher incidence of side effects (see below). The
side effects are markedly reduced when the drug is admin- • Patients with bronchospastic lung disease with ongoing
istered as a 30 s infusion rather than a bolus. Most com- wheezing or a history of significant reactive airway dis-
mon reactions to administration of regadenoson are ease should not undergo regadenoson stress testing.
shortness of breath, headache, and flushing. The minor • Second- or third-degree AV block or sinus node dysfunc-
adverse events are headache (29%), dyspnea (25%), tion without a functioning pacemaker.
flushing (17%), chest discomfort (11%), chest pain (8%), • Sinus node disease, such as sick sinus syndrome or symp-
angina (8%), dizziness (7%), nausea (6%), and abdominal tomatic bradycardia, without a functioning pacemaker.
discomfort (6%). • Systolic BP less than 90 mmHg. The risk of serious hypo-
• Rhythm or conduction abnormalities are seen in 26% of tension may be higher in patients with autonomic dys-
patients. First-degree AV block is detected in 3% and function, hypovolemia, left main coronary artery stenosis,
second-degree AV block in 0.1%. Asystole and QT inter- stenotic valvular heart disease, pericarditis or pericardial
val prolongation have also been reported. effusions, or stenotic carotid artery disease with cerebro-
• Most adverse reactions begin soon after dosing and gen- vascular insufficiency.
erally resolve within <5 min, except for headache, which • Uncontrolled hypertension (systolic BP >200 mmHg or
resolves in most patients within 30 min. diastolic BP >110 mmHg).
• Aminophylline may be administered in doses ranging • Recent (<48 h) use of dipyridamole or dipyridamole-
from 50 to 250 mg by slow intravenous injection (50– containing medications.
100 mg over 30–60 s) to attenuate severe and/or persistent • Known hypersensitivity to adenosine or regadenoson.
adverse reactions at least 1 min after tracer injection. • Unstable angina, acute coronary syndrome, or less than
• New-onset or recurrent atrial fibrillation and atrial flutter 2–4 days after an acute myocardial infarction.
have been reported following regadenoson administration.
• New-onset or recurrence of convulsive seizures has Relative Contraindications
occurred following regadenoson administration.
• Refractory ischemia/myocardial infarction, hemorrhagic, • Profound sinus bradycardia (heart rate <40/min).
and ischemic cerebrovascular accidents have been • Mobitz Type I second-degree AV block (Wenckebach).
reported. • Severe aortic stenosis (see “Special Populations” section).
• Refractory ischemia and myocardial infarction have • Ingestion of caffeinated foods or beverages (e.g., coffee,
occurred. tea, sodas) within the last 12 h should be avoided.
• Seizure disorder: regadenoson may lower seizure thresh-
Hemodynamic Effects old, and aminophylline should not be used in cases of sei-
In clinical studies, the majority of patients have had an zures associated with regadenoson; these seizures may be
increase in heart rate and a decrease in BP within 15 min of new onset or may be recurrences. In addition, some
after administration of regadenoson. Systolic BP decreased seizures are prolonged and may require urgent anticon-
by 13 ± 14 mmHg and diastolic BP decreased 10 ± 8 mmHg. vulsive management.
Indications for reversal of regadenoson (50- to 250-mg amino- Procedure

phylline intravenously at least 1 min after the tracer injection)
Although usually given by an infusion pump, dipyridamole
include the following:
can also be administered intravenously by manual injection.
• Severe hypotension (systolic BP <80 mmHg). A Y-connector is not needed. ECG and blood pressure moni-
• Development of symptomatic, persistent second-degree toring should be carried out as with exercise stress testing.
or complete heart block. Dipyridamole Dose
• Other significant cardiac arrhythmia. Dipyridamole should be given as a continuous intravenous
• Wheezing. infusion at 140 μg/kg/min over 4 min. The radiopharmaceuti-
• Severe chest pain associated with ST depression of 2 mm cal is injected 3–5 min after the completion of the infusion.
or greater. Absolute Contraindications
• Signs of poor perfusion (pallor, cyanosis, cold skin).
• Patients with bronchospastic lung disease with ongoing
Dipyridamole wheezing or a history of significant reactive airway dis-
Dipyridamole is an indirect coronary artery vasodilator ease should not undergo dipyridamole stress testing.
that increases the tissue levels of adenosine by prevent- • Systolic BP less than 90 mmHg. The risk of serious hypo-
ing the intracellular reuptake and deamination of ade- tension may be higher in patients with autonomic dys-
nosine. This results in a 3.8- to 7-fold increase in function, hypovolemia, left main coronary artery stenosis,
coronary blood flow velocity. Dipyridamole-induced stenotic valvular heart disease, pericarditis or pericardial
hyperemia lasts for more than 50 min. Peak vasodilation effusions, or stenotic carotid artery disease with cerebro-
after dipyridamole administration occurs on average vascular insufficiency.
6.5 min after the start of the infusion. The half-life of • Uncontrolled hypertension (systolic BP >200 mmHg or
dipyridamole is approximately 30–45 min. It is metabo- diastolic BP >110 mmHg).
lized in the liver to the glucuronic acid conjugate and • Ingestion of caffeinated foods or beverages (e.g., coffee,
excreted in the bile. tea, sodas) within the last 12 h should be avoided.
Side Effects of Dipyridamole • Known hypersensitivity to dipyridamole.
• Unstable angina, acute coronary syndrome, or less than
• Minor side effects are common and occur in approxi- 2–4 days after an acute myocardial infarction.
mately 50% of patients. These adverse events include the
Note: in patients taking oral dipyridamole, intravenous dipyri-
following: Chest pain (20%), headache (12%), dizziness damole may be administered safely and efficaciously.
(12%), ventricular extrasystoles (5%), nausea (5%),
hypotension (5%), and flushing (3%). Chest pain is non- Relative Contraindications
specific and is not necessarily indicative of the presence
of CAD. • Profound sinus bradycardia (heart rates <40/min).
• The incidence of AV block with dipyridamole is less than • Second- or third-degree AV block without a functioning
that observed with adenosine (2%). pacemaker.
• ST-segment and T-wave changes occurred (8%); however, • Severe aortic stenosis.
unlike chest pain, ST changes may be indicative of true • Seizure disorder. Methylxanthine (aminophylline) use is
ischemia. not recommended in patients who experience seizures in
• Fatal or nonfatal myocardial infarction is extremely rare association with dipyridamole stress testing.
but has been reported.
Indications for Early Termination of Dipyridamole
Symptoms may last for a longer period of time than with Infusion and for Reversal of Dipyridamole (50- to 250-mg
other vasodilators (15–25 min) and may vary significantly in Aminophylline Intravenously At Least 1 min After the Tracer
individual patients. Aminophylline (50–250 mg intrave- Injection) Include the Following:
nously) is often required to reverse these side effects; prefer-
ably it must be injected not earlier than 3 min after the • Severe hypotension (systolic BP <80 mmHg).
injection of the radiopharmaceutical. • Development of symptomatic, persistent second-degree
Hemodynamic Effects or complete heart block.
Dipyridamole administration results in a modest increase • Other significant cardiac arrhythmia.
in heart rate and a modest decrease in both systolic and dia- • Wheezing.
stolic BPs. Systolic BP decreased by 14 ± 15 mmHg, while • Severe chest pain associated with ST depression of 2 mm
heart rate increased by 17 ± 11 bpm. Systolic BP fell to or greater.
<90 mmHg in 2% of patients. • Signs of poor perfusion (pallor, cyanosis, cold skin).
Sympathomimetic Agents Stress Testing with Dobutamine should be continued for 2 min after the radiopharmaceutical
Dobutamine infusion results in direct β1 and β2 adrener- injection. Atropine (0.25 mg i.v. one to three times with 30-s
gic stimulation with a dose-related increase in heart rate, intervals) can be added if heart rate does not reach 85% of
BP, and myocardial contractility and, hence, an increase in age-predicted maximal heart rate. Patients should be
myocardial oxygen demand (in severe ischemic heart dis- informed of possible difficulties while driving in 2 h follow-
ease, contractility may be reduced with high doses of ing atropine administration, due to reduced ocular accom-
dobutamine). modation. Beta-blockers are dobutamine antagonists and
Dobutamine increases regional myocardial blood flow should be discontinued for at least three to four half-lives
based on physiologic principles of coronary flow reserve. A before the test.
similar dose-related increase in subepicardial and subendo- Indications
cardial blood flow occurs within vascular beds supplied by
normal coronary arteries. However, blood flow increases • Dobutamine is a secondary pharmacologic stressor that is
minimally within vascular beds supplied by significantly ste- recommended only in patients who cannot undergo exer-
nosed arteries, with most of the increase occurring within the cise stress and who also have contraindications to phar-
subepicardium rather than the subendocardium. Furthermore, macologic vasodilator stressors (mainly bronchospastic
at a dose of 20 μg/kg/min, dobutamine induces coronary flow airway disease).
heterogeneity, which is similar to exercise stress but pro- • Dobutamine perfusion imaging has not been studied as
duces less heterogeneity than that induced by adenosine or extensively as the vasodilator stress perfusion imaging in
dipyridamole. the evaluation and prognostication of patients with CAD.
The plasma half-life of dobutamine is 2 min with the
onset of action within 1–2 min; however, up to 10 min may Absolute Contraindications to Dobutamine
be required to obtain the peak effect.
Side Effects of Dobutamine • Unstable angina, acute coronary syndrome, or less than
2–4 days after an acute myocardial infarction.
• The common side effects are palpitations (29%), chest • Hemodynamically significant left ventricular outflow
pain (31%), headache (14%), flushing (14%), dyspnea tract obstruction.
(14%), and significant supraventricular or ventricular • Atrial tachyarrhythmias with uncontrolled ventricular
arrhythmias (8–10%). response.
• Ischemic ST-segment depression occurs in approximately • Prior history of ventricular tachycardia.
one-third of patients undergoing dobutamine infusion. • Uncontrolled hypertension (systolic BP >200 mmHg or
• Severe side effects may require intravenous administra- diastolic BP >110 mmHg).
tion of a short-acting β-blocker (esmolol, 0.5 mg/kg over • Patients with aortic dissection.
1 min); i.v. metoprolol (5 mg) can also be used. • Known hypersensitivity to dobutamine.
Hemodynamic Effects Relative Contraindications

The hemodynamic response to dobutamine infusion is
dose dependent and varies based on the maximal infusion • Patients who are on β-blockers, in whom the heart rate
rate obtained. In studies titrating to a maximal dose of 40 μg/ and inotropic responses to dobutamine will be
kg/min, heart rate increased by 45 ± 18 bpm, while systolic attenuated.
BP increased by 30 ± 21 mmHg in one study, and by • Severe aortic stenosis.
12 ± 29 mmHg in another study. • Patients with symptomatic or large aortic aneurysm.
Dobutamine Procedure and Dose • Left bundle branch block.
An infusion pump is necessary for dobutamine adminis- • Paced ventricular rhythm.
tration. Two separate i.v. lines or one i.v. line with a
Y-connector is required for injection of radioisotope during Contraindications to Atropine Administration Under
dobutamine infusion. ECG monitoring and blood pressure Dobutamine Stress
monitoring should be performed as with other pharmaco-
logical stressors. Dobutamine is infused incrementally, start- • Narrow angle glaucoma.
ing at a dose of (5 to) 10 μg/kg/min and increasing at 3–5-min • Obstructive uropathy, including bladder neck obstruction
intervals to 20, 30, and 40 μg/kg/min. The radiopharmaceuti- from prostatic hypertrophy.
cal should be injected when the heart rate is ≥85% of the • Atrial fibrillation with an uncontrolled heart rate.
age-predicted maximum heart rate. Dobutamine infusion • Obstructive gastrointestinal disease or paralytic ileus.
Indications for Early Termination of Dobutamine Infusion Table 20.3 Main properties of tracers for SPECT-MPI
SPECT-MPI 99m
Tc- 99m
Tc-
• Achieving >85% of the age-predicted peak heart rate radiotracers Tl-chloride
201
tetrofosmin sestamibi
(after maintaining for 1 min following radiotracer Physical half-life 73 h 6 h 6 h
Production Cyclotron On-site On-site
injection).
labelling labelling
• Severe hypotension (systolic BP <80 mmHg). Uptake Potassium Lipophilic Lipophilic
• Severe hypertension (systolic BP >230 mmHg or dia- mechanism analogue cation cation
stolic pressure >115 mmHg). First pass 80–85 50–65 50–65
• Significant cardiac arrhythmia. Termination for ventricu- extraction (%)
lar tachycardia or atrial tachyarrhythmia is more likely Energy of photon 68–82 140 140
emission (keV)
with dobutamine than with other stressors.
• Severe chest pain associated with ST depression of 2 mm
or greater. Termination for ST-segment depression is while prolonged retention depends on the integrity of the cell
more likely with dobutamine than with other stressors. membrane and hence on viability. Redistribution of 201Tl
• Signs of poor perfusion (pallor, cyanosis, cold skin). (nonuniform clearance of thallium from the myocardium,
Technical problems with the monitoring equipment. with slower clearance from areas of ischemia and more rapid
clearance from areas regions of normal perfusion) occurs
Reversal of Complications and Side Effects of Dobutamine over several h after administration [15, 16, 18]. For this rea-
Severe side effects, arrhythmia, or ST changes may son, sensitivity in the detection of ischemia can be enhanced
require the i.v. administration of a short-acting β-blocker injecting a single dose of 74–111 MBq (92–130 MBq accord-
(esmolol, 0.5 mg/kg over 1 min); intravenous metoprolol ing to American guidelines) [14] of 201Tl-chloride prior to
(5 mg) can also be used [15–17]. peak exercise stress or at peak pharmacologic vasodilatation
and starting SPECT imaging 10–15 min later. Redistribution
(rest) imaging is done 2.5–4.0 h later.
Key Learning Points In cases where standard stress-redistribution imaging
• In patients with suspected or known CAD, dynamic shows a fixed or minimally reversible perfusion abnormal-
exercise is the stress test of choice. ity, myocardial viability can be assessed with a rest image
• Dynamic exercise should not be performed in at 18–24 h, or following injection of an additional 37 MBq
patients who cannot achieve an adequate hemody- activity of 201Tl [15, 16, 18]. An alternative method for
namic response. viability assessment is injection of 92–130 MBq of 201Tl at
• Two groups of drugs are commonly used as substi- rest followed by 3- to 4-h or 18- to 24-h redistribution
tutes for exercise stress testing: Vasodilators (ade- imaging [17].
nosine, regadenoson, and dipyridamole), which Although 201Tl is a good tracer for myocardial perfusion
induce coronary hyperemia, and β-receptor agonists imaging, it has several limitations:
(dobutamine) that increase myocardial oxygen
demand. • Relatively long physical half-life: high radiation burden
to the kidneys.
• Relatively low injected activity: low signal-to-noise ratio;
images can be suboptimal particularly in obese patients.
20.2.5 SPECT-MPI: Radiopharmaceuticals • Relatively low energy emission: low-resolution images
and significant scatter and attenuation by soft tissue
Radiopharmaceuticals used in SPECT myocardial perfusion [16, 18].
imaging protocols are 201Tl-chloride and two 99mTc-labelled
agents: 99mTc-2-methoxyisobutylisonitrile (99mTc-sestamibi) 99m
Tc-sestamibi and 99mTc-tetrofosmin have very similar
and 99mTc-1,2-bis[bis(2-ethoxyethyl) phosphino] ethane characteristics: lipid-soluble, cationic, agents. The 6-h physi-
(99mTc-tetrofosmin) (see also Table 20.3). cal half-life and emission of 140 keV photons are favorable
Thallium-201 is an analogue of potassium (monovalent for imaging. These tracers, however, have a first-pass extrac-
cation), with a physical half-life of 73.1 h. It decays by elec- tion substantially less than 201Tl (~60% versus 88%). Within
tron capture emitting gamma photons of 135 and 167 keV the cell 99mTc-labelled tracers are retained in intact mitochon-
(12% abundance) and X-rays with 67–82 keV energy (88% dria, reflecting viable myocytes. Excretion of 99mTc-sestamibi
abundance). is primarily biliary (~80%), while excretion of 99mTc-tetrofos-
Myocardial uptake of 201Tl is directly related to perfusion min is ~50% biliary and 50% renal. Two protocols can be
and viability: the initial uptake is proportional to perfusion, used for imaging: a 2-day protocol or a 1-day protocol.
Two-day protocol. Ideally, stress and rest imaging with injection. Therefore, the ideal imaging time should reflect
99m
Tc-agents should be performed on two separate days to the best compromise between high myocardial count rate
avoid the contamination by residual activity (“shine-through” and the lowest surrounding organ uptake (mostly liver and
or “crosstalk”) from the first injection that inevitably inter- gallbladder). For this reason it was initially recommended to
feres with interpretation of images reflecting the second wait approximately 60–75 min after 99mTc-sestamibi i.v.
injection. Usually the administered activity is 350–700 MBq/ injection before starting the image acquisition. Subsequently,
study. In larger patients (e.g., >113 kg or BMI >35) or in it was suggested to decrease this relatively long-time interval
women where significant breast attenuation is anticipated, a between injection and imaging [19].
low activity of 99mTc-radiotracer may result in suboptimal Many studies suggest that early post-exercise imaging is
images and a 2-day imaging protocol with higher activities feasible and can potentially improve the detection of post-
(666–1110 MBq) for each injection may be preferable. ischemic wall motion impairment without compromising
One-day protocol. For many patients, 2-day imaging is image quality and perfusion data [20, 21]. Myocardial stun-
impractical, and thus stress and rest studies are usually per- ning is a transient postischemic left ventricle (LV) dysfunc-
formed using a 1-day protocol. This requires administration tion induced by a brief episode of severe ischemia in the
of a lower activity (approximately 250–400 MBq: one-fourth absence of irreversible cellular damage, which persists when
of the total activity) for the first injection and a higher activ- myocardial perfusion has returned to normal [22]. The sever-
ity (approximately three-fourths of the total activity) for the ity and duration of regional LV dysfunction depends on the
second injection. severity of the ischemic insult as well as on the metabolic
One-day stress/rest and rest/stress 99mTc-radiotracer proto- condition of the myocardium before the insult.
cols are now typically performed with no significant delay Although some studies have shown that regional dyssyn-
between obtaining the first set of images and the second tracer ergy may persist for several hours after completion of exercise,
injection at stress or rest. A 3:1 ratio of activities with a 2-h it is generally accepted that LV dysfunction tends to resolve
delay and a 4:1 ratio with no delay provide similar results. more rapidly, usually within 30 min. Myocardial distribution of
For both 99mTc-labelled tracers: the perfusion tracer at the time of tomographic acquisition rep-
resents the relative myocardial perfusion pattern at the time of
• Splanchnic uptake and excretion are markedly higher tracer injection, but the gated tomogram represents LV function
than for 201Tl, so the interpretation of the inferior wall per- at the time of acquisition, which can be anywhere from 15 min
fusion could be hampered. to 1 h after the completion of stress [23], therefore underesti-
• The tracer molecules taken up by the cardiac myocytes mating the degree of left ventricular dysfunction.
remain within the cells: usually two visits on two different Therefore, gated SPECT imaging performed early after
days are necessary to obtain optimal stress and rest radiotracer administration could potentially detect more
images. post-exercise LV dysfunction. According to literature, imag-
• The uptake of both 99mTc-labelled tracers as a function of ing should begin:
myocardial perfusion is less avid than in the case of 201Tl, For 99mTc-sestamibi:
and so defects may be less pronounced.
• Radiation exposure to staff members is higher compared • Exercise stress: 15–20 min after injection.
to the use of 201Tl. • Pharmacological stress (associated with higher subdia-
phragmatic activity): 45–60 min after injection (to permit
On the other hand, the 99mTc-labelled tracers offer other at least partial clearance from the liver and bowel adjacent
and probably more important advantages over 201Tl-chloride: to the diaphragm).
• Rest: 45–60 min after injection.
• The higher energy of 99mTc leads to better quality images
because of less attenuation and scatter. For 99mTc-tetrofosmin:
• The shorter half-life of 99mTc permits much higher activi-
ties to be administered, yielding better counting statistics. • Exercise stress: 10–15 min after injection.
• Pharmacological stress (higher subdiaphragmatic activ-
When a 99mTc-labelled tracer is injected, the myocardial ity; see above): 45 min after injection.
uptake is approximately 1.5% of the total activity, while the • Rest: 30–45 min after injection.
liver and biliary uptake represents almost 30% of the admin-
istered activity. The effective half-life of clearance (which The differences between the two radiotracers are related to
includes both the biological half-life and the physical half- the difference in their respective biological characteristics.
life of 99mTc) from the heart is approximately 3–4 h and for Myocardial perfusion imaging is most commonly
the liver is approximately 30 min after a rest or exercise performed on a dual-head SPECT system with rotating
detectors and NaI(Tl)-crystals. Newer dedicated single-head Number of rejected beats should be limited to below 25%.
gamma cameras have been designed specifically for heart The beat length acceptance window (usually 20%) aims to
imaging and are usually based on the use of semiconductor eliminate data from beats that are “too short” or “too long”
detectors. Images should be acquired in a gated mode using while still accepting enough beats. If there is an irregularly
an ECG trigger. This modality offers at least three important irregular rhythm, the data should be acquired without gating,
advantages: to assure that the myocardial perfusion information is obtained.
• Evaluation of LV ejection fraction (EF) and volumes and

evaluation of LV regional wall motion and thickening as
Key Learning Points
well as diastolic function.
• Two protocols can be adopted for myocardial perfu-
• Improvement of the diagnostic accuracy of perfusion
sion scintigraphy: The standard 2-day protocol or
imaging in the event of attenuation problems (apparently
the 1-day protocol.
irreversible perfusion defects due to attenuation artifacts
• According to many studies, early post-exercise
that may be recognized as wrongly interpreted scar tissue,
imaging is feasible and can potentially improve the
when function is maintained).
detection of postischemic stunning without com-
• Performance of phase analysis for assessment of LV
promising image quality and perfusion data.
dyssynchrony.
• The gating of a SPECT acquisition is easily imple-
mented using the QRS complex of the ECG signal,
The gating of a SPECT acquisition is easily implemented
since the R-wave corresponds to end-diastolic
using the QRS complex of the ECG signal, since the R-wave
stage.
corresponds to end-diastolic stage. The gating hardware
interfaces with the acquisition computer that controls the
gantry. Data corresponding to each frame are automatically
sorted by the camera into the appropriate image matrix. The 20.2.6 PET-MPI: Radiopharmaceuticals
R-wave has to be positive in most triggering systems. Ideally,
the R-wave should be at least threefold higher than other A unique advantage of PET over other noninvasive imaging
positive waves (P and T) and should have the steepest rising modalities is the ability to image and quantify tracer concen-
slope of the ECG cycle. trations. Table 20.4 lists some of the more commonly used
Using ECG-gated SPECT, the heartbeat is usually divided cardiac PET tracers and their characteristics.
into 16 temporal frames or bins. The R-wave of the QRS Rubidium-82 (82Rb), first described for myocardial perfu-
complex serves as the signal and starting point (triggering sion imaging in the 1970s, is a potassium analogue localiz-
point) of the cardiac cycle. The cardiac cycle is divided into ing in cells through the Na+/K+ ATPase pump. 82Rb has such
frames representing different phases of the cardiac cycle. a short physical half-life, 75 s, that the tracer must be pro-
After a gated SPECT acquisition with 16 bins and 64 projec- duced on site. 82Rb is produced by a strontium-82/rubidium-
tions, the count statistics should provide accurate edge defi- 82 generator as follows: 82Sr decays by electron capture to
nitions of the ventricular walls and the temporal resolution 82
Rb, which is eluted from the generator with saline directly
that is required to record the end-systolic period. into the patient [24–26]. The 75-s half-life allows the genera-
For gated acquisition, a 3-lead ECG is sufficient. The tor to be eluted every 10 min, making 82Rb suitable for
electrodes should be positioned ventrally in supine posi- repeated and sequential perfusion studies. The short half-life
tion. Before the acquisition, it is important to verify that the reduces the total study duration to approximately 30–45 min.
ECG monitor and acquisition display show the same heart A diffusible, inert tracer such as [15O]water is ideal for
rate and stable triggering. Patients with a fairly regular perfusion quantification. The tracer rapidly diffuses into tis-
heart rhythm can be easily studied with gated sue, behaving in similar fashion to a microsphere. [15O]water
SPECT. Patients with arrhythmias (atrial fibrillation, sinus has a very high tissue extraction (~90%) allowing the use of
arrhythmia, frequent premature beats, intermittent and simple kinetic models to calculate blood flow. Because of the
dual-chamber pacing, etc.) can also be studied with ECG very short half-life (125 s), like rubidium, it can be used to
triggering if the arrhythmia is not too irregular, but acquisi- measure myocardial blood flow (MBF) [27]. Measuring
tion times will be significantly prolonged with resulting [15O]water myocardial blood flow in conjunction with coro-
loss of accuracy. Regular brady- or tachycardia (e.g., AV nary CT angiography offers a more complete assessment of
block, atrial flutter, etc.) does not interfere with acquisition the significance of a specific stenosis than either measure-
of gated SPECT but may have profound influence on LVEF ment alone. This tracer has a washout rate proportional to
and volume values observed, mostly due to longer or flow. Indeed, the relationship between tracer clearance and
shorter LV filling intervals. MBF is linear, making it an ideal tracer for absolute
Table 20.4 Main properties of tracers for PET-MPI

PET-MPI radiotracers 82
Rb [13N]ammonia 18
F-flurpiridaz [15O]water
Physical half-life 76 s 10 min 109 min 2 min
Activity for 3D stress (MBq) 740–1850 740 222–370 370–1110
Scan duration (stress) 6 min 10 min 5–15 min 6 min
Production 82
Sr/82Rb generator Cyclotron Cyclotron Cyclotron
Myocardial extraction fraction 65% 80% 94% 100%
Image quality Good Excellent Excellent Poor
Semiquantitative image interpretation Yes Yes Yes No
Myocardial blood flow quantification Yes Yes Yes Yes
Stress modality Pharmacologic Pharmacologic or exercise Pharmacologic or exercise Pharmacologic
Average positron energy (MeV) 1.48 0.49 0.25 0.74
Max positron range in tissues (mm) 8.6 2.5 1.0 4.1
Modified from [42]
easurement of MBF. However, its myocardial retention is

m o xidoreductase (also known as mitochondrial complex-1) of
virtually 0%, and thus it is not suitable for assessment of the electron transport chain [29]. 18F-Flurpiridaz inhibits mito-
cardiac function: ejection fraction, wall motion, and thicken- chondrial complex-1 by competing for binding with ubiqui-
ing. The myocardium is not clearly visualized because of the none without affecting the viability of cardiomyocytes. This
absence of contrast between the myocardium and the blood radiotracer exhibits rapid myocardial uptake and slow wash-
pool due to the lack of retention. Consequently, the use of out from cardiomyocytes. Experimental PET imaging demon-
factor analysis allows identification of the heart and defini- strated a high and sustained cardiac uptake that is proportional
tion of volumes of interest (VOI) without the need for an to blood flow [30]. In rats, the first-pass myocardial extraction
additional blood pool scan; nevertheless, no true morpho- fraction of 18F-flurpiridaz is 94%. The flow-independent
logical images are available, and gated PET studies are extraction fraction of 18F-flurpiridaz implies a linear relation-
highly demanding. Moreover, due to the short scan period ship between uptake and MBF, which is an important attribute
and short half-life of 15O, the [15O]water technique is charac- for stress MBF measurements. In a pig model, 18F-flurpiridaz
terized by poor statistics. To use 15O, a cyclotron must be exhibits higher activity ratios of the myocardium versus the
available on site. blood, liver, and lungs compared to [13N]ammonia [31].
[13N]ammonia also requires an on-site cyclotron, but its Moreover, 18F-flurpiridaz has an excellent correlation with
production is easier than that of [15O]water. Because of its radioactive microspheres in assessing absolute quantitation of
longer physical half-life (9.96 min), the time schedule for a regional MBF [32]. 18F-Flurpiridaz is not yet approved for
complete rest-stress study is slightly longer than with [15O] clinical use, but preliminary data are promising.
water [27]. Once injected, neutral [13N]ammonia is in equi- PET-MPI is superior to SPECT due to improved image
librium with its charged ammonium (NH4) ion [28]. The neu- quality, spatial/temporal resolution, and diagnostic accuracy.
tral NH3 molecule readily diffuses across plasma and cell However, PET-MPI does have some limitations, including its
membranes, leading to virtually complete extraction from costs (particularly with low patient volumes, <4 patients/
the blood. In the myocyte, ammonia equilibrates with the day). Due to the short half-lives of current PET radiopharma-
ionized ammonium. The ammonium ion is trapped in the ceuticals, only pharmacological stress can be performed with
myocyte due to activity of the enzyme glutamine synthase. these imaging protocols. On the other hand, the recently pro-
Ammonia has a first-pass myocardial extraction of 83% at posed PET-MPI radiotracer 18F-flurpiridaz, due to the longer
resting coronary blood flow. High-quality perfusion images half-life, is potentially able to overcome this latter limitation.
and gated images to determine ejection fraction and regional
wall can be obtained (see Fig. 20.5).
The [13N]ammonia should be administered in less than Key Learning Points
30 s and a delay after injection of 1.5–3 min should occur • A unique advantage of PET over other noninvasive
prior to initiation of imaging. However, if myocardial blood imaging modalities is the ability to measure abso-
flow is measured, imaging should begin immediately before lute tracer concentrations due to the very high sen-
injection to capture the input function. A sufficient differ- sitivity of PET instrumentation.
ence between stress and rest radiopharmaceutical counts • Advantages of PET with respect to SPECT for MPI
may be achieved by decay and/or administered activities. include markedly improved image quality, spatial/
An 18F-labelled tracer, 18F-flurpiridaz, is a structural ana- temporal resolution, better attenuation correction,
logue of the insecticide pyridaben, a known inhibitor of the and diagnostic accuracy.
nicotinamide adenine dinucleotide hydrate: ubiquinone
Anterior
• PET-MPI provides more detailed information for
risk stratification. Left anterior descending artery Left circumflex artery
• 82Rb and [13N]ammonia PET-MPI cannot be com- 100
bined with exercise stress testing due to the half-life
and clearance of these tracers.
% of maximum uptake
80
Lateral
Septal
60
20.2.7 MP Image Interpretation
40
Before the images are interpreted, data should be reviewed
for attenuation artifacts or zones of unexpected increased 20
activity that may alter the appearance of the myocardium.
Planar and SPECT images should be viewed on a computer Right coronary artery 0
Posterior
display to permit adjustment of contrast and brightness, opti-
mized to the myocardium. Before reconstruction, the SPECT Fig. 20.4 Schematic rendering of typical coronary anatomy overim-
projection data should be reviewed as a cine display to detect posed on left ventricle bull’s eye view. The left descending artery
patient motion. Significant patient motion during image (LDA), left circumflex (LCx), and right coronary artery (RCA) are the
acquisition may necessitate the reprocessing or reacquisition three main coronary arteries that perfuse the LV myocardium. The col-
ored polar map is a representative example image of relative tracer
of these studies. The data should be reconstructed using an uptake (% of maximum) in the LV (reproduced with permission from
iterative reconstruction algorithm. Perfusion PET images [33])
should be carefully reviewed to detect cardiac displacement
between rest and stress. On the [13N]ammonia perfusion
images, the radiotracer activity in the lateral wall of the left (SRS), which are global metrics of defect extent and severity
ventricular myocardium may be diminished in healthy vol- [34]. The summed difference score (SDS) is a global indica-
unteers, a factor that needs to be considered when patient tor of ischemia. SSS >3 is a commonly used threshold for the
studies are interpreted. presence of disease, while SRS >3 and SDS >1 indicate the
In the absence of artifacts, regional myocardial perfusion presence of nonreversible scar. A limitation of these scores is
is evaluated visually by comparing tracer uptake in different the dependence on segment alignment and visual assessment
regions of the myocardium against a reference region of (including overriding of automatically generated scores),
maximum uptake, under the assumption that it is normally which reduces their reproducibility (Figs. 20.5 and 20.6).
perfused. Regions with homogeneous uptake at both stress In the clinical setting, one of the main advantages of
and rest are considered as normal. Regions with relatively nuclear techniques over other modalities, such as stress echo-
reduced tracer uptake at stress but normalizing at rest are cardiography or cardiac MRI, is the development of standard-
interpreted as ischemic, while regions with relatively reduced ized methods for automated quantitation. Accordingly,
uptake at both rest and stress are interpreted as myocardial automated analysis of three-dimensional SPECT and PET
scar. The rest and stress images are reoriented to a standard images is now routine for both clinical and research purposes.
orientation that can be viewed as a series of short-axis slices Current software can automatically segment the LV, quantify
transecting the LV from apex to base. Orthogonal slices may LVEF, establish myocardial perfusion maps, and estimate
be displayed to generate horizontal long axis (HLA) and ver- global and local measures of stress/rest perfusion, all with
tical long axis (VLA) views of the heart. A more concise minimal user input. These methods have demonstrated better
presentation of the LV can be done by means of polar maps, reproducibility and at least similar diagnostic accuracy as
sampling the mid myocardium contour (bull’s eye). The apex qualitative visual analysis by expert readers. Perfusion
is presented at the center of the polar map with the base at the defects, fixed perfusion defects, LV function, LV volumes,
radial extreme; the septum and anterior, lateral, and posterior and regional wall motion and thickening can all be assessed
walls are on the left, top, right, and bottom, respectively, as with this software. Computer-based quantitation improves the
illustrated in Fig. 20.4 [33]. LV polar maps can be segmented consistency of interpretation [35]. A number of validated
into regions, which are each scored individually to generate software packages distributed by the main vendors of nuclear
a semiquantitative score of defect severity. Most commonly medicine imaging equipment are available for automated
a 17-segment model is used with scores ranging from 0 to 4 quantification (QPS/QGS, Emory Toolbox, 4D-MSPECT,
corresponding to normal perfusion, mild reduction, moder- and Wackers-Liu CQ) [36–39]. The basic principles are simi-
ate reduction, severe reduction, and an absence of tracer lar for each of these software packages: after segmentation of
uptake, respectively. Summation of the segment scores the LV, normalized relative radiotracer uptake in recon-
produces summed stress score (SSS) and summed rest scores structed slices is quantitatively compared to normal data files.
Fig. 20.5 Example of hybrid PET/CT MPI with [13N]ammonia, pro- branch. Fused PET and CCTA images revealed a perfusion defect
viding simultaneous assessment of comprehensive anatomical and downstream from the coronary stenosis. Invasive coronary angiography
functional information within a single scanning session. Case example showed angiographic significant luminal narrowing of the obtuse mar-
of a 46-year-old man with typical anginal chest pain. PET showed an ginal branch (reproduced with permission from: Driessen RS,
inferolateral perfusion defect with an abnormal hyperemic perfusion of Raijmakers PG, Stuijfzand WJ, Knaapen P. Myocardial perfusion imag-
1.89 mL min−1 × g−1 and a myocardial flow reserve of 1.75. CCTA ing with PET. Int J Cardiovasc Imaging. 2017;33:1021–31)
displayed an obstructive soft plaque located in the obtuse marginal
In the interpretation of functional MPI results, also the can be analyzed by means of various programs, mostly those
type of stressor should be considered. When exercise stress already extensively employed for gated SPECT.
triggers ischemia, clinical markers of ischemic threshold A wide array of criteria has been proposed for classifying
and exercise tolerance add diagnostic and prognostic infor- quantitative perfusion PET results as normal or abnormal. The
mation [11]. Conversely, both adenosine [40] and adenosine tracer used is the main variable. To obtain a correct CFR value,
A2A receptor agonists [41] exert a stimulus for maximal it is important to correct the baseline MBF for the resting rate
vasodilation [40]. pressure product. With [15O]water, previous studies have shown
The parameters that can be derived from quantitative that a CFR threshold of 2.5 is the most effective for identifying
PET-MPI approaches are resting and maximal myocardial patients with CAD defined as lumen narrowing of >50% on
blood flow (MBF) and their ratio, usually identified as coro- coronary angiography plus fractional flow reserve <0.8 [45].
nary flow reserve (CFR) or MBF reserve [42]. These param- For [13N]ammonia, the first proposed CFR threshold was 2.74;
eters are generally calculated for the whole left ventricular however, different thresholds were proposed. Surprisingly, for
myocardium and for each coronary territory. In studies per- the most used perfusion tracer, 82Rb, the thresholds are less
formed with [15O]water, myocardial perfusion images are not well defined. In a very large patient population, Johnson et al.
available but can be derived from MBF parametric images. identified the threshold of 1.74 for CFR [46]. No data on the
On the other hand, both [13N]ammonia and 82Rb provide clinical reliability of 18F-flurpiridaz for CFR quantification
perfusion images and their qualitative assessment [43].
have been published, as this tracer has not yet been clinically
A semiautomatic procedure based on comparison with a nor- approved. However, the feasibility of quantitative perfusion
mal database can also be used [44]. Additionally, gated PET imaging with this radiotracer has been demonstrated [47].
Fig. 20.6 Mismatch between

123
I-MIBG washout rate and
Perfusion MIBG Early MIBG Delayed
99m
Tc-tetrofosmin uptake in
DCM compared to ischemic HF
in vertical long axis myocardial
SPECT images. Upper panel
shows a case of DCM-HF in Case #1
which 99mTc-tetrofosmin stress/ C.L.
rest G-SPECT showed a DCM
homogenous perfusion in each LVEF 17%
myocardial wall, while Normal C.A.
123
I-MIBG SPECT washout rate
shows reduced tracer retention
particularly evident in lateral LV
wall. By contrast, lower panel
displays a case of ischemic HF in
which a match between a fixed
perfusion defect and reduced
123
I-MIBG retention area is
appreciable in the apex Case #2
(reproduced with permission B.B.
from [133]: Bauckneht M, CAD
Sambuceti G, Pomposelli E, Fiz LVEF 25%
F, Marini C. Pathophysiological Anterior MI
basis of myocardial innervation
imaging in heart failure. Clin
Transl Imaging. 2015;3:347–55)
20.2.8 Multimodality Imaging of Coronary

Key Learning Points Atherosclerotic Disease (Tables 20.5
• In the absence of artifacts, regional myocardial per- and 20.6)
fusion is evaluated visually by comparing tracer
uptake in different regions of the myocardium The simultaneous acquisition of MR images together with
against a reference region of maximum uptake, SPECT or PET might enable differentiation of epicardial and
which is assumed to be normally perfused. endocardial blood flow [48], which might vary in different
• Regions with homogeneous uptake at both stress patient groups such as arterial hypertension, diabetes melli-
and rest are considered as normal. tus, or kidney failure patients. On the other hand, MR imag-
• Regions with relatively reduced tracer uptake at ing might allow a detailed structural characterization of the
stress, normalizing at rest are interpreted as heart, enabling identification of myocardial fibrosis or scar
ischemic. tissue following myocardial infarction. As a consequence, the
• Regions with relatively reduced uptake at both rest addition of MR might help to differentiate between hibernat-
and stress are interpreted as myocardial scar. ing myocardium and myocardial scar in patients showing
• One of the main advantages of nuclear techniques fixed perfusion defects [49]. Finally, MR imaging represents
over other modalities, such as stress echocardiogra- the “gold standard” approach in the calculation of left ven-
phy or cardiac MRI, is the development of stan- tricular ejection fraction and wall motion and might improve
dardized methods for automated quantitation. the assessment of functional consequences of the ischemic
• Current software packages can automatically seg- burden during MPI. However, before its introduction into the
ment the LV, quantify LVEF, establish myocardial clinical practice, several technical issues should be solved.
perfusion maps, and estimate global and local
measures of stress/rest perfusion, all with minimal
user input. Key Learning Points
• The parameters that can be derived from quantita- • The most promising area of application for hybrid
tive PET-MPI approaches are resting and maximal myocardial combined SPECT/CT and PET/CT
myocardial blood flow (MBF) and their ratio, usu- imaging is CAD, in which functional modalities are
ally identified as coronary flow reserve (CFR) or able to accurately reflect coronary physiology, while
MBF reserve. cardiac CT is able to describe coronary anatomy.
Table 20.5 Strengths and limitations of different imaging methods in the evaluation of CAD
Diagnostic
Method accuracy Strengths Limitations Future perspectives
SPECT SE 87–89% Highly available and low costs Radiation exposure CZT cameras recently improved
SP 70–76% Extensive standardization and validation Poor spatial resolution resolution and image quality and
Radionuclides with long half-lives Long acquisition times reduced acquisition time and
MPI useful in patients who experienced Attenuation, motion, scatter, and radiation exposure
acute chest pain in the last 24 h partial volume artifacts
Applicable to patients with chronic
kidney disease
PET SE 90–92% Gold standard approach for MPI Radiation exposure Novel 18F-labelled tracers such
SP 81–88% Lower radiation exposure with respect High costs and limited availability as 18F-flurpiridaz might
to SPECT-MPI Poorer spatial resolution with respect overcome the major limitations
Absolute quantification of MBF to MDCT and MR of PET-MPI
Lower attenuation artifacts Short half-lives of radiotracers and
Assessment of myocardial viability required in-site generator/cyclotron
Short scan time
Applicable to patients with chronic
kidney disease
MDCT SE 82–99% High anatomical definition and spatial Radiation exposure, particularly in Potential simultaneous
SP 89–98% resolution case of MDCT perfusion imaging assessment of coronary anatomy
Short acquisition time and fast Iodinated contrast media and myocardial perfusion
post-processing time contraindications Potential clinical application in
Widely available and relatively low Pharmacologic stress only emergency department,
costs Artifacts provided by arrhythmias particularly in young patients
Quantification of coronary calcium and tachycardia, calcifications, and (high negative predictive value)
burden patient motion
No consensus of optimal scan
techniques and protocols
MRI SE 83–91% High spatial and temporal resolution Limited availability Novel T1-mapping sequences
SP 81–86% No ionizing radiation exposure Gadolinium media contraindications might provide better spatial
Absolute quantification of perfusion Pacemakers and ICD still limit the granularity, allowing the
Gold standard approach for noninvasive applicability differentiation of myocardial
evaluation of left ventricular ejection Claustrophobia tissue characteristics in patients
fraction, regional wall motion and Long acquisition times with diverse cardiomyopathies,
cardiac structure Lack of standardized sequences for without the need of contrast
Assessment of scar tissue, myocardial perfusion imaging media
edema, deposit, and inflammatory
pathologies
SE sensitivity, SP specificity
Table 20.6 Added value of hybrid imaging in the diagnosis of CAD

• The combined use of SPECT or PET with MDCT Higher diagnostic performance with respect to single imaging
provides a better assessment of myocardial perfu- modalities alone
sion defects together with co-localization to the cul- Combination of anatomical and functional information, potentially
prit coronary stenosis, directly influencing clinical able to reassign an entire perfusion defect to another coronary
territory [Liga R, Vontobel J, Rovai D, Marinelli M, Caselli C,
decision-making and prognosis. Pietila M, et al. Multicentre multi-device hybrid imaging study of
• Two different ways to obtain image co-registration can coronary artery disease: results from the EValuation of INtegrated
be used: Software- or hardware-based co-registration. Cardiac Imaging for the detection and characterization of
ischaemic heart disease (EVINCI) hybrid imaging population. Eur
Heart J Cardiovasc Imaging. 2016;17:951–60]
Identification of culprit lesion in multivessel disease
Combined evaluation of myocardial viability and coronary anatomy
20.3 Multimodality Imaging able to guide coronary revascularization
in Cardiovascular Infections Additional prognostic value
Cardiac infections include a group of conditions involving cases of implantable devices. Despite their relatively low
the myocardium, the pericardium, and/or the endocardial incidence, these conditions are associated with high morbid-
surface of the heart [50]. Infections can involve any foreign ity and mortality [51], involving a relevant burden of diag-
agent in the vasculature, including valves, stents, wires, and nostic workup. Moreover, the number of patients presenting
with suspected cardiac infections is progressively rising lonephritis, and peripheral stigmata, occur when IE remains
because of the increased use of prosthetic valve and cardio- undiagnosed for a long period. Vascular and immunological
vascular electronic device implants. Early diagnosis is cru- phenomena such as splinter hemorrhages and Roth’s spots
cial for adequate patient’s management, as early treatment (retinal hemorrhages with white or pale centers) are com-
improves the prognosis; unfortunately, the clinical manifes- mon. However, atypical presentations may occur in elderly
tations are often non-specific. patients or in immunocompromised patients.
Blood culture is the most important initial laboratory test.
If antibiotic therapy has been administered prior to the col-
20.3.1 Clinical Background on Infections lection of blood cultures, the rate of positive cultures
of the Cardiovascular System declines. In cases of suspected culture-negative infective
endocarditis (IE), other microbiological testing approaches
20.3.1.1 Infective Endocarditis may be useful. Echocardiography is the first-line imaging
Infective endocarditis (IE) is an infection of the endocardial modality that plays a key role in both the diagnosis and man-
surface of the heart that can involve the leads in case of agement of IE. In the latest update of the European Society
implanted devices. IE is a life-threatening disease, associated of Cardiology (ESC) Guideline for the management of IE
with a mortality rate of approximately 10% at initial admis- [60], multimodality imaging has been included in the diag-
sion which might rise up to 20% in the first year [52]. nostic algorithm. Therefore, along with blood cultures and
Although the overall incidence of the disease has remained echocardiography, which remains the first imaging test that
stable, ranging annually from 3 to 7 per 100,000 per year in plays a central role in both the diagnosis and the subsequent
the most recent population surveys [53–55], over the last clinical management of patients with IE, other multimodality
years, epidemiology of IE has become more complex, and imaging techniques have been introduced in the diagnostic
the epidemiological profile has changed substantially flowchart of IE, such as autologous radiolabelled leukocyte
because of several reasons: (1) a decrease in patients with (99mTc-HMPAO-WBC) scintigraphy and [18F]FDG PET/CT.
rheumatic heart disease, (2) an increasing percentage of In particular, molecular imaging techniques are useful to
elderly people often affected by several comorbidities, (3) confirm/exclude IE in case of “possible” or “rejected” IE
more frequent nosocomial infections, (4) increasing utiliza- (similarly as for fever of unknown origin) and to assess the
tion of prosthetic valves and intracardiac devices, (5) rising embolic burden in case of “definite” IE. The main added
number of subjects with intravenous drug addictions, and (6) value of these techniques is the reduction of the rate of mis-
increasing population of hemodialysis patients. diagnosed IE (classified in the “possible IE” category by
A microbiological shift from streptococci to staphylo- using the Duke criteria alone) and the detection of peripheral
cocci as the most frequent pathogens has also been reported embolic and metastatic infectious events. Evidence is higher
[56, 57]. New prophylaxis guidelines have been published, in case of prosthetic valve IE (PVE); however, accuracy of
new antimicrobial drugs are available, and bacterial resis- 99mTc-HMPAO-WBC scintigraphy and [18F]FDG PET/CT is
tances are increasing. high also in presence of native IE (NVE) and in case of
The clinical history of IE is highly variable according to inconclusive clinical findings. In particular, in case of PVE,
the causative microorganism, the presence or absence of pre- three imaging-based findings are now included as either
existing cardiac disease, the presence or absence of pros- major or minor criteria: (1) the identification of paravalvular
thetic valves or cardiac devices, and the presentation. IE may lesions by cardiac CT should be considered as a major crite-
present clinically as an acute, rapidly progressive infection rion; (2) in the setting of suspected PVE, abnormal uptake on
or as a subacute or chronic disease with low-grade fever and [18F]FDG PET/CT or 99mTc-HMPAO-WBC SPECT/CT
non-specific symptoms, which may thwart or confound ini- should be considered as a major criterion; and (3) the identi-
tial assessment [58]. Therefore, patients may be referred to a fication by imaging of recent embolic events or infectious
variety of specialists who may consider a range of alternative aneurysms (silent events) should be considered as a minor
diagnoses. criterion.
The diagnosis of IE is essentially clinical [59] and should
be suspected in all patients presenting with fever of unknown 20.3.1.2 Infections of Cardiovascular
origin, particularly when fever (up to 90% of the cases) is Implantable Electronic Devices (CIEDs)
associated with laboratory signs of infection, anemia, and The use of CIEDs has increased significantly over the last
microscopic hematuria and when septic embolic manifesta- decade. The number of devices that are implanted is esti-
tions are present (brain, lung, or spleen, in about 30% of the mated to exceed one million per year [61]. Simultaneously
cases). The main cardiac signs include heart murmur (up to with the rise in device implantations, the rate of infectious
85% of the cases) and progressive heart failure. Systemic complications is also increasing, by an estimated 5% per
signs, typically represented by spleen enlargement, glomeru- year [62]. The majority of CIED infections are caused by
either Staphylococcus aureus or coagulase-negative staphy- high of missing the presence and/or underestimating the
lococci; a variety of other bacteria and fungi are less com- extent of infection. Blood cultures are recommended in all
monly identified as causes of CIED infection. suspected cases of CIED infection, regardless of whether
CIED-associated infections cause significant morbidity and the patient is febrile or has other signs or symptoms of sys-
are burdened by a high death rate, particularly in case of temic infection. However, blood cultures may be negative
endovascular infection (20%) [63]. The incremental cost for despite CIED infection, particularly in patients with pocket-
managing CIED infection has been estimated to be about $ site infection and in those receiving antibiotics shortly
28,676 to $ 53,349, nearly half of this amount being due to before blood samples are drawn for culture. Moreover, posi-
intensive care procedures [64]. Furthermore, device replace- tive blood cultures may be due to a source other than an
ment procedures (that are periodically necessary for battery infected CIED. The likelihood of CIED infection when
depletion and/or for upgrading) are associated with infection blood cultures are positive varies according to the pathogen
rates higher than those occurring after initial implantation. detected, the number and duration of positive blood cultures,
The interval between CIED implant or revision and the and the presence of other findings that suggest device-
onset of infection varies widely, from days to years. The related infections [65].
clinical presentation of CIED infection depends on several TEE is recommended for patients with bacteremia, espe-
factors, including the site of infection (e.g., generator’s cially if the bloodstream infection is due to staphylococcal
pocket versus intravascular leads or epicardial leads), the species or if the source is not identified, and for patients with
type of microorganism, and the origin of the infection (e.g., signs of systemic infection, regardless of the results of blood
pocket erosion, localized infection of the generator’s pocket, culture [66]. The main purpose of TEE is to identify compli-
bacteremia from a remote site). Early infections, i.e., those cations such as valvular vegetations or myocardial or peri-
occurring within a few months after implantation, manifest valvular abscesses. In adults, TEE is more sensitive than
as acute or subacute infections of the pulse-generator’s TTE for detecting signs of an intracardiac infection. However,
pocket. Bacteremia may occur even without clinical signs false-negative and false-positive echocardiographic studies
and symptoms. Fever is the most common finding and are not rare, and the Duke criteria are difficult to apply in
remains the only sign in approximately 33% of the patients. these patients because of lower sensitivity than in the case of
Although some cases of CIED infection present without IE, even when the modified Duke criteria are used. Because
obvious inflammatory changes of the skin, the diagnosis is of the frequently difficult diagnosis of the disease, and
most often (about 70% of cases) based on findings at the because of some limitations of echocardiography, multimo-
generator’s pocket site, including local pain, swelling, red- dality imaging has been successfully applied in patients with
ness, drainage, and skin and soft-tissue ulceration. The first CIEDs infections [67].
sign of infection may be erosion through the skin at the site
of the generator’s pocket, with external exposure of the gen- 20.3.1.3 I nfections of Implantable Left
erator, of one or more leads, or of both the generator and Ventricular Assist Devices (LVADs)
leads, with or without local inflammatory changes. Late LVAD represents a major medical development for end-stage
infections occur up to several years after implantation or heart failure in selected patients. This treatment is currently
reimplantation and have much more subtle manifestations. used as a bridge-to-transplantation, as a bridge-to-recovery,
Most often the transvenous or epicardial leads are involved; or as destination therapy as the last resort in patients with
in the latter circumstance, complications such as pericarditis, neither perspectives of recovery nor heart transplant [68].
mediastinitis, and right-sided endocarditis are often present. Implantable LVADs intended for long-term use rely on a per-
The diagnostic workup of CIED infections is problem- cutaneous driveline, to carry electric signals and energy from
atic, since patients can present with a variety of manifesta- the controller and batteries to the implanted pump. As with
tions including subtle signs of systemic or local infection. any other implantable foreign device, LVAD-related infec-
Final diagnosis of CIED infection is generally based on tions may occur, with a 23–58% prevalence and an associ-
microbiological tests (blood cultures and culture of material ated high mortality rate (15–44%). The major sites of
from exposed sites of the device) and ultrasound evaluation infection include the mediastinum drivelines and device sur-
of the cardiac region (either transthoracic echocardiogra- face, identified as LVAD endocarditis [69].
phy, TTE, and/or transesophageal echocardiography, TEE) The major pathogens involved in these emerging foreign
and of the venous pathway of the device. The above tests device-related infectious diseases are the “big five”:
constitute the basis for defining the patients’ likelihood to Staphylococcus aureus, Enterobacteriaceae, Pseudomonas
have CIED according to the Duke criteria [59]. However, in aeruginosa, coagulase-negative staphylococci, and
case of CIED infections, the Duke criteria, originally devel- Corynebacterium spp. [70]. Due to the lack of specific guide-
oped for the diagnosis of IE, may be inadequate; even with lines, the clinical management of LVAD infections is not
the addition of clinical parameters, the possibility remains standardized and is mainly derived from the available
recommendations for other CIED infections (prosthetic The 99mTc-HMPAO-WBC scans are classified as positive
valves or vascular prosthesis), although their characteristicsfor infection when at least one focus of abnormal uptake/
differ significantly. The only available specific recommenda- accumulation characterized by time-dependent increase in
tion to assist therapeutic decisions (i.e., the use of antimicro-
radioactivity from early planar to delayed images is observed.
bial treatment and surgery) in this challenging context is In case of IE, SPECT/CT acquisition overcomes the failure
based on observational studies and experts’ opinion [71]. of planar images alone to detect the site and extent of infec-
Regarding the diagnostic workup of patients with sus- tions in the cardiovascular system. Therefore, in this condi-
pected LVAD infections, the use of CT as main diagnostic tion SPECT/CT images confirm and localize findings seen
imaging is based on the possibility to detect the presence of with planar imaging and increase diagnostic accuracy. If
edema as primary sign of infection, a finding that is often semiquantitative evaluation of the 99mTc-HMPAO-WBC scan
highly non-specific. is used for diagnostic purposes, it is very important that the
planar and SPECT/CT images are corrected for isotope
decay. SPECT/CT images should cover the thorax in case of
Key Learning Points IE and the thorax-upper abdominal area in case of CIEDs
• Infective endocarditis is a potentially fatal disease and LVAD infections, ensuring that all components of the
whose incidence is growing in parallel with chang- device are included in the field of view, considering all the
ing lifestyles (e.g., intravenous drug addiction) and possible generator positions (i.e., abdomen in some cases)
with increasing application of invasive treatments [72, 73].
(e.g., cardiac prosthetic valves and intracardiac When present, focal uptake indicating infection is further
devices, hemodialysis). classified as pertaining to the heart (Fig. 20.7) and/or to
• Although diagnosis of infective endocarditis based extracardiac sites (Fig. 20.8) by SPECT/CT. In this respect,
on clinical criteria alone (including blood cultures with 99mTc-HMPAO-WBC scintigraphy sites of embolism
and echocardiography) is sufficiently reliable in the might appear as areas of increased uptake/accumulation as in
majority of patients with infection affecting native the case of the lung (hot spot), whereas for septic emboli in
cardiac valves, in the case of prosthetic valve infec- the spleen or spine, the typical finding appears as a cold spot.
tion or in the presence of intracardiac devices, accu- However, the detection of cold spots is not itself indicative of
rate diagnosis requires a combination of imaging septic embolism, since it might be present in a number of
modalities—especially imaging based on radionu- other clinical conditions (i.e., metastasis, angiomas, verte-
clide techniques. bral crushes); therefore, this image finding needs further
• The bacterial strains more frequently causing infec- confirmation by CT or MR.
tive endocarditis are Staphylococcus aureus, Analysis of the SPECT/CT images includes visual inspec-
Enterobacteriaceae, Pseudomonas aeruginosa, coag- tion to exclude mis-registration between the SPECT and the
ulase-negative staphylococci, and Corynebacterium CT components and side-by-side inspection of both CT
spp., frequently in antibiotic-resistant forms. attenuation-corrected and non-attenuation-corrected images,
to minimize metal-related artifacts. SPECT/CT is mandatory
to correctly interpret and localize the site and extent of 99mTc-
HMPAO- WBC uptake/accumulation indicating infection.
20.3.2 Radionuclide Imaging Techniques The images must be reconstructed with and without attenua-
in Cardiac Infections tion correction to identify potential reconstruction artifacts.
False-positive findings have been described for 99mTc-
20.3.2.1 Overall Background on Scintigraphy HMPAO-WBC imaging in IE and in CIED infections, even
with Radiolabelled Autologous in case of very early infections. On the other hand, false-
Leukocytes (99mTc-HMPAO-WBC) negative scans have been observed in the presence of IE
99m
Tc-HMPAO-WBC scintigraphy consists of sequential caused by some specific organisms [72]. The same limitation
acquisitions that include whole-body scan and spot planar must always be considered in case of CIED infections, in
images of the thorax and of any additional area of interest at particular in the presence of very small vegetation(s) along
30 min (early imaging), 4–6 h (late imaging), and 20–24 h the electrocatheter.
(delayed imaging, if needed) after the reinfusion of 370– Disadvantages of scintigraphy with radiolabelled WBC
555 MBq of autologous leukocytes radiolabelled with 99mTc- are the requirement of blood handling for radiopharmaceuti-
HMPAO. SPECT/CT of the chest-upper abdomen is cal preparation, duration of the procedure that is more time
generally obtained at 4–6 h and repeated at 24 h in case of consuming as compared to PET/CT, and a lower spatial
negative or doubtful imaging at 6 h. If necessary, additional resolution and photon detection efficiency compared with
SPECT/CT images might be acquired at the same time point. PET/CT.
Fig. 20.7 99mTc-HMPAO- a b 30 min

WBC scintigraphy in a patient
with aortic and mitral
mechanical prosthesis. The
patient presented with fever,
increased CRP and ESR, and
negative rheumatoid factor.
Echocardiography (both
transthoracic and
transesophageal) was
negative. Blood culture was 6h
positive for Enterococcus
faecalis. SPECT/CT shows a
focal area of increased uptake
in the perivalvular aortic
region, at the medial aspect
(from left to right MIP,
transaxial, coronal, and
sagittal view and
reconstructed 3D images,
respectively). The added value
of SPECT/CT images is
20 h
obvious, since planar images
alone are not able to diagnose
IE due to the activity of the
sternum that covers the
valvular uptake
20.3.2.2 O verall Background on [18F]FDG yocardial uptake of [18F]FDG include blood glucose levels
m
PET/CT and a low-carbohydrate diet. In particular, while age and
Performing an [18F]FDG PET/CT for cardiovascular infec- fasting time do not affect physiologic [18F]FDG uptake in the
tions is more complex than a simple translation of the stan- myocardium, blood glucose levels may have a nonlinear
dard protocols commonly employed in oncology. Starting effect.
from patient preparation, some specific issues of the imaging Physiological [18F]FDG uptake by the left ventricular
protocol and imaging reading should be considered [74]. myocardium can hamper image interpretation due to the
Patient preparation is very important to minimize the blurring effect caused by cardiac cycle and respiratory activ-
physiological uptake of [18F]FDG in the myocardium. In ity. Similarly, the marked and often unpredictable heteroge-
fact, variable focal of diffuse physiologic [18F]FDG uptake is neity of its distribution can either mask or potentially mimic
often observed in the normal myocardium of fasting nondia- the presence of a local infection. Minimizing glucose
betic patients (6–12 h to overnight) with normal glucose lev- metabolism of myocardial cells as much as possible is thus
els. Uptake of [18F]FDG is most notable in the left ventricular mandatory. To achieve this, a prolonged reduction in sugar
myocardium, which has a greater muscle mass than other intake decreases serum glucose and insulin levels, markedly
cardiac chambers. Uptake in the wall of the right ventricle is lowering myocardial glucose consumption by reducing
typically equal to or less intense than that in the left ventricu- GLUT4 docking to the sarcolemma. Metabolism of inflam-
lar myocardium; uptake in the wall of the right and left matory infiltrates, on the other hand, remains relatively sta-
atria is usually not detected. Factors possibly influencing ble, since transmembrane glucose transport in white blood

different patterns of uptake in
patients with IE and septic
embolism. Upper panel:
spleen and spine metastatic
infections at [18F]FDG PET/
CT, showing increased tracer
uptake in the spleen
corresponding to a wedge-
shaped low CT attenuation
(axial fused and CT sections
on the left) and to increased
uptake in the spine. Lower
panel: 99mTc-HMPAO-WBC
SPECT/CT imaging showing
focal “cold” areas in the
spleen (left) and in the spine,
due to the physiological high
accumulation of the
radiolabelled WBCs both in
the spleen and in the spine
cells relies on the constitutive availability of GLUT1 and The relevance of this pathway in allowing an adequate
GLUT3. Cardiomyocyte intracellular nutrient asset thus “infection-to-background” ratio has been extensively vali-
shifts toward a reduction in glucose availability, facing an dated: preceding [18F]FDG injection by 12 h of
increased concentration in free fatty acid derivatives. This carbohydrate-restricted, very high-fat, and protein-allowed
modification switches cardiac metabolism toward beta-oxi- diet followed by 12-h fasting period results in an “absent”
dation and activates the peroxisome proliferator-activated cardiac uptake in about 55% of patients compared to about
receptor-α that transcriptionally promotes expression of 30% of subjects studied with the standard 6-h fasting pro-
enzymes governing fatty acid metabolism and inhibits that tocol. Adding a single bolus of heparin (50 IU/kg 15 min
of glycolytic enzymes, thus causing a further reduction in prior to [18F]FDG administration) further increases this
cardiomyocyte utilization of glucose. response up to about 90% [75], probably due to the
Table 20.7 Methods to improve glucose utilization by normal Table 20.8 Methods to suppress glucose utilization by normal
myocardium myocardium
Oral glucose load Oral glucose load of 25–100 g. Serum Method Technique
protocols glucose level (SGL) should be measured Prolonged fast Fasting state for 12–18 h
before and after 60 min. If a 60-min SGL High-fat/ Two meals 24 h before the study, followed by an
ranges between 100 and 140 mg/dL, [18F] low- overnight fast. High fat and protein with low/no
FDG injection is recommended. In case of a carbohydrate carbohydrate content meal are allowed
60-min SGL >140 mg/dL, 2 IU of regular diet
insulin should be injected before [18F]FDG
Intravenous 15–50 IU of unfractionated heparin 15 min prior
Intravenous glucose Intravenous glucose administration can be unfractionated to [18F]FDG administration or 500 IU 45 and
administration considered; however, it is not recommended heparin 15 min prior to [18F]FDG administration (total
in case of SGL >120 mg/dL and in diabetic 1000 IU). Ensure patient has no
patients. 13–25 g of 50% dextrose solution contraindications to administration of
should be administered depending on intravenous heparin including bleeding
SGL. After 30–60 min, [18F]FDG should be tendencies, allergies, or history of heparin-
injected induced thrombocytopenia with thrombosis
Hyperinsulinemic- This protocol can be used both in diabetic Combined High-fat/low-carbohydrate diet for two meals,
euglycemic clamp and nondiabetic patients, and it is based on methods 1 day prior, followed by overnight fast, and
the simultaneous infusion of insulin and intravenous regular heparin before [18F]FDG
glucose acting on the tissue as a metabolic administration
challenge in order to maximize myocardial
[18F]FDG uptake. After SGL estimation, Adapted from [26]
insulin should be injected in declining
infusion rates starting with 0.16 IU/min/kg The [18F]FDG activity recommended in the joint EANM/
for 4 min, followed by 0.08 IU/min/kg for
SNMMI guidelines on PET imaging in inflammation/infec-
3 min and 0.04 IU/min/kg at 8 min.
Co-infusion of 20% dextrose should be tion is 2.5–5.0 MBq/kg (175–350 MBq for a 70-kg standard
carried out in order to achieve a constant, adult) [77]. Although antimicrobial treatment is expected to
euglycemic blood glucose concentration. decrease the intensity of [18F]FDG accumulation, there is no
[18F]FDG can be injected when SGL reaches
evidence at this stage to routinely recommend treatment dis-
the range 80–110 mg/dL. For the best image
quality, insulin and dextrose infusion should continuation before performing PET/CT. Whereas, cortico-
be maintained for 30 min steroid drugs should be discontinued or at least reduced to
Acipimox oral This compound induces the inhibition of the the lowest possible dose in the 24 h preceding the scan.
administration peripheral lipolysis, thus decreasing free Image acquisition generally starts after an uptake time of
fatty acid plasma levels. The lack of this
energetic substrate shifts myocardial 45–60 min, and the emission time per bed position depends
metabolism toward preferential glucose on sensitivity of the scanner. As in oncology, the field of
utilization (glucose switch). The usual acquisition generally includes from the skull base to mid
protocol suggests oral 250 mg acipimox thighs (total body). Whole-body images including the lower
administration 90–60 min prior to [18F]FDG
injection, after overnight fasting [26] limbs might be suggested to detect complications of IE such
SGL serum glucose level
as mycotic aneurysms that may require specific treatment by
embolization to prevent rupture. An additional separate bed
on the cardiac region is useful to record gated images.
lipolytic activity of heparin that can induce a fivefold Diagnostic CT angiography (CTA) imaging might also be
increase in the blood free fatty acid levels. Although still performed, to maximize the diagnostic information provided
suboptimal, this feature actually approaches the clinical by the exam. Although delayed imaging has been proposed
applicability. Nevertheless, repeatability of this procedure to increase specificity in diagnosing infection of cardiovas-
has not been thoroughly tested, particularly in PVE cular implants, recent data suggested that in IE delayed
patients in whom a sugar-free diet has been maintained images are more prone to false-positive results.
either for 12 h without any further intervention or for 48 h Also in case of [18F]FDG PET/CT, image reconstruction
and combined with heparin administration. Accordingly, with and without CT-based attenuation correction is recom-
further studies to optimize dietary preparation for this pro- mended to identify potential reconstruction artifacts.
cedure to optimize the diagnostic potential of [18F]FDG Techniques for metal artifact reduction are useful to mini-
PET/CT imaging in PVE diagnosis. Dietary and/or phar- mize overcorrection, even if they do not always recover com-
macologic preparations prior to myocardial [18F]FDG PET pletely quality of the PET images. PET/CT images must be
imaging have been proposed [76], as summarized in Tables visually evaluated for increased [18F]FDG uptake, taking into
20.7 and 20.8. consideration the pattern (focal, linear, diffuse), the intensity,
and the relationship to areas of physiologic distribution. PET and 93–100% negative predictive values, even in the early
information is compared with morphologic information post-intervention phase [79, 80]. In addition, the intensity of
obtained by CT and, possibly, CTA. 99m
Tc-HMPAO-WBC accumulation in the perivalvular area
A physiological variant that might constitute a confound- represents an interesting marker of local infectious activity, in
ing factor while reading the images is the presence of increased that patients with a mild activity on the first exam disappearing
metabolic activity along the posterior part of the heart, where on the second imaging evaluation seem to have a favorable
lipomatous hypertrophy of the interatrial septum may appear outcome [79]. This observation opens the very interesting per-
as a fat-containing mass with increased [18F]FDG uptake. spective for the use of molecular multimodality imaging for
Patients who have recently undergone cardiac surgery might the assessment of antimicrobial treatment response.
present postoperative inflammation that results in non-specific In a recent systematic review on the assessment of PVE,
[18F]FDG uptake in the immediate postoperative period. To [18F]FDG PET/CT sensitivity and specificity have been
minimize the risk of false-positive findings, the European reported to be 73–100% and 71–100%, respectively, with
Society of Cardiology Guidelines recommend not utilizing 67–100% positive and 50–100% negative predictive values,
[18F]FDG PET in the 3-month period following valve implanta- respectively [80]. Addition of [18F]FDG PET/CT to the mod-
tion. If the scan is performed early after surgery, the possibility ified Duke criteria increased sensitivity for definite IE from
of false-positive results should be taken into account. 52–70% to 91–97%, by reducing the number of possible
A number of additional conditions can mimic the pattern PVE cases [81]. When [18F]FDG PET/CT is associated with
of focally increased [18F]FDG uptake that is typically CT angiography ([18F]FDG PET/CTA), both sensitivity and
observed in IE, such as active thrombi, soft atherosclerotic specificity for the diagnosis of IE increased to 91%, with
plaques, vasculitis, primary cardiac tumors, cardiac metasta- 93% positive and 88% negative predictive values [82]. In
sis from a non-cardiac tumor, postsurgical inflammation, and association with the Duke criteria, [18F]FDG PET/CTA
foreign body reactions. allowed reclassification of 90% of the cases initially classi-
fied as “possible” IE and provided a more conclusive diagno-
sis (definite/reject) in 95% of the patients. This combined
Key Learning Points multimodality procedure should be considered in all the
• Scintigraphy with radiolabelled autologous leuko- patients where echocardiography presents significant limita-
cytes (99mTc-HMPAO-WBC) is the cornerstone of tions. In fact, its ability to provide relevant information on
radionuclide imaging of cardiac infections. the local extent of the disease such as the presence of pseu-
• SPECT/CT imaging is crucial for increased speci- doaneurysms, fistulas, thrombosis, and coronary involve-
ficity and for accurately localizing the foci of ment is significant for the subsequent clinical and surgical
infection. decision-making. In addition, by adding CTA to PET/CT in
• PET/CT with [18F]FDG for imaging cardiac infec- IE patients, it is possible to assess the entire chest, therefore
tions is an especially attractive alternative option, to identify septic pulmonary infarcts and abscesses, as well
thanks to better spatial resolution and less cumber- as to evaluate the aorta and the coronary arteries in the per-
some labelling procedure versus 99mTc-HMPAO- spective of surgery. Figures 20.9 and 20.10 show examples
WBC scintigraphy. of [18F]FDG PET/CT and [18F]FDG PET/CTA contribution
• When performing [18F]FDG PET/CT in patients to correct diagnosis in patients with suspected IE after
with suspected cardiac infection, special attention implantation of prosthetic valves.
must be paid to preparation of the patient so to min-
imize physiological myocardial uptake of [18F]
FDG. 20.3.4 Radionuclide Imaging of Native Valve
Endocarditis
Despite the fact that imaging interpretation might be more

20.3.3 Radionuclide Imaging of Prosthetic straightforward than in patients with PVE, the diagnostic
Valve Endocarditis value of [18F]FDG PET/CT has not been well determined in
case of native valve endocarditis (NVE). In fact, most studies
The sensitivity of 99mTc-HMPAO-WBC SPECT/CT for pros- included mainly PVE or a mixed patient population with
thetic valve endocarditis has been reported at an overall both native and prosthetic valves. The reported low sensitiv-
64–90%, with 36–100% specificity, and 85–100% positive ity of [18F]FDG PET/CT in NVE is likely to be mainly related
and 47–81% negative predictive values [72, 74, 78]. In case of to the location and size of the lesions. In NVE the presence
abscess formation, 99mTc-HMPAO-WBC SPECT/CT has of a vegetation is the main finding, at least in the initial phase
83–100% sensitivity, 78–87% specificity, and 43–71% p ositive of the disease. It should also be noticed that this is a clinical
Fig. 20.9 [18F]FDG PET/CT in a 50-year-old man with an aortic bio- of myocardial [18F]FDG uptake, PET/CT clearly shows a focal area of
logical prosthesis positioned 2 years earlier. The patient developed increased uptake in the perivalvular region, at the medial aspect (upper
fever, increased ESR, and positive blood culture for Enterococcus fae- panel, from left to right transaxial CT, PET emission, and fused PET/
calis. Echocardiography was negative. Despite suboptimal suppression CT images)
a b
c d
Fig. 20.10 Examples of postoperative inflammation versus infective aortic valve replacement 1 month earlier. Fused PET/CTA thin valve leaf-
endocarditis in prosthetic aortic valves as evaluated by [18F]FDG PET/CT lets and mild, homogeneous perivalvular [18F]FDG uptake (SUVmax 1.58),
and CT angiography (CTA). (a) A 38-year-old man submitted to aortic similar to activity in the blood pool. (d) A 79-year-old woman with early
valve replacement with a bileaflet mechanical prosthesis. Fused PET/CTA (3 months postimplantation) bioprosthetic aortic valve infective endocar-
transverse view of the aortic valve shows a normally inserted aortic pros- ditis. Fused PET/CTA shows intense (SUVmax 13), heterogeneous peri-
thesis with mild, homogeneous valvular inflammatory reaction (SUVmax prosthetic [18F]FDG uptake (arrows), thickened valve leaflets, and a
3.1) (arrows). (b) A 56-year-old man with mechanical aortic valve infec- multiloculated perivalvular pseudoaneurysm (asterisk) (reproduced from:
tive endocarditis (valve implantation 3 years previously). PET/CTA shows Pizzi MN, Roque A, Cuéllar-Calabria H, Fernández-Hidalgo N, Ferreira-
intense (SUVmax 13.33) periprosthetic [18F]FDG uptake and a poorly González I, González-Alujas MT, et al. 18F-FDG-PET/CTA of prosthetic
delimited perivalvular soft tissue lesion, consistent with a periprosthetic cardiac valves and valve-tube grafts: infective versus inflammatory pat-
abscess (arrowheads). (c) A 78-year-old man submitted to bioprosthetic terns. JACC Cardiovasc Imaging. 2016;9:1224–1227)
setting where echocardiography has a very high accuracy; ficity of 88% (95% CI, 77–94%), and AUC of 0.861 [83].
therefore, the use of multimodality imaging is reserved to a Such a finding is mainly related to the small size of the veg-
limited number of patients, i.e., patients with severe valve etations along the leads, which are often well below spatial
calcific degeneration with suboptimal acoustic window. resolution of the system.
[18F]FDG PET/CT and 99mTc-HMPAO-WBC scan findings
in association with Duke criteria also allowed reclassifying
20.3.5 Radionuclide Imaging most of cases initially classified as “possible” IE, distinguish-
of CIED Infections ing infection limited to the pocket or leads from a more severe
infection affecting the whole device and identifying patients
Similarly as described above for PVE, 99mTc-HMPAO-WBC requiring device extraction. Semiquantitative parameters such
SPECT/CT and [18F]FDG PET/CT can be used to confirm/ as semiquantitative ratio of maximum count rate of the pocket
exclude infection and characterize the extension of the infec- device over mean count rate of lung parenchyma or normaliza-
tious process, including extracardiac workup. The diagnosis tion of SUVmax around the CIEDs to the mean hepatic blood
of local infections is quite straightforward. A recent meta- pool ratio activity might help in differentiated mild postopera-
analysis provides pooled specificity and sensitivity in this tive residual inflammation (lasting up to 2 months after device
subgroup of 93% (95% CI, 84–98%) and 98% (95% CI, implantation) versus infection. Finally, those patients with
88–100%), respectively, and AUC was 0.98 for [18F]FDG suspected infection but without [18F]FDG uptake have been
PET/CT. The largest study with 99mTc-HMPAO-WBC scin- shown to have a favorable outcome following antibiotic ther-
tigraphy reported a sensitivity of 94%, with 100% specific- apy, thus suggesting the absence of bacterial colonization of
ity. The diagnostic accuracy for lead infections is lower, with CIEDs. Figure 20.11 shows an example of the use of [18F]
overall pooled sensitivity of 65% (95% CI, 53–76%), speci- FDG PET/CT in a patient with CIED infection.
Fig. 20.11 [18F]FDG PET/CT in a patient previously submitted to PET/CT images from left to right). Infection involved also the intra-
CIED implantation, with fever lasting since several weeks with mild vascular portion of the electrocatheters (middle row), as well as the
increased CRP and ESR. Echocardiography was negative. intracardiac portion (bottom row). Based on these imaging findings,
Antimicrobial treatment was initiated. Increased [18F]FDG uptake the patient was treated with removal of the device and prolonged anti-
around the pocket (upper row: transaxial CT, emission PET, and fused microbial treatment
20.3.6 Radionuclide Imaging of LVAD-Related Early detection of embolic events has been reported using
Infections [18F]FDG PET/CT with a high sensitivity (87–100%) and
specificity (80%) [80] at a reasonable cost-effectiveness
The usefulness of 99mTc-HMPAO-WBC SPECT/CT and of ratio, especially in patients with Gram-positive bacteremia
[18F]FDG PET/CT in the diagnosis of LVAD-related infec- [88]. Extra-cerebral peripheral localizations of IE were
tion has been shown in small patient groups under routine found in 24–74% among the definite IE population; most of
clinical conditions. Molecular imaging allows precise ana- these peripheral localizations were silent (50–71%) and
tomic location and accurate extent of a suspected infection revealed by [18F]FDG PET/CT. In a case-control study, [18F]
with sensitivity of 100% and a specificity of 94% in case of FDG PET/CT detected peripheral localizations in 57.4% of
[18F]FDG PET/CT [84]. The use of the metabolic volume has IE patients, representing the only initially positive imaging
been recently reported to be associated with increased diag- technique in about half of the patients with embolic events
nostic accuracy as compared to the SUVmax index in a series [89]. Detection of metastatic infection by [18F]FDG PET/CT
of 48 patients. In particular, the negative predictive value and led to change of treatment in up to 35% of patients [90], with
sensitivity increased up to >95% by using the metabolic vol- a twofold reduction in the number of relapses [89]. [18F]FDG
ume compared to 87.5% when using SUVmax [85]. PET/CT is very accurate in organs with low physiological
uptake, therefore not applicable in ruling out the presence of
brain embolism [91], where CT/MRI are more sensitive and
20.3.7 Extracardiac Workup of Patients specific. 99mTc-HMPAO-WBC SPECT/CT shares with PET/
with Infective Endocarditis CT the possibility of acquiring whole-body images. [73].
Also in the case of CIED infections, accurate evaluation
The most frequent extracardiac manifestations of either of the whole-body imaging might detect septic embolisms
native valve and prosthetic valve endocarditis (reported in and identify the possible infection portal of entry, impacting
30–80% of patients) are embolic stroke or septic emboliza- on the subsequent therapeutic management and reducing the
tion to the bone, spleen, or kidneys, although only some of risk of relapse. Indeed, in CIED infection the detection of
these are symptomatic [86]. The majority of embolisms lung embolisms, considered as a major criterion of the Duke
occur within the first 14 days after treatment initiation; nev- score, has shown to increase the diagnostic sensitivity.
ertheless, they may appear as the initial symptom leading to
the diagnosis and frequently are recurrent [87]. The localiza- 20.3.7.2 I dentification of the Infection Portal
tion of the emboli and their cerebral/extracerebral proportion of Entry
vary according to the studies, in particular according to the The extracardiac diagnostic workup in patients with IE is
frequency and modalities of imaging and the proportion of important not only for the identification of sites of embolism
right-sided and left-sided IE. and metastatic infections but also for the identification of the
infection portal of entry. Locating the portal of entry lowers
20.3.7.1 Detection of Embolic Events the risk of recurrence in patients with IE. In a recent study,
The search for asymptomatic embolic events through sys- systematic search for the portal of entry identified the site of
tematic extracardiac imaging has become a very important primary infection in 74% of patients, mainly cutaneous (40%),
issue especially after inclusion of the detection of asymp- followed by oral or dental (29%) and gastrointestinal (23%)
tomatic embolic events as a minor Duke criterion in the [92]. [18F]FDG PET/CT reveals the source of infection in most
2015 ESC criteria. The panel of imaging modalities used of the patients, including cases where the portal of entry was a
routinely to evaluate patients with extracardiac infective neoplasm (colonic cancer) [82]. Several studies support the
processes includes dental radiography, abdominal ultra- association between bacteremia or IE due to Streptococcus
sound, CT scan of the brain, whole-body CT, or MRI scan. gallolyticus and Streptococcus infantarius and gastrointestinal
CT scan (including cerebral) has long been considered the disease, mostly represented by colon cancer, adenomatous
main imaging technique for the diagnosis of embolic events polyps, diverticulosis, and biliary lesions [93]. Multiple por-
in IE patients; MRI is a valuable alternative in case of cere- tals of entry are also possible. Once the portal of entry has
bral embolism, having the advantage of a higher sensitivity been identified, risk modification can be performed.
in detecting recent ischemic lesions and small ischemic or
hemorrhagic lesions without the need of administering an
iodinated contrast medium. Both [18F]FDG PET/CT and 20.3.8 Clinical Background on Vascular
99m
Tc-HMPAO- WBC SPECT/CT enable whole-body Prosthetic Infections
exploration and therefore the possibility of revealing with a
single imaging procedure both cardiac and extracardiac Vascular prosthetic infection (VPI) is the most serious com-
sites of infection. plication following surgical or endovascular implantation.
Despite a relative low incidence (between 0.5% and 5%) [94], signal intensity differences between T1- and T2-weighted
it represents one of the most challenging conditions with very images. However, MRI shares the same limitations as CT
high morbidity and mortality rates (around 50% and 25–75%, imaging in the early postoperative period because of non-
respectively). It is more common in the inguinal region (about specific signal abnormalities in the vicinity of the perigraft
13%) followed by aorto-bifemoral bypass and femoropopli- [98]. Nevertheless, MRI accuracy increases to 90–95% when
teal bypass. Despite the shift in trends of aortic repair in the it is performed 3–4 months after surgery.
past two decades with increased proportion of endovascular
repairs, the expected reduction in the rates of VPI did not 20.3.8.1 R adionuclide Imaging of Vascular
occur. Infection might arise as a consequence of vascular Prosthetic Infections
graft or stent-graft perioperative contamination or during bac- Nuclear medicine techniques have generally been reserved
teremia with consequent seeding of the implant material. for cases with equivocal conventional imaging findings or
Mechanical erosion of the graft/stent-graft into an adjacent for patients managed by high-expertise multidisciplinary
structure (esophagus, bronchial system, and duodenum) is a groups including nuclear medicine physicians. Molecular
very rare but challenging complication. Depending on the multimodality imaging has demonstrated high accuracy in
timing of clinical presentation, prosthetic graft infection can detecting graft infection in patients with aortic graft and
be classified as early versus late or very late infections, the without specific signs of infection (low-grade phases).
threshold for late infection being 4–6 months after the pri- Overall, 99mTc-HMPAO-WBC imaging has sensitivity
mary surgery or endovascular intervention. Late infection is ranging from 82% to 100%, with 75–100% specificity. The
more indolent and does not present with signs of septicemia; extensive use of SPECT/CT allows accurate characterization
one of the most suggesting signs is the absence of graft incor- of abnormal foci of 99mTc-HMPAO-WBC accumulation, par-
poration with surrounding tissues and the presence of peri- ticularly by visualizing extension of the lesion and by con-
graft fluid (and gas particles) containing large amounts of firming or rejecting graft involvement even in the presence of
leukocytes. The majority of cases of VPI are due to postsurgical distortions and in complex anatomical sites. In
Staphylococcus aureus, Escherichia coli, and Staphylococcus this regard, the use of SPECT/CT is associated with a signifi-
epidermidis, whereas Klebsiella, Pseudomonas, Enterobacter, cant reduction in the false-positive results of planar imaging
and Proteus account for most of the remaining portion [95]. (i.e., abdominal non-specific accumulation). High specificity
Diagnosis of VPI is difficult, since no single diagnostic is maintained even when scintigraphy is performed during
procedure has 100% of accuracy; therefore, a combination of the first month after surgery [99] and also in case of late
physical examination, laboratory tests, and several imaging low-grade VPI [100]. Figures 20.12 and 20.13 show exam-
techniques is mandatory. In fact, patients may report a vari- ples of the use of 99mTc-HMPAO-WBC imaging (planar and
ety of clinically equivocal complaints. Furthermore, blood SPECT/CT) in patients with VPI infections.
chemistry parameters can only show moderately elevated [18F]FDG PET/CT has emerged as a valuable tool for the
WBC counts and ESR and/or CRP values, a rather common, evaluation of patients with suspected VPI. By considering
non-specific finding. Once a vascular graft infection is sus- the presence of focal [18F]FDG uptake around the prosthesis
pected, prompt and accurate detection is required for the cor- as a diagnostic criterion, sensitivity reaches about 93% and
rect choice of treatment. Success of surgical intervention is specificity 70–91%, with positive and negative predictive
closely dependent on early diagnosis [96]. values of 82–88% and 88–96%, respectively [101].
Ultrasonography with color flow Doppler is often a first- Figures 20.14 and 20.15 show examples of the use of these
line imaging procedure, but in case of aortic graft, the predic- criteria in the diagnosis of VPI infections.
tive value is limited both by air content in the intestinal A semiquantitative approach, by means of SUVmax and/
lumen and, sometimes, by abundant subcutaneous fat. or the tissue-to-background ratio, can be also used. More
CT angiography is the technique of choice for both con- accurate quantification methods are also possible such as,
firmation of the infection and the detection of complications, e.g., textural analysis to evaluate [18F]FDG uptake hetero-
with 94% sensitivity and 85% specificity. While the presence geneity in aortic VPI. The short-run high gray-level
of fluid and air surrounding the aortic graft is a normal find- emphasis demonstrated higher values for the infected
ing in the early postoperative period, finding gas in the peri- compared to the uninfected prosthetic grafts. The short-
prosthetic tissue on the CT scan should be considered run high gray-level emphasis was most efficient in identi-
abnormal beyond 6–8 weeks after surgery. Despite several fying aortic graft infection within the suspected group,
advantages (high specificity, guidance for needle aspiration whereas for the same task, the performances of SUVmax,
and microbiological analysis, speed of execution), the main tissue-to-background ratio, and visual grading scale mea-
limitation of CT imaging is its low sensitivity in detecting surements were all limited [102].
early postsurgical infections and low-grade infections [97]. A particular case of cardiovascular infections arises after
Magnetic resonance imaging (MRI) enables to distin- correction of aorta defects with a special surgical approach,
guish between perigraft fluid and perigraft fibrosis, thanks to the Bentall procedure. This procedure consists in positioning
a a1 a2 c c1 c2
30 min 30 min
2h 2h
b d
b1 d1
b2 d2
Fig. 20.12 99mTc-HMPAO-WBC scintigraphy in a patient with sus- cular graft. After treatment with antimicrobial therapy, the repeat 99mTc-
pected late infection of an aortobisiliac vascular prosthesis presenting HMPAO- WBC scintigraphy showed persistence of uptake at the
with cellulitis and erysipelas of the left lower leg and fever. Swab cul- vascular prosthesis, although less intense than at baseline. Thereafter
ture isolated Enterobacter sakazakii and Escherichia coli. Anterior (a1 the patient underwent debridement of the prosthesis and in situ replace-
and c1) and posterior (a2 and c2) planar images at 30 min and 2 h after ment with autologous femoral veins; microbiology showed the pres-
infusion of 99mTc-HMPAO-WBCs at baseline (a) and during follow-up ence of Candida albicans (reproduced from: Erba PA, Leo G, Sollini M,
(c). Coronal (b1 and d1) and transaxial (b2 and d2) slices: CT (left) and Tascini C, Boni R, Berchiolli RN, et al. Radiolabelled leucocyte scintig-
corresponding fused SPECT/CT sections (right) at baseline (b) and raphy versus conventional radiological imaging for the management of
during follow-up (d). The baseline planar and SPECT/CT images show late, low-grade vascular prosthesis infections. Eur J Nucl Med Mol
increased 99mTc-HMPAO-WBC accumulation corresponding to the vas- Imaging. 2014;41:357–368)
a composite aortic graft (combining a vascular tube graft with condition. There is a predominance of Staphylococcus aureus
an attached mechanical or biologic valve) to replace the prox- infections (35%), with a recent 20% increase in methicillin-
imal ascending aorta and the aortic valve. The main indica- resistant Staphylococcus aureus strains [104].
tions for performing Bentall surgery are the presence of aortic The goal of diagnostic imaging in patients with infection
regurgitation, Marfan’s syndrome, aortic dissection, and aor- after the Bentall procedure is to demonstrate and distinguish
tic aneurysm. Infections after Bentall surgery are reported in the presence of aortic valve-root involvement or of infection
about 3% of the cases [103]; despite such relatively low inci- localized to the vascular aortic graft. These two conditions
dence, the risk of death is high in patients affected by this might also coexist and involve the surrounding structures
Fig. 20.13 99mTc-HMPAO-WBC scintigraphy in a patient with sus- femoral portions. Microbiology showed that the infection was caused
pected late infection of a right iliofemoral and femoropopliteal vascular by polymicrobial agents. Lower left panel: 3D surface rendering
prosthesis. Planar anterior and posterior images at 30 min and at 4 h SPECT/CT images better demonstrating 99mTc-HMPAO-WBC accu-
shown in upper left panel, demonstrating increased 99mTc-HMPAO- mulation all along the vascular graft (reproduced from: Erba PA, Leo G,
WBC accumulation in the right inguinal region. Right panels show Sollini M, Tascini C, Boni R, Berchiolli RN, et al. Radiolabelled leuco-
axial and coronal sections at different levels (from left to right: SPECT, cyte scintigraphy versus conventional radiological imaging for the
CT, fused SPECT/CT images) localizing the focal accumulation of management of late, low-grade vascular prosthesis infections. Eur J
radiolabelled leucocytes to the vascular graft in both the iliac and the Nucl Med Mol Imaging. 2014;41:357–368)
Fig. 20.14 [18F]FDG PET/CT images show increased uptake at the site of an aortic aneurysm (MIP in left panel; right panel from left to right,
CT, emission PET, and fused PET/CT images in the coronal, transaxial, and sagittal views)
Fig. 20.15 [18F]FDG PET/CT images show increased tracer uptake at the thoracic aortic prosthesis after Bentall procedure (MIP in left; right
panels, coronal emission PET, CT, and fused PET/CT images)
such as the mediastinal soft tissue and the sternum.

Identification of patients with infection limited to the vascu- • However, the limited spatial resolution hampers
lar portion of the thoracic aortic grafts (VPI, about 2%) is its diagnostic sensitivity, even when SPECT is
very important since the ideal treatment, replacement of the combined with MDCT in the hybrid SPECT/CT
graft, carries a high mortality, especially in cases of long- approach.
lasting infections or severe comorbidities [105]. No specific • PET/CT imaging displays the high avidity of
guidelines for the management of aortic valve-root-vascular inflammatory infiltrates for [18F]FDG.
prosthesis infections are available, the standard • When using [18F]FDG PET/CT for imaging patients
recommendations for infectious endocarditis and prosthetic with suspected infective endocarditis, special atten-
graft infections being usually followed, including echocar- tion must be paid to time interval from surgery,
diography and ce-CT and/or MRI in short interval. TEE patient preparation, image reconstruction, and eval-
plays a key role in the assessment of IE, but is not always uation criteria in order to maximize its diagnostic
conclusive due to the numerous artifacts related to the pres- potential for infective endocarditis.
ence of the prosthesis. In most of the published reports, the
diagnosis of infection in composite aortic grafts requires a
combination of echocardiography, CT, and PET/CT [106]
(see Fig. 20.15). Also in the presence of infection after 20.4 Clinical Indications, Technique,
Bentall procedure in relation to possible comorbidities, the and Principles of Noninfectious,
risk of distant sites of infections should not be underesti- Non-atherosclerotic Cardiovascular
mated. Prompt extracardiac workout will allow identification Diseases
of both embolic events or concomitant source of infection/
inflammation. 20.4.1 Clinical Background on Myocarditis
Myocarditis, or inflammation of the heart muscle, may pres-

Key Learning Points ent as an acute, subacute, or chronic illness. A large propor-
• 99m
Tc-HMPAO-WBC SPECT is probably the radionu- tion of individuals with this condition may be asymptomatic.
clide imaging technique with the longest experience in The causes of myocarditis include a variety of infectious or
patients with suspected infective endocarditis. systemic diseases, toxins, and drugs. Viruses, especially
• This technique is highly specific both in endocardi- enteroviruses, are the most important causes of myocarditis
tis of native structures and in prosthetic valve endo- in developed countries. The enterovirus genome has been
carditis (PVE). identified in the myocardium of patients with myocarditis or
with dilated cardiomyopathy. Until the introduction of
e ndomyocardial biopsy in 1962, the diagnosis of myocarditis in a patient with chronic active EBV infection presenting
during life was often presumptive and not always correct, with fever, dyspnea on exertion, general malaise, and hepato-
mainly in young patients with heart failure preceded by a splenomegaly; the patient subsequently developed heart fail-
febrile illness. Most cases were diagnosed as late as at ure and myocarditis was confirmed by endomyocardial
autopsy. Prospective and retrospective studies have identi- biopsy. This condition must be distinguished from conges-
fied myocardial inflammation in 1–9% of routine postmor- tive heart failure, right ventricular strain, and hypertrophy
tem examinations. Myocarditis is a major cause of sudden, due to elevated pulmonary artery pressure, which can also
unexpected death (up to 20% of cases) in adults younger than lead to increased [18F]FDG uptake—but in the right ventricu-
40 years. According to the 1995 World Health Organization lar myocardium [111].
(WHO)/International Society and Federation of Cardiology Tuberculous and viral myocarditis should be considered
Task Force on the Definition and Classification of for the differential diagnosis in patients with heterogeneous
Cardiomyopathy, inflammatory cardiomyopathy is now [18F]FDG uptake at PET/CT imaging performed for diagnos-
included as a subtype of specific cardiomyopathies, defined ing infiltrative cardiomyopathy. Unlike tuberculosis and sar-
as myocarditis in association with cardiac dysfunction. coidosis, viral myocarditis does not typically manifest with
Infectious, autoimmune, and idiopathic forms of inflamma- perfusion defects. Given the similar appearance of myocar-
tory heart disease have been recognized. dial [18F]FDG uptake in sarcoidosis and tuberculosis by PET
In 1986, the Dallas criteria for the histologic diagnosis of imaging, a detailed medical history and histologic correla-
myocarditis were defined, based on the identification of infil- tion are crucial for differentiating tuberculous myocarditis
trating lymphocytes and of myocytolysis [107]. Patients with and sarcoidosis [112].
lymphocytic infiltration, but without myocytolysis, were
classified as having borderline or ongoing myocarditis.
However, less than 10% of the patients with suspected myo-
carditis had positive biopsies when assessed by the Dallas Key Learning Points
criteria, thus raising the issue of low sensitivity and high • The condition of myocarditis (a major cause of sud-
inter-observer variability. A second clinico-pathological den, unexpected death in up to 20% in adults aged
classification system was proposed in 1991 but has not been <40 year) includes a variety of acute, subacute, or
widely accepted [108]. chronic illnesses caused by a variety of infectious
Although there have not been major advances in the iden- or systemic diseases, toxins, and drugs.
tification of the etiology of myocarditis in recent years, new • MRI is very useful to distinguish myocarditis from
molecular techniques such as polymerase chain reaction and acute myocardial infarction.
genomic hybridization have allowed confirmation of the eti- • PET/CT with [18F]FDG is a useful adjunct in the
ology in some cases, which would have otherwise remained diagnosis of different forms of myocarditis.
undiagnosed. The serum level of creatine kinase-MB (CK-
MB) has high specificity but limited sensitivity for the diag-
nosis of myocarditis; on the other hand, serum troponin-I is
more often elevated than CK-MB in patients with 20.4.2 Clinical Background on Acute
myocarditis. Pericarditis
Echocardiography performed in the setting of acute myo-
carditis may demonstrate either normal heart function or Pericarditis is an inflammatory process that can be caused
global or regional left ventricular hypokinesis, with an over- by localized or systemic diseases, including infection,
all low sensitivity [109]. connective tissue disorders, and uremia, or can occur after
MRI allows discrimination of myocarditis from myocardial radiation therapy. Although dyspnea and pleuritic chest
infarction by depicting scattered areas of hyper-enhancement pain that changes with patient’s position are the typical
with a nonvascular distribution (in mid-myocardial or subepi- signs of the disease, clinical manifestations can be more
cardial locations) in patients with myocarditis. These areas of variable, depending on the extent of pericardial disease,
hyper-enhancement correspond to inflammation and cell the rate at which it develops, and its effect on cardiac func-
necrosis and are most commonly seen in the inferior and infer- tion [113]. Additional signs include pericardial friction rub
olateral myocardial segments [110]. at physical examination and non-specific ST-segment and
T-wave changes on ECG. Discriminating pericarditis from
20.4.1.1 Radionuclide Imaging of Myocarditis myocardial disease or pulmonary infarction can be diffi-
Diffusely increased metabolic activity in the left ventricular cult on the basis of physical examination alone; imaging
myocardium with mild heterogeneity suggesting the occur- may instead be helpful for distinguishing these different
rence of myocarditis has been observed at [18F]FDG PET/CT conditions.
Echocardiography, CT, and MRI are frequently employed exploited the ability of [18F]FDG PET/CT to distinguish
to identify and characterize the pericardium and pericardial acute tuberculous from idiopathic pericarditis. Patients with
space, the typical findings of pericarditis including pericar- acute tuberculous pericarditis showed a diffuse or multifocal
dial thickening, pericardial enhancement, and pericardial pattern of [18F]FDG uptake in the pericardium associated
effusions [114]. However, echocardiographic assessment of with mediastinal and supraclavicular lymph nodes with
pericardial thickening can be difficult, and the entire pericar- increased uptake [116]. Similarly, the lack of significant [18F]
dium often cannot be fully evaluated because of poor acous- FDG accumulation in the pericardial space in a patient pre-
tic window. CT and MRI allow clearer characterization of senting with idiopathic pericarditis associated with striking
the pericardium and pericardial space and can therefore help hypermetabolism of the thoracic and abdominal aortic wall
to distinguish between pericardial fluid effusion and (highly suggestive of large vessel vasculitis) is consistent
thickening [115]. with pericarditis as the initial manifestation of giant cell
arteritis [117].
20.4.2.1 R adionuclide Imaging of Acute
Pericarditis
[18F]FDG PET/CT is generally employed as a complement to
other imaging modalities, since the different pattern of [18F] Key Learning Points
FDG accumulation in the pericardium or in the pericardial • Discriminating pericarditis from myocardial dis-
fluid might help in distinguishing infectious/inflammatory ease or pulmonary infarction can be difficult on the
conditions from neoplastic pericardial involvement. In fact, basis of physical examination alone; imaging may
noninfectious and inflammatory processes involving the instead be helpful for distinguishing these different
pericardium or pericardial space generally present mild to conditions.
moderate increase or no increase in [18F]FDG uptake (see • Although echocardiography, CT, and MRI are fre-
Fig. 20.16), whereas neoplastic conditions generally present quently employed to identify and characterize the
intense metabolic activity often associated with a focal soft- pericardium and pericardial space, equivocal cases
tissue mass. Incidental reports of cardiac infection or inflam- still remain even after extensive imaging with pre-
mation with increased [18F]FDG activity have been described dominantly morphologic techniques.
in patients evaluated with [18F]FDG PET/CT for fever of • PET/CT with [18F]FDG is a promising imaging
unknown origin. In patients with meningococcal sepsis and alternative in the diagnostic workup of acute
AIDS, intense pericardial uptake at [18F]FDG PET allowed pericarditis.
identification of infectious pericarditis. A recent study
Fig. 20.16 [18F]FDG PET/CT in a 73-year-old woman with cardiac the pericardial site (left panel coronal and transaxial images from left to
pacemaker implanted 3 years earlier, presenting with fever since several right, emission PET, CT, and fused PET/CT, and right panel transaxial
weeks with mildly increased CRP and ESR. Antimicrobial treatment CT and fused PET/CT)
was initiated. The PET/CT images show increased [18F]FDG uptake at
20.4.3 Clinical Background on Cardiac PET/CT imaging is currently most commonly used for imag-
Sarcoidosis ing myocardial inflammation in clinical practice. Nevertheless,
[18F]FDG PET/CT has not been officially included in univo-
About 64% of patients with cardiac sarcoidosis (CS) present cally accepted diagnostic guidelines. This is due to its rela-
with isolated cardiac involvement without sarcoidosis of tively high sensitivity but low specificity and high variability,
other organs. The annual detection rate of CS has increased which limits the use of [18F]FDG PET/CT for strictly diagnos-
>20-fold over the last 25 years, reaching 0.31 in 100,000 ing CS. The main reason for low specificity and high variabil-
adults between 2008 and 2012. The clinical manifestations ity of this imaging technique is physiological [18F]FDG uptake
of CS relate primarily to the location and inflammatory in the normal myocardium. Despite these limitations [18F]
effects of the granulomas. The presence of granulomas in the FDG PET/CT imaging is most commonly used for imaging
ventricular myocardium may lead to abnormal automaticity myocardial inflammation in clinical practice.
and re-entrant tachyarrhythmias manifesting as palpitations In sarcoidosis, the uptake of [18F]FDG is related to inflam-
or syncope. Involvement of the conduction system may lead matory cell infiltrates. Since [18F]FDG accumulates in active
to bradyarrhythmias and syncope. Congestive heart failure lesions where inflammatory cells utilize glucose as an energy
may result from widespread sarcoidosis of the myocardium, source, active sarcoid lesions in various organs are visualized
with progressive decline in left ventricular ejection fraction by increased [18F]FDG uptake on PET/CT imaging. Currently
(EF) and death. The revised Japanese Ministry of Health and used strategies for myocardial suppression include prolonged
Welfare (JMHW) guidelines for the diagnosis of CS are cur- fasting, dietary modifications, and a heparin load before imag-
rently used in many studies as a reference. A more recent ing, as reported in a recent meta-analysis by Tang et al. [122].
publication by the Heart Rhythm Society (HRS), in collabo- This preparation (especially heparin administration) will also
ration with other organizations including the World improve accuracy of the procedure. [18F]FDG imaging starts
Association for Sarcoidosis and Other Granulomatous after an uptake period of 60–90 minutes. Following CT for
Disorders (WASOG), published an expert consensus recom- attenuation correction purpose, non-gated cardiac [18F]FDG
mendation on criteria for the diagnosis of CS [118]. imaging is acquired first. Thereafter, whole-body images are
obtained along with a CT for attenuation correction. [18F]FDG
20.4.3.1 R adionuclide Imaging of Cardiac PET is often combined with a cardiac perfusion scan.
Sarcoidosis Comparison of the perfusion images with the [18F]FDG PET
Due to the multifocal, patchy pattern of myocardial sarcoid images will allow for differentiation of scar tissue from nor-
involvement, the sensitivity of endomyocardial biopsy is as mal myocardium or change of active sarcoid granuloma to
low as 20% [119]. Whereas, even patchy active inflammation normal myocardium or progression to scar tissue in response
can be detected as patchy areas with increased [18F]FDG to therapy. It is also important to exclude epicardial coronary
uptake on the PET/CT scan, thus providing important infor- artery disease in these patients (with either invasive or
mation not only for diagnostic and/or prognostic purposes CT-based coronary angiography), since the patterns of perfu-
but also to guide endomyocardial biopsy [120]. The prognos- sion/metabolism mismatch in patients with left ventricular
tic information provided by [18F]FDG PET/CT is enhanced systolic dysfunction may reflect either hibernating myocar-
when combined with perfusion imaging; in particular, dium or myocardial inflammation according to the clinical
patients with concomitant abnormalities on the 82Rb PET context. CS can affect any part of the heart, the most frequently
scan and on the [18F]FDG scan had a hazard ratio of 3.9 for involved area being the ventricular septum. Several visual
ventricular tachycardia and death [121]. score have been used for evaluating the intensity of [18F]FDG
Imaging modalities recommended in the documents men- uptake, with or without the combination with myocardial per-
tioned above include [18F]FDG PET or PET/CT, 67Ga-citrate fusion with either [13N]ammonia or 82Rb PET imaging.
scintigraphy, and cardiac MRI. [18F]FDG PET/CT has 89% sensitivity and 78% specific-
Although 67Ga-citrate scintigraphy is included in the major ity for CS detection [123]. In addition, patients with CS pres-
diagnostic criteria for CS, its sensitivity for detecting CS is ent a significantly higher SUVmax than non-CS patients. With
reported to be as low as 36.4%, which is significantly lower a cutoff value of 4.0 for myocardial SUVmax, sensitivity was
than that of [18F]FDG PET. Therefore 67Ga-citrate scintigra- 97.3% and specificity 83.6% for diagnosing CS. The speci-
phy has largely been replaced with [18F]FDG PET/CT for ficity of quantitative analysis was greater than that of visual
diagnosing CS. In fact, [18F]FDG PET/CT has several practi- analysis. Moreover, SUVmax on [18F]FDG PET/CT was the
cal and technical advantages over 67Ga-citrate scintigraphy, only independent predictor among clinical and imaging vari-
including favorable tracer kinetics, lower radiation exposure, ables for diagnosing CS. Therefore, [18F]FDG PET/CT using
and better quality images. Furthermore, [18F]FDG PET/CT is quantitative analysis is expected to facilitate correct diagno-
more accurate than 67Ga-citrate and allows better evaluation of sis of CS. Several studies compared late enhancement in
extrapulmonary involvement, including CS. Hence, [18F]FDG CMR and [18F]FDG uptake in PET/CT, showing only mild to
Fig. 20.17 [18F]FDG PET/CT in a young man with sudden arrhythmic FDG uptake in multiple mediastinal lymph nodes, as well as cardiac
episodes. Sarcoidosis was diagnosed by biopsy of a mediastinal lymph involvement by sarcoidosis (left panel MIP, right panel transaxial emis-
node guided by the PET/CT images. The images show increased [18F] sion, CT, and fused PET/CT at different levels)
moderate correlation. This finding is due to the fact that, 20.4.4 Clinical Background on Cardiac
while delayed enhancement represents cardiac damage and Amyloidosis
scarring, [18F]FDG uptake represents active inflammation.
When CMR and [18F]FDG PET were compared with the Cardiac amyloidosis (CA) may be genetic/familial (ATTR) or
Japanese Ministry of Health and Welfare guidelines, CMR non-genetic/non-familial (AL/prealbumin, senile); the genetic
had a higher specificity but a lower sensitivity. Figure 20.17 variant usually involves deposition of the transthyretin/prealbu-
shows an example of CS evaluated with [18F]FDG PET. min protein (and is therefore defined as the “ATTR” form),
[18F]FDG PET can also be used to assess response to ther- while the non-familial variant usually involves deposition of
apy [124]. Following corticosteroid therapy, [18F]FDG light-chain amyloid (and is therefore defined as the “AL” form).
uptake in cardiac lesions declines. Similarly, in a retrospec- Imaging plays a major role in assessing patients with cardiac
tive study of 23 patients with cardiac sarcoidosis treated with amyloidosis, according to the following clinical scenarios:
immunosuppressive therapies guided by serial PET scans,
reduction in the intensity and extent of inflammation, as • Differential diagnosis between hypertrophic and restric-
quantified by [18F]FDG uptake, was associated with increas- tive cardiomyopathy.
ing LVEF values. • Evaluation of cardiac involvement in patients with either
form of amyloidosis.
• Staging of the disease and monitoring of response to
Key Learning Points therapy.
• Non-infective active inflammation (e.g., cardiac sar-
coidosis) can be detected as patchy uptake of [18F] Echocardiography is a first-line technique in cardiac amy-
FDG on PET/CT, linked to the inflammatory cells loidosis and is recommended for the diagnosis, functional
infiltrate accumulation within myocardial walls. and anatomical characterization, differential diagnosis, and
• Moreover, [18F]FDG PET/CT imaging has been prognostic stratification of these patients. Nevertheless, it
shown to be able to guide endomyocardial biopsy provides relatively late non-specific information on thicken-
and to predict response to therapy. ing of amyloidotic myocardium.
• Physiologic myocardial [18F]FDG uptake requires Gadolinium-based contrast MR (CMR) offers particular
dedicated patient preparation to avoid the pitfall of advantages in patients with suspected cardiac amyloidosis,
inadequately suppressed basal uptake. because of its high spatial resolution and its ability to charac-
terize myocardial tissue. CMR is often used after CA is
s uspected by echocardiography to confirm or refute the diag-

several phosphate derivatives labelled with 99mTc, resulting in
nosis and in experienced hands represents a powerful tool with
the current clinical use of the following tracers: 99mTc-2,3-
important diagnostic, with reasonably high sensitivity [125]
dicarboxypropane-1,1-diphosphonate (99mTc-DPD) mainly in
and prognostic implications. Late gadolinium enhancement Europe and Asia and 99mTc-pyrophosphate (99mTc-PYP) in the
(LGE) is typically observed in a circumferential subendocar-
United States. Their main advantage is avid uptake by ATTR
dial pattern or with bilateral subendocardial LGE of the sep-
amyloidosis and minimal uptake with the AL variant, thus
tum with a dark mid-wall (zebra pattern); combination of this
providing one of the best noninvasive ways to differentiate
finding with altered gadolinium blood pool kinetics is highly
these two subtypes of cardiac amyloidosis [126]. In AL amy-
suggestive for cardiac amyloidosis. Patterns of focal LGE may
loidosis, cardiac uptake is found in less than half of patients
also be found but are less specific for cardiac amyloidosis.
and is generally less intense (likely due to the lower concen-
Myocardial non-contrast T1 values are longer in cardiac amy-
tration of calcium-containing products in AL amyloid).
loidosis than in controls, a finding with higher sensitivity for
Furthermore, AL patients generally exhibit no muscular
detecting early subclinical cardiac involvement than LGE.99m
Tc-DPD or 99mTc-HDP uptake, while visceral uptake (liver,
It should be noted that neither echocardiography nor spleen) is more common (see Fig. 20.18). 99mTc-DPD has also
CMR enable to distinguish between the two variants of car-
been shown to provide independent prognostic value.
diac amyloidosis. Scintigraphy with either 99mTc-DPD or 99mTc-PYP scin-
tigraphy shows early signs of amyloid involvement of myo-
20.4.4.1 Radionuclide Imaging of Cardiac cardial tissue and might be helpful in distinguishing between
Amyloidosis ATTR and AL amyloidosis [126]. The most reasonable
Radiolabelled bone-seeking phosphates were incidentally explanation for accumulation of 99mTc-labelled bisphospho-
noted to localize in amyloid deposits using 99mTc- nates in the amyloidotic myocardium is related to their affin-
diphosphonate. This observation led to the development of ity for the high calcium content of transthyretin amyloid.
99mTc-PYP 99mTc-DPD
a b
Fig. 20.18 Example of cardiac SPECT in a patient with ATTR amyloidosis. 99mTc-PYP (a) and 99mTc-DPD (b) (courtesy of Selene Capitanio, MD,
Nuclear Medicine Unit, University Hospital of Trieste, Trieste, Italy)
The intensity of 99mTc-DPD uptake in ATTR patients is 20.5 Heart Failure

negatively correlated with the LV ejection fraction assessed
by echocardiography, being a predictor of major adverse car- Cardiac pump dysfunction is the final endpoint of a wide
diac outcomes. Moreover, the concomitance of 99mTc-PYP range of cardiac disorders including ischemic heart disease,
uptake and of myocardial perfusion defects suggests the replacement fibrosis, infiltration, inflammation, and myocar-
opportunity to combine the 99mTc-PYP scan with SPECT or dial iron deposition. Establishing the specific etiology of
PET perfusion imaging, for enhanced diagnostic accuracy. heart failure (HF) is crucial because it is associated with
After intravenous administration of 740 MBq of 99mTc- prognosis and it guides specific therapies. Multimodality
DPD/HDP or 99mTc-PYP, a whole-body scan is performed 3 h imaging in the setting of HF aims to join the different physi-
or 1 h later (anterior and posterior projections). If there is active cal bases of image formation to create an image set with
uptake in the heart, chest SPECT is performed (Fig. 20.18). complementary (or even synergistic) information to demon-
The analysis is performed by semiquantitative visual scoring of strate the occurrence of cardiac dysfunction, to display the
cardiac uptake as compared to bone uptake (scores from 0 to 3) underlying structural heart disease, and finally to determine
and by computing the ratio, after correction for background the HF mechanisms (HF with reduced or preserved systolic
activity, of the mean counts in the heart region divided by the function) [131].
mean counts in the contralateral chest (H/CL ratio). Hence, pla- In patients with ischemic HF, identifying residual viable
nar 99mTc-DPD/HDP scintigraphy helps to phenotype cardiac myocardium has a major impact on prognosis and therapy.
amyloidosis, particularly for distinguishing of ATTR from AL Current guidelines recommend myocardial revascularization
amyloidosis. This property is shared by PET/CT with in patients with ejection fraction ≤35% only in the presence
18
F-fluoride, which can distinguish ATTR from AL cardiac of viable myocardium. Detection of residual myocardial glu-
amyloidosis [127] with similar diagnostic accuracy but greater cose uptake with [18F]FDG PET/CT is a widely accepted
spatial resolution. In this regard, it can be speculated that PET/ approach to identify these patients with viable (increased
MR with 18F-fluoride may combine optimal diagnostic perfor- [18F]FDG uptake) myocardium, therefore with potential for
mance of the two imaging modalities. recovery of contractile function after revascularization. On
Finally, novel PET tracers originally developed for brain the other hand, a matching flow-metabolic defect indicates
beta-amyloid imaging, such as the 11C-labelled Pittsburgh com- scar formation. The substrate and hormonal levels in the
pound B ([11C]PiB) [128] or 18F-florbetapir [129, 130], appear blood must switch metabolism in the myocardium to glucose
promising for imaging cardiac amyloidosis (particularly AL over fatty acids. To achieve this, a glucose load (25–100 g) is
amyloidosis) and are currently under clinical investigation. administered orally to induce an endogenous insulin response
and the consequent overexpression of GLUT-4 on cardiomy-
ocytes. Intravenous loading may also be used. Supplemental
Key Learning Points insulin administration can be also considered, as needed. In
• Scintigraphy with either 99mTc-DPD or 99mTc-PYP case of diabetic patients, a major challenge occurs, either
shows early signs of amyloid involvement of myo- because they have limited endogenous insulin production or
cardial tissue and might be helpful for distinguish- because their cells respond less to insulin stimulation.
ing ATTR from AL amyloidosis. Potential alternative protocols in this clinical scenario are
• Excellent agreement has been shown between echo- insulin administration along with close monitoring of blood
cardiography findings and 99mTc-DPD uptake in ATTR glucose or to delay image acquisition from 2 to 3 h after [18F]
patients, the intensity of 99mTc-DPD uptake being neg- FDG injection (see also Tables 20.7 and 20.8). In a pooled
atively correlated with the LV ejection fraction. meta-analysis of 24 studies in 756 patients, the weighed mean
• 99mTc-DPD uptake in the myocardium predicts sensitivity and specificity of [18F]FDG PET/CT for myocar-
major adverse cardiac outcomes. dial viability were 92% and 63%, respectively [132].
• Although PET/CT has been intensively investi- A wide series of radiotracers have also been proposed to
gated, its clinical role is still limited. investigate myocardial innervation of the failing heart [133].
• While [18F]FDG PET/CT imaging does not accu- Currently, 123I-metaiodobenzylguanidine (123I-MIBG), a
rately display cardiac involvement in patients with radioiodinated norepinephrine analogue, is the most clini-
systemic amyloidosis, several non-[18F]FDG tracers cally used for myocardial SPECT imaging of sympathetic
show promising results. function. 123I-MIBG is administered i.v. in the amount of
• Since either modality is potentially able to provide 185 MBq over 1–2 min, after thyroid blockade by oral
complementary information, PET/MRI might administration of 400 mg of potassium perchlorate. Ten-min
become the hybrid imaging method of choice for planar images of the thorax in standard anterior view (with
patients with cardiac amyloidosis. matrix 128 × 128) are recorded at 30 min (early image) and
3–4 h (late image) after tracer administration. For early and
late planar images, the heart-to-mediastinum (H/M) ratio is

computed by dividing the mean counts per pixel within the • Cardiac uptake of 123I-MIBG is reduced in HF
myocardium by the mean counts per pixel within the medias- patients, indicating the prognostic power of myo-
tinum. The washout rate (WR) is calculated using the follow- cardial innervation imaging.
ing equation: • For early and late planar images, the heart-to-medi-
astinum (H/M) ratio is computed by dividing the
WR = ( early H / M - late H / M ) / early H / M ´ 100. mean counts per pixel within the myocardium by
the mean counts per pixel within the mediastinum.
Various quantitation methods for measuring myocardial • The washout rate (WR) is calculated using the fol-
uptake have been proposed, although the simplest index is lowing formula: (early H/M – Late H/M)/early
the heart-to-mediastinum ratio (HMR), which is a crude but H/M × 100.
practical parameter. However, the practical simplicity of the
HMR does not necessarily result in its consistency, as vari-
ous factors are known to cause variations in this score; in
particular, choice of the collimator introduces considerable 20.6 Large Vessels Vasculitis
variations [134, 135]. Since the HMR is determined by an
averaged count ratio of heart and mediastinum, the location, Large vessel vasculitis consists of a group of conditions caus-
size, and shape of the region of interest (ROI) can all induce ing inflammation of the aorta and its branches. It is an uncom-
variability. Although in clinical practice such variations mon but well-known cause of non-specific symptoms that
between different centers exist, reproducibility of the HMR constitute a relevant diagnostic challenge. Large vessel vas-
is generally considered to be good for inter- and intra-patient culitis includes giant cell arteritis (GCA), Takayasu’s arteritis,
comparisons in the same center [136]. However, some and isolated aortitis; other associated conditions are Behçet’s
variations still exist, depending on ROI setting methods; pre- disease, Cogan’s syndrome, systemic lupus erythematosus,
defined or semiautomatic methods for ROI definition are spondyloarthritis, and sarcoidosis. Other types of vasculitis,
preferable [137, 138]. There is no general consensus even such as the ANCA-positive vasculitis and polyarteritis
regarding the prognostic threshold of 1.6 that has been pro- nodosa, may occasionally involve the large vessels walls.
posed for the HMR. [18F]FDG PET, CT coronary angiography (CTCA), and
Many studies have demonstrated that cardiac uptake of MRI demonstrate that extracranial involvement in GCA is
123
I-MIBG is reduced in HF patients, thus suggesting a quite frequent, occurring in 30–74% of patients. In patients
prognostic power for myocardial innervation imaging with typical clinical presentation, the diagnosis of GCA is
[139]. However, widespread application of 123I-MIBG not difficult, and temporal artery biopsy (TAB) is considered
imaging is still limited, partly due to suboptimal target-to- to be the “gold standard” for diagnostic confirmation,
background ratios and to the use of planar imaging for although hampered by low sensitivity (with 15–40% false-
quantitation of myocardial activity relative to background negative rate). Instead, diagnosis may be challenging when
activity in the mediastinum or for the estimation of myocar- symptoms are non-specific, given the wide range of clinical
dial washout. manifestations of GCA.
Due to the inflammatory nature of the disease, [18F]FDG
PET/CT can detect vascular involvement in GCA patients.
Key Learning Points Other imaging techniques (i.e., Doppler ultrasonography,
• In ischemic HF, identification of residual viable CT, and MRI) have been proposed for the assessment of
myocardium profoundly impacts prognosis and vascular inflammation in GCA patients. However, while
guides therapy. they are able to demonstrate anatomic changes in the
• Assessment of residual myocardial viability with affected vessels (wall thickening, dilatation and aneurysms,
[18F]FDG PET/CT identifies ischemic myocardium and enhancement of perivascular connective tissue) if the
that is dysfunctional but viable and, therefore, has inflammatory process is advanced, they are not sensitive
potential for recovery of contractile function after enough to diagnose early inflammatory changes, which are
revascularization. potentially reversible following therapy. Furthermore,
• Different radiotracers have been proposed to patient follow-up and assessment of response to medical
explore myocardial innervation of the failing heart. treatment are not easy to perform on the basis of morpho-
• Currently, 123I-MIBG is the most widely used tracer logical information alone [140]. The main limitation of [18F]
for myocardial planar and/or SPECT imaging of FDG PET/CT as a reliable diagnostic tool is the lack of a
sympathetic function. standardized definition of vascular inflammation based on
the intensity of [18F]FDG uptake.
Qualitative and semiquantitative methods for assessing

the presence and for grading the severity of GCA-related • Several qualitative and semiquantitative methods
vascular inflammation on [18F]FDG PET/CT imaging have for assessing the presence and for grading the sever-
been proposed, qualitative analysis of [18F]FDG uptake being ity of vasculitis-related inflammation on [18F]FDG
currently the most widely adopted approach. Qualitative PET/CT imaging have been proposed.
analysis has been used both to perform dichotomous assess- • For quantitative analysis, the SUV-based approach
ment (confirm or exclude the presence of vascular inflamma- has not been sufficiently validated for evaluating
tion) [141] and to grade the severity of vascular involvement inflammation and infection.
according to ordinal scales, [18F]FDG uptake in the vessel • SUV-based methods should be used with caution in
wall being visually analyzed or compared with that of a ref- the clinical practice.
erence structure. Among the visual grading systems, the
vessel-to-liver ratio is the most frequently adopted. The
advantage of qualitative methods, in addition to being more
References
immediate and rapid than semiquantitative methods, lies in
their high specificity, with sensitivity ranging from 56% to 1. Rossi A, Uitterdijk A, Dijkshoorn M, Klotz E, Dharampal A,
77%. Moreover, qualitative methods are characterized by van Straten M, et al. Quantification of myocardial blood flow by
high inter-observer agreement (90%) and intra-observer adenosine-stress CT perfusion imaging in pigs during various
reproducibility (93.3%). Nevertheless, the diagnostic dis- degrees of stenosis correlates well with coronary artery blood
flow and fractional flow reserve. Eur Heart J Cardiovasc Imaging.
crimination between vasculitis and other vascular diseases 2013;14(4):331–8.
(e.g., atherosclerosis) is still challenging. 2. Marini C, Seitun S, Zawaideh C, Bauckneht M, Morelli MC,
As to the semiquantitative methods, several authors have Ameri P, et al. Comparison of coronary flow reserve estimated
proposed arterial SUVmax as a diagnostic factor [140], while by dynamic radionuclide SPECT and multi-detector X-ray CT. J
Nucl Cardiol. 2017;24:1712–21.
others have normalized arterial SUVmax to background activ- 3. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis
ity, represented by the mean uptake value of a selected refer- GJ. Compensatory enlargement of human atherosclerotic coro-
ence structure. When comparing the SUVmax-based approach nary arteries. N Engl J Med. 1987;316(22):1371–5.
without normalization with simple visual qualitative analysis, 4. Libby P, Pasterkamp G. Requiem for the 'vulnerable plaque. Eur
Heart J. 2015;36(43):2984–7.
high sensitivity and low specificity has been reported by using 5. Hackett D, Davies G, Maseri A. Pre-existing coronary stenoses in
a cutoff value. Nevertheless, SUVmax cutoff values obtained patients with first myocardial infarction are not necessarily severe.
without normalization appear to be population-specific, being Eur Heart J. 1988;9(12):1317–23.
therefore difficult to apply to the general population. In a 6. Sambuceti G, Marini C, Morbelli S, Paoli G, Derchi M, Pomposelli
E. Witnessing ischemia or proofing coronary atherosclerosis: two
recent meta-analysis, Puppo et al. found that semiquantitative different windows on the same or on different pathways precipitat-
methods normalized to background activity seem to outper- ing cardiovascular events? J Nucl Cardiol. 2009;16(3):447–55.
form qualitative approaches and semiquantitative methods 7. Gould KL, Lipscomb K, Hamilton GW. Physiologic basis for
without normalization [140]. They also noticed that the power assessing critical coronary stenosis. Instantaneous flow response
and regional distribution during coronary hyperemia as measures
of discrimination between GCA and control groups also of coronary flow reserve. Am J Cardiol. 1974;33(1):87–94.
depends on the anatomical structure chosen as the back- 8. Sambuceti G, Marzilli M, Mari A, Marini C, Schluter M, Testa
ground. An emerging semiquantitative approach is the arte- R, et al. Coronary microcirculatory vasoconstriction is hetero-
rial-to-background ratio proposed by Moosig et al. [142], geneously distributed in acutely ischemic myocardium. Am J
Physiol Heart Circ Physiol. 2005;288(5):H2298–305.
which outperformed the other methods where liver and lung 9. Saracco A, Bauckneht M, Verna E, Ghiringhelli S, Repetto R,
are used as reference structures for background. Sambuceti G, Provasoli S, Storace M. A mathematical model for
the vessel recruitment in coronary microcirculation in the absence
of active autoregulation. Microvasc Res. 2016;104:38–45.
10. Sambuceti G, Marzilli M, Fedele S, Marini C, L’Abbate
A. Paradoxical increase in microvascular resistance during
Key Learning Points tachycardia downstream from a severe stenosis in patients with
• Due to the inflammatory nature of the disease, [18F] coronary artery disease: reversal by angioplasty. Circulation.
FDG PET/CT imaging is a feasible noninvasive 2001;103(19):2352–60.
11. Marini C, Acampa W, Bauckneht M, Daniele S, Capitanio S,
method to detect vascular involvement in patients Cantoni V, et al. Added prognostic value of ischaemic threshold
with large vessel vasculitis. in radionuclide myocardial perfusion imaging: a common-sense
• The main limitation of [18F]FDG PET/CT as a reli- integration of exercise tolerance and ischaemia severity. Eur J
able diagnostic tool is the lack of a standardized Nucl Med Mol Imaging. 2015;42(5):750–60.
12. Sambuceti G, Giorgetti A, Corsiglia L, Marini C, Schneider-
definition of vascular inflammation based on the Eicke J, Brunelli C, et al. Perfusion-contraction mismatch during
intensity of [18F]FDG uptake. inotropic stimulation in hibernating myocardium. J Nucl Med.
1998;39(3):396–402.
13. Ward RP, Al-Mallah MH, Grossman GB, Hansen CL, Hendel RC, 28. Di Carli MF, Dorbala S, Meserve J, El Fakhri G, Sitek A,
Kerwin TC, et al. American Society of Nuclear Cardiology review Moore SC. Clinical myocardial perfusion PET/CT. J Nucl Med.
of the ACCF/ASNC appropriateness criteria for single-photon 2007;4:783–93.
emission computed tomography myocardial perfusion imaging 29. Huisman MC, Higuchi T, Reder S, Nekolla SG, Poethko T, Wester
(SPECT MPI). J Nucl Cardiol. 2007;14(6):e26–38. HJ, et al. Initial characterization of an 18F-labeled myocardial per-
14. Hesse B, Lindhardt TB, Acampa W, Anagnostopoulos C, fusion tracer. J Nucl Med. 2008;49(4):630–6.
Ballinger J, Bax JJ, et al. EANM/ESC guidelines for radionu- 30. Yu M, Guaraldi MT, Mistry M, Kagan M, McDonald JL, Drew K,
clide imaging of cardiac function. Eur J Nucl Med Mol Imaging. et al. BMS-747158-02: a novel PET myocardial perfusion imag-
2008;35(4):851–85. ing agent. J Nucl Cardiol. 2007;14(6):789–98.
15. Hendel RC, Berman DS, Di Carli MF, Heidenreich PA, Henkin 31. Nekolla SG, Reder S, Saraste A, Higuchi T, Dzewas G, Preissel
RE, Pellikka PA, et al. ACCF/ASNC/ACR/AHA/ASE/SCCT/ A, et al. Evaluation of the novel myocardial perfusion positron-
SCMR/SNM 2009 appropriate use criteria for cardiac radionu- emission tomography tracer 18F-BMS-747158-02: comparison to
clide imaging: a report of the American College of Cardiology 13
N-ammonia and validation with microspheres in a pig model.
Foundation Appropriate Use Criteria Task Force, the American Circulation. 2009;119(17):2333–42.
Society of Nuclear Cardiology, the American College of 32. Sherif HM, Nekolla SG, Saraste A, Reder S, Yu M, Robinson S,
Radiology, the American Heart Association, the American Society et al. Simplified quantification of myocardial flow reserve with
of Echocardiography, the Society of Cardiovascular Computed flurpiridaz F-18: validation with microspheres in a pig model. J
Tomography, the Society for Cardiovascular Magnetic Resonance, Nucl Med. 2011;52(4):617–24.
and the Society of Nuclear Medicine. J Am Coll Cardiol. 33. Klein R, deKemp RA. Cardiac PET imaging: principles and new
2009;53(23):2201–29. developments. In: Khalil M, editor. Basic science of PET imaging.
16. Verberne HJ, Acampa W, Anagnostopoulos C, Ballinger J, Bengel Cham: Springer; 2017. p. 451–83.
F, De Bondt P, et al. EANM procedural guidelines for radionuclide 34. Canepa M, Bezante G, Vianello P, Ameri P, Milaneschi Y, Aste M,
myocardial perfusion imaging with SPECT and SPECT/CT: 2015 et al. Diagnostic value of ischemia severity at myocardial perfu-
revision. Eur J Nucl Med Mol Imaging. 2015;42(12):1929–40. sion imaging in elderly persons with suspected coronary disease.
17. Henzlova MJ, Duvall WL, Einstein AJ, Travin MI, Verberne J Cardiovasc Med (Hagerstown). 2016;17(10):719–28.
HJ. ASNC imaging guidelines for SPECT nuclear cardiol- 35. Slomka P, Xu Y, Berman D, Germano G. Quantitative analy-
ogy procedures: stress, protocols, and tracers. J Nucl Cardiol. sis of perfusion studies: strengths and pitfalls. J Nucl Cardiol.
2016;23(3):606–39. 2012;19(2):338–46.
18. Hesse B, Tagil K, Cuocolo A, Anagnostopoulos C, Bardies M, 36. Garcia EV, Faber TL, Cooke CD, Folks RD, Chen J, Santana
Bax J, et al. EANM/ESC procedural guidelines for myocardial C. The increasing role of quantification in clinical nuclear cardiol-
perfusion imaging in nuclear cardiology. Eur J Nucl Med Mol ogy: the Emory approach. J Nucl Cardiol. 2007;14(4):420–32.
Imaging. 2005;32(7):855–97. 37. Germano G, Kavanagh PB, Slomka PJ, Van Kriekinge SD, Pollard
19. Taillefer R. Should early post-stress imaging be performed on a G, Berman DS. Quantitation in gated perfusion SPECT imaging:
routine clinical basis for myocardial perfusion studies? J Nucl the cedars-Sinai approach. J Nucl Cardiol. 2007;14(4):433–54.
Cardiol. 2014;21(6):1177–80. 38. Ficaro EP, Lee BC, Kritzman JN, Corbett JR. Corridor4DM:
20. Mut F, Giubbini R, Vitola J, Lusa L, Sobic-Saranovic D, Peix A, the Michigan method for quantitative nuclear cardiology. J Nucl
et al. Detection of post-exercise stunning by early gated SPECT Cardiol. 2007;14(4):455–65.
myocardial perfusion imaging: results from the IAEA multi- 39. Liu YH. Quantification of nuclear cardiac images: the Yale
center study. J Nucl Cardiol. 2014;21(6):1168–76. approach. J Nucl Cardiol. 2007;14(4):483–91.
21. Kaufmann BA, Pfisterer ME, Viswanathan S, Muller-Brand J, 40. Rossen JD, Quillen JE, Lopez AG, Stenberg RG, Talman CL,
Zellweger MJ. Stunning and left ventricular function--how long Winniford MD. Comparison of coronary vasodilation with
is the ventricle knocked out? Left ventricular function correlated intravenous dipyridamole and adenosine. J Am Coll Cardiol.
with ischemic burden and recovery time after stress. Int J Cardiol. 1991;18(2):485–91.
2006;112(2):223–8. 41. Mekkaoui C, Jadbabaie F, Dione DP, Meoli DF, Purushothaman
22. Emmett L, Iwanochko RM, Freeman MR, Barolet A, Lee DS, K, Belardinelli L, et al. Effects of adenosine and a selective A2A
Husain M. Reversible regional wall motion abnormalities on exer- adenosine receptor agonist on hemodynamic and thallium-201
cise technetium-99m-gated cardiac single photon emission com- and technetium-99m-sestaMIBI biodistribution and kinetics.
puted tomography predict high-grade angiographic stenoses. J Am JACC Cardiovasc Imaging. 2009;2(10):1198–208.
Coll Cardiol. 2002;39(6):991–8. 42. Bauckneht M, Piva R, Ferrarazzo G, Ceriani V, Buschiazzo A,
23. Johnson LL, Verdesca SA, Aude WY, Xavier RC, Nott LT, Morbelli S, et al. Non-invasive measurement of coronary flow
Campanella MW, et al. Postischemic stunning can affect left reserve: uniqueness of radionuclide methods and alternative tech-
ventricular ejection fraction and regional wall motion on niques. Q J Nucl Med Mol Imaging. 2016;60(4):324–37.
post-stress gated sestamibi tomograms. J Am Coll Cardiol. 43. Sampson UK, Dorbala S, Limaye A, Kwong R, Di Carli
1997;30(7):1641–8. MF. Diagnostic accuracy of rubidium-82 myocardial perfusion
24. Heller GV, Mann A, Hendel R. Nuclear cardiology: technical imaging with hybrid positron emission tomography/computed
applications. New York: McGraw-Hill; 2009. tomography in the detection of coronary artery disease. J Am Coll
25. Yoshinaga K, Klein R, Tamaki N. Generator-produced rubidium- Cardiol. 2007;49(10):1052–8.
82 positron emission tomography myocardial perfusion imaging- 44. Dorbala S, Vangala D, Sampson U, Limaye A, Kwong R, Di Carli
from basic aspects to clinical applications. J Cardiol. 2010;55(2): MF. Value of vasodilator left ventricular ejection fraction reserve
163–73. in evaluating the magnitude of myocardium at risk and the extent
26. Dilsizian V, Bacharach SL, Beanlands RS, Bergmann SR, Delbeke of angiographic coronary artery disease: a 82Rb PET/CT study. J
D, Dorbala S, et al. ASNC imaging guidelines/SNMMI procedure Nucl Med. 2007;48(3):349–58.
standard for positron emission tomography (PET) nuclear cardiol- 45. Kajander S, Joutsiniemi E, Saraste M, Pietila M, Ukkonen H, Saraste
ogy procedures. J Nucl Cardiol. 2016;23(5):1187–226. A, et al. Cardiac positron emission tomography/computed tomogra-
27. Maddahi J, Packard RR. Cardiac PET perfusion tracers: current phy imaging accurately detects anatomically and functionally sig-
status and future directions. Semin Nucl Med. 2014;44(5):333–43. nificant coronary artery disease. Circulation. 2010;122(6):603–13.
46. Johnson NP, Gould KL. Physiological basis for angina and 65. Camus C, Leport C, Raffi F, et al. Sustained bacteremia in 26
ST-segment change PET-verified thresholds of quantitative stress patients with a permanent endocardial pacemaker: assessment of
myocardial perfusion and coronary flow reserve. JACC Cardiovasc wire removal. Clin Infect Dis. 1993;17:46–55.
Imaging. 2011;4(9):990–8. 66. Tarakji KG, Chan EJ, Cantillon DJ, et al. Cardiovascular implant-
47. Packard RR, Huang SC, Dahlbom M, Czernin J, Maddahi able electronic device infections: presentation, management, and
J. Absolute quantitation of myocardial blood flow in human sub- patient outcomes. Heart Rhythm. 2010;7:1043–7.
jects with or without myocardial ischemia using dynamic flur- 67. Holman WL, Naftel DC, Eckert CE, et al. Durability of left
piridaz F-18 PET. J Nucl Med. 2014;55(9):1438–44. ventricular assist devices: interagency registry for mechanically
48. Ko SM, Hwang HK, Kim SM, Cho IH. Multi-modality imaging assisted circulatory support (INTERMACS). J Thorac Cardiovasc
for the assessment of myocardial perfusion with emphasis on Surg. 2006–2011;2013(146):437–41.
stress perfusion CT and MR imaging. Int J Cardiovasc Imaging. 68. Xie A, Phan K, Yan TD. Durability of continuous-flow left ven-
2015;31(Suppl 1):1–21. tricular assist devices: a systematic review. Ann Cardiothorac
49. Alexanderson-Rosas E, Guinto-Nishimura GY, Cruz-Mendoza Surg. 2014;3:547–56.
JR, Oropeza-Aguilar M, De La Fuente-Mancera JC, Barrero- 69. Wickline SA, Fischer KC. Can infections be imaged in implanted
Mier AF, et al. Current and future trends in multimodality imag- devices? ASAIO J. 2000;46:80–1.
ing of coronary artery disease. Expert Rev Cardiovasc Ther. 70. Siméon S, Flécher E, Revest M, et al. Left ventricular assist
2015;13(6):715–31. device-related infections: a multicentric study. Clin Microbiol
50. Osler W. The Gulstonian lectures, on malignant endocarditis. Br Infect. 2017;23:748–51.
Med J. 1885;1(1264):577–9. 71. Koval CE, Rakita R. AST Infectious Disease Community of
51. Thuny F, Giorgi R, Habachi R, Ansaldi S, Le Dolley Y, Casalta JP, Practice Ventricular assist device related infections and solid
et al. Excess mortality and morbidity in patients surviving infec- organ transplantation. Am J Transplant. 2013;13:348–54.
tive endocarditis. Am Heart J. 2012;164(1):94–101. 72. Erba PA, Conti U, Lazzeri E, et al. Added value of 99mTc-HMPAO-
52. Gelijns AC, Moskowitz AJ, Acker MA, et al. Management prac- labeled leukocyte SPECT/CT in the characterization and man-
tices and major infections after cardiac surgery. J Am Coll Cardiol. agement of patients with infectious endocarditis. J Nucl Med.
2014;64:372–81. 2012;53:1235–43.
53. Donlan RM. Biofilms and device-associated infections. Emerg 73. Erba PA, Lancellotti P, Vilacosta I, et al. Recommendation on
Infect Dis. 2001;7:277–81. nuclear and multi-modality imaging in IE and CIED infections.
54. von Eiff C, Jansen B, Kohnen W, Becker K. Infections associated Eur J Nucl Med Mol Imaging. 2018;45(10):1795–815.
with medical devices. Drugs. 2005;65:179–214. 74. Rouzet F, Chequer R, Benali K, et al. Respective performance
55. Baddour LM, Wilson WR, Bayer AS, Fowler VG Jr, Tleyjeh IM, of 18F-FDG PET and radiolabeled leucocyte scintigraphy for
Rybak MJ, et al. Infective endocarditis in adults: diagnosis, anti- the diagnosis of prosthetic valve endocarditis. J Nucl Med.
microbial therapy, and management of complications: a scientific 2014;55:1980–5.
statement for healthcare professionals from the American Heart 75. Scholtens AM, Verberne HJ, Budde RP, Lam MG. Additional
Association. Circulation. 2015;132(15):1435–86. heparin preadministration improves cardiac glucose metabolism
56. Jensen AG, Wachmann CH, Poulsen KB, Espersen F, Scheibel J, suppression over low-carbohydrate diet alone in 18F-FDG PET
Skinhøj P, et al. Risk factors for hospital-acquired Staphylococcus imaging. J Nucl Med. 2016;57(4):568–73.
aureus bacteremia. Arch Intern Med. 1999;159(13):1437–44. 76. Dilsizian V. Transition from SPECT to PET myocardial perfusion
57. Jensen AG, Wachmann CH, Espersen F, et al. Treatment and out- imaging: a desirable change in nuclear cardiology to approach
come of Staphylococcus aureus bacteremia: a prospective study of perfection. J Nucl Cardiol. 2016;23(3):337–8.
278 cases. Arch Intern Med. 2002;162:25–32. 77. Jamar F, Buscombe J, Chiti A, Christian PE, Delbeke D, Donohoe
58. Perez de Isla L, Zamorano J, Lennie V, et al. Negative blood cul- KJ, et al. EANM/SNMMI guideline for 18F-FDG use in inflamma-
ture infective endocarditis in the elderly: long-term follow-up. tion and infection. J Nucl Med. 2013;54:647–58.
Gerontology. 2007;53:245–9. 78. Jimenez-Ballve A, Perez-Castejon MJ, Delgado-Bolton RC,
59. Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG Jr, Ryan T, et al. Sanchez-Enrique C, Vilacosta I, Vivas D, et al. Assessment of the
Proposed modifications to the Duke criteria for the diagnosis of diagnostic accuracy of 18F-FDG PET/CT in prosthetic infective
infective endocarditis. Clin Infect Dis. 2000;30(4):633–8. endocarditis and cardiac implantable electronic device infection:
60. Habib G, Lancellotti P, Antunes MJ, Bongiorni MG, Casalta JP, comparison of different interpretation criteria. Eur J Nucl Med
Del Zotti F, et al. 2015 ESC guidelines for the management of Mol Imaging. 2016;43(13):2401–12.
infective endocarditis: The Task Force for the Management of 79. Hyafil F, Rouzet F, Lepage L, et al. Role of radiolabelled leuco-
Infective Endocarditis of the European Society of Cardiology cyte scintigraphy in patients with a suspicion of prosthetic valve
(ESC). Endorsed by: European Association for Cardio-Thoracic endocarditis and inconclusive echocardiography. Eur Heart J
Surgery (EACTS), the European Association of Nuclear Medicine Cardiovasc Imaging. 2013;14:586–94.
(EANM). Eur Heart J. 2015;36(44):3075–128. 80. Gomes A, Glaudemans AW, Touw DJ, et al. Diagnostic value of
61. Neelankavil JP, Thompson A, Mahajan A. Managing cardiovascu- imaging in infective endocarditis: a systematic review. Lancet
lar implantable devices during perioperative care. The Anesthesia Infect Dis. 2017;17:e1–e14.
Patient Safety Foundation [Internet]. 2013. Available from: https:// 81. Saby L, Laas O, Habib G, et al. Positron emission tomogra-
www.apsf.org/newsletters/html/2013/fall/01_cieds.htm. phy/computed tomography for diagnosis of prosthetic valve
62. Podoleanu C, Deharo JC. Management of cardiac implant- endocarditis: increased valvular 18
F-fluorodeoxyglucose
able electronic device infection. Arrhythm Electrophysiol Rev. uptake as a novel major criterion. J Am Coll Cardiol. 2013;61:
2014;3:184–9. 2374–82.
63. Voigt A, Shalaby A, Saba S. Rising rates of cardiac rhythm man- 82. Pizzi MN, Roque A, Fernández-Hidalgo N, et al. Improving
agement device infections in the United States: 1996 through the diagnosis of infective endocarditis in prosthetic valves and
2003. J Am Coll Cardiol. 2006;48:590–1. Intracardiac devices with 18F-Fluordeoxyglucose positron emis-
64. Margey R, McCann H, Blake G, et al. Contemporary management sion tomography/ computed tomography angiography: initial
of and outcomes from cardiac device related infections. Europace. results at an infective endocarditis referral center. Circulation.
2010;12:64–70. 2015;132:1113–26.
83. Vilacosta I, Graupner C, San Roman JA, et al. Risk of emboliza- in diagnosing aortic prosthetic graft infection. Eur J Nucl Med
tion after institution of antibiotic therapy for infective endocardi- Mol Imaging. 2017;44:886–94.
tis. J Am Coll Cardiol. 2002;39:1489–95. 103. Tossios P, Karatzopoulos A, Tsagakis K, et al. Successful surgical in
84. Dell’Aquila AM, Mastrobuoni S, Alles S, et al. Contributory role situ treatment of prosthetic graft infection by staged procedure after
of fluorine 18-Fluorodeoxyglucose positron emission tomography/ Bentall operation and total aortic arch replacement. Springerplus.
computed tomography in the diagnosis and clinical Management 2014;3:172. https://doi.org/10.1186/2193-1801-3-172.
of Infections in patients supported with a continuous-flow left 104. Legout L, D’Elia PV, Sarraz-Bournet, et al. Diagnosis and man-
ventricular assist device. Ann Thorac Surg. 2016;101:87–94. agement of prosthetic vascular graft infections. Méd Mal Infect.
85. Avramovic N, Dell’Aquila AM, Weckesser M, et al. Metabolic 2012;42:102–9.
volume performs better than SUVmax in the detection of left ven- 105. Hargrove WC III, Edmunds LH Jr. Management of infected tho-
tricular assist device driveline infection. Eur J Nucl Med Mol racic aortic prosthetic grafts. Ann Thorac Surg. 1984;37:72–7.
Imaging. 2017;44:1870–7. 106. Takano T, Terasaki T, Wada Y, et al. Treatment of prosthetic graft
86. Di Salvo G, Habib G, Pergola V, et al. Echocardiography pre- infection after thoracic aorta replacement. Ann Thorac Cardiovasc
dicts embolic events in infective endocarditis. J Am Coll Cardiol. Surg. 2014;20:304–9.
2001;37:1069–107. 107. Aretz HT, Billingham ME, Edwards WD, et al. Myocarditis: a
87. Juneau D, Golfam M, Hazra S, Zuckier LS, Garas S, Redpath C, histopathologic definition and classification. Am J Cardiovasc
Bernick J, et al. Positron emission tomography and single-photon Pathol. 1987;1:3–14.
emission computed tomography imaging in the diagnosis of car- 108. Lieberman EB, Hutchins GM, Herskowitz A, et al. Clinico-
diac implantable electronic device infection: a systematic review pathologic description of myocarditis. J Am Coll Cardiol.
and meta-analysis. Circ Cardiovasc Imaging. 2017;10(4). pii: 1991;18:1617–26.
e005772. https://doi.org/10.1161/CIRCIMAGING.116.005772. 109. Magnani JW, Dec GW. Myocarditis: current trends in diagnosis
88. Vos FJ, Bleeker-Rovers CP, Sturm PD, et al. 18F-FDG PET/CT and treatment. Circulation. 2006;113:876–90.
for detection of metastatic infection in gram-positive bacteremia. 110. Laissy JP, Hyafil F, Feldman LJ, et al. Differentiating acute
J Nucl Med. 2010;51:1234–40. myocardial infarction from myocarditis: diagnostic value of
89. Orvin K, Goldberg E, Bernstine H, et al. The role of FDG-PET/ early- and delayed-perfusion cardiac MR imaging. Radiology.
CT imaging in early detection of extra-cardiac complications of 2005;237:75–82.
infective endocarditis. Clin Microbiol Infect. 2015;21:69–76. 111. Takano H, Nakagawa K, Ishio N, et al. Active myocarditis in a
90. Kestler M, Munoz P, Rodriguez-Creixems M, et al. Role of 18F- patient with chronic active Epstein–Barr virus infection. Int J
FDG PET in patients with infectious endocarditis. J Nucl Med. Cardiol. 2008;130:e11–3.
2014;55:1093–8. 112. Oikawa M, Kagaya Y, Otani H, et al. Increased [18F]fluorodeoxy-
91. Ozcan C, Asmar A, Ggill S, et al. The value of FDG-PET/CT glucose accumulation in right ventricular free wall in patients with
in the diagnostic work-up of extracardiac infectious manifes- pulmonary hypertension and the effect of epoprostenol. J Am Coll
tations in infectious endocarditis. Int J Cardiovasc Imaging. Cardiol. 2005;45:1849–55.
2013;29:1629–37. 113. O’Leary SM, Williams PL, Williams MP, et al. Imaging the peri-
92. Delahaye F, M’Hammedi A, Guerpillon B, et al. Systematic search cardium: appearances on ECG-gated 64-detector row cardiac
for present and potential portals of entry for infective endocarditis. computed tomography. Br J Radiol. 2010;83:194–205.
J Am Coll Cardiol. 2016;67:151–8. 114. Wang ZJ, Reddy GP, Gotway MB, et al. CT and MR imaging of
93. McCoy CW, Mason JM. Enterococcal endocarditis associated pericardial disease. Radiographics. 2003;23(Spec No):S167–80.
with carcinoma of the sigmoid; report of a case. J Med Assoc State 115. Alter P, Figiel JH, Rupp TP, et al. MR, CT, and PET imaging in
Ala. 1951;21:162–6. pericardial disease. Heart Fail Rev. 2013;18:289–306.
94. Berger P, Vaartjes I, Moll FL, et al. Cumulative incidence of graft 116. Dong A, Dong H, Wang Y, et al. 18F-FDG PET/CT in differenti-
infection after primary prosthetic aortic reconstruction in the ating acute tuberculous from idiopathic pericarditis: preliminary
endovascular era. Eur J Vasc Endovasc Surg. 2015;49:581–5. study. Clin Nucl Med. 2013;38:e160–5.
95. Hicks RC, Greenhalgh RM. The pathogenesis of vascular graft 117. Couturier B, Huyge V, Soyfoo MS. Pericarditis revealing large
infection. Eur J Vasc Endovasc Surg. 1997;14:5–9. vessel vasculitis. ISRN Rheumatol. 2011;2011:648703. https://
96. Kilic A, Arnaoutakis DJ, Reifsnyder T, Black JH III, et al. doi.org/10.5402/2011/648703.
Management of infected vascular grafts. Vasc Med. 2016;21:53–60. 118. Birnie DH, Sauer WH, Bogum F, Cooper JM, Culver DA,
97. Low RN, Wall SD, Jeffrey RB Jr, et al. Aortoenteric fistula and peri- Duvernoy CS, et al. HRS expert consensus statement on the diag-
graft infection: evaluation with CT. Radiology. 1999;175:157–62. nosis and management of arrhythmias associated with cardiac sar-
98. Orton DF, LeVeen RF, Saigh JA, et al. Aortic prosthetic graft coidosis. Heart Rhythm. 2014;11:1304–23.
infections: radiologic manifestations and implications for man- 119. Cooper LT, Baughman KL, Feldman AM, Frustaci A, Jessup
agement. Radiographics. 2000;20:977–93. M, Kuhl U, et al. The role of endomyocardial biopsy in the
99. Liberatore M, Misuraca M, Calandri E, et al. White blood cell management of cardiovascular disease: a scientific statement
scintigraphy in the diagnosis of infection of endovascular pros- from the American Heart Association, the American College of
theses within the first month after implantation. Med Sci Monit. Cardiology, and the European Society of Cardiology Endorsed
2006;12:5–9. by the Heart Failure Society of America and the Heart Failure
100. Erba PA, Leo G, Sollini M, et al. Radiolabelled leucocyte scintig- Association of the European Society of Cardiology. Eur Heart J.
raphy versus conventional radiological imaging for the manage- 2007;28(24):3076–93.
ment of late, low-grade vascular prosthesis infections. Eur J Nucl 120. Simonen P, Lehtonen J, Kandolin R, Schildt J, Marjasuo S, Miettinen
Med Mol Imaging. 2014;41:357–68. H, et al. F-18-fluorodeoxyglucose positron emission tomography-
101. Spacek M, Belohlavek O, Votrubova J, Sebesta P, Stadler guided sampling of mediastinal lymph nodes in the diagnosis of
P. Diagnostics of “non-acute” vascular prosthesis infection using cardiac sarcoidosis. Am J Cardiol. 2015;116(10):1581–5.
18
F-FDG PET/CT: our experience with 96 prostheses. Eur J Nucl 121. Blankstein R, Osborne M, Naya M, Waller A, Kim CK, Murthy
Med Mol Imaging. 2009;36:850–8. VL, et al. Cardiac positron emission tomography enhances prog-
102. Saleem BR, Beukinga RJ, Boellaard R, et al. Textural features of nostic assessments of patients with suspected cardiac sarcoidosis.
18
F-fluorodeoxyglucose positron emission tomography scanning J Am Coll Cardiol. 2014;63(4):329–36.
122. Tang R, Wang JT, Wang L, Le K, Huang Y, Hickey AJ, Emmett 134. Verberne HJ, Feenstra C, de Jong WM, Somsen GA, van Eck-
L. Impact of patient preparation on the diagnostic performance Smit BL, Busemann Sokole E. Influence of collimator choice
of 18F-FDG PET in cardiac sarcoidosis: a systematic review and and simulated clinical conditions on 123I-MIBG heart/medias-
meta-analysis. Clin Nucl Med. 2016;41:e327–39. tinum ratios: a phantom study. Eur J Nucl Med Mol Imaging.
123. Youssef G, Leung E, Mylonas I, Nery P, Williams K, Wisenberg 2005;32(9):1100–7.
G, et al. The use of 18F-FDG PET in the diagnosis of cardiac 135. Verschure DO, de Wit TC, Bongers V, Hagen PJ, Sonneck-

sarcoidosis: a systematic review and metaanalysis including the Koenne C, D’Aron J, et al. 123I-MIBG heart-to-mediastinum
Ontario experience. J Nucl Med. 2012;53(2):241–8. ratio is influenced by high-energy photon penetration of colli-
124. Schneider S, Batrice A, Rischpler C, Eiber M, Ibrahim T, Nekolla mator septa from liver and lung activity. Nucl Med Commun.
SG. Utility of multimodal cardiac imaging with PET/MRI in car- 2015;36(3):279–85.
diac sarcoidosis: implications for diagnosis, monitoring and treat- 136. Veltman CE, Boogers MJ, Meinardi JE, Al Younis I, Dibbets-
ment. Eur Heart J. 2014;35(5):312. Schneider P, Van der Wall EE, et al. Reproducibility of planar
125. Ruberg FL, Appelbaum E, Davidoff R, Ozonoff A, Kissinger KV, 123
I-meta-iodobenzylguanidine (MIBG) myocardial scintigra-
Harrigan C, et al. Diagnostic and prognostic utility of cardiovas- phy in patients with heart failure. Eur J Nucl Med Mol Imaging.
cular magnetic resonance imaging in light-chain cardiac amyloi- 2012;39(10):1599–608.
dosis. Am J Cardiol. 2009;103(4):544–9. 137. Klene C, Jungen C, Okuda K, Kobayashi Y, Helberg A, Mester J,
126. Perugini E, Guidalotti PL, Salvi F, Cooke RM, Pettinato C, Riva et al. Influence of ROI definition on the heart-to-mediastinum ratio
L, et al. Noninvasive etiologic diagnosis of cardiac amyloidosis in planar 123I-MIBG imaging. J Nucl Cardiol. 2018;25(1):208–16.
using 99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid scin-
138. Okuda K, Nakajima K, Hosoya T, Ishikawa T, Konishi T,
tigraphy. J Am Coll Cardiol. 2005;46(6):1076–84. Matsubara K, et al. Semi-automated algorithm for calculating
127. Trivieri MG, Dweck MR, Abgral R, Robson PM, Karakatsanis heart-to-mediastinum ratio in cardiac Iodine-123 MIBG imaging.
NA, Lala A, et al. 18F-sodium fluoride PET/MR for the assessment J Nucl Cardiol. 2011;18(1):82–9.
of cardiac amyloidosis. J Am Coll Cardiol. 2016;68(24):2712–4. 139. Jacobson AF, Senior R, Cerqueira MD, Wong ND, Thomas

128. Antoni G, Lubberink M, Estrada S, Axelsson J, Carlson K, Lindsjo GS, Lopez VA, et al. Myocardial iodine-123 meta-
L, et al. In vivo visualization of amyloid deposits in the heart with iodobenzylguanidine imaging and cardiac events in heart failure.
11
C-PIB and PET. J Nucl Med. 2013;54(2):213–20. Results of the prospective ADMIRE-HF (AdreView Myocardial
129. Dorbala S, Vangala D, Semer J, Strader C, Bruyere JR Jr, Di Imaging for Risk Evaluation in Heart Failure) study. J Am Coll
Carli MF, et al. Imaging cardiac amyloidosis: a pilot study using Cardiol. 2010;55(20):2212–21.
18
F-florbetapir positron emission tomography. Eur J Nucl Med
140. Puppo C, Massollo M, Paparo F, Camellino D, Piccardo A,
Mol Imaging. 2014;41(9):1652–62. Shoushtari Zadeh Naseri M, et al. Giant cell arteritis: a systematic
130. Law WP, Wang WY, Moore PT, Mollee PN, Ng AC. Cardiac amy- review of the qualitative and semiquantitative methods to assess
loid imaging with 18F-Florbetaben PET: a pilot study. J Nucl Med. vasculitis with 18F-fluorodeoxyglucose positron emission tomog-
2016;57(11):1733–9. raphy. Biomed Res Int. 2014;2014:574248.
131. Capitanio S, Marini C, Bauckneht M, Sambuceti G. Nuclear 141. Hooisma GA, Balink H, Houtman PM, Slart RH, Lensen
cardiology in heart failure. Curr Cardiovasc Imaging Rep. KD. Parameters related to a positive test result for FDG PET(/CT)
2014;7(3):9256. for large vessel vasculitis: a multicenter retrospective study. Clin
132. Schinkel AF, Bax JJ, Poldermans D, Elhendy A, Ferrari R, Rheumatol. 2012;31(5):861–71.
Rahimtoola SH. Hibernating myocardium: diagnosis and patient 142. Moosig F, Czech N, Mehl C, Henze E, Zeuner RA, Kneba M,
outcomes. Curr Probl Cardiol. 2007;32(7):375–410. et al. Correlation between 18-fluorodeoxyglucose accumulation
133. Bauckneht M, Sambuceti G, Pomposelli E, Fiz F, Marini in large vessels and serological markers of inflammation in poly-
C. Pathophysiological basis of myocardial innervation imaging in myalgia rheumatica: a quantitative PET study. Ann Rheum Dis.
heart failure. Clin Transl Imaging. 2015;3(5):347–55. 2004;63(7):870–3.
Radionuclide Imaging of Benign
Pulmonary Diseases 21
Federica Guidoccio, Edoardo Airò, and Giuliano Mariani
Contents
21.1 General Background on the Role of Radionuclide Imaging
in Benign Pulmonary Diseases 500
21.2 Anatomy of the Lungs 500
21.3 Radiopharmaceuticals Used in Benign Pulmonary Diseases 502
21.4 Lung Perfusion Imaging 502
21.5 Lung Ventilation Scintigraphy with Radioaerosol 503
21.6 Lung Ventilation Scintigraphy with Inert Radioactive Gases 503
21.6.1 Ventilation Imaging with 133Xe 503
21.6.2 Ventilation Imaging with 81mKr 503
21.7 Sequence of Imaging 506
21.8 SPECT/Low-Dose CT (SPECT/CT) 506
21.9 Ga-Citrate Scintigraphy
67
506
21.10 Clinical Indications 507
21.11 Acute Pulmonary Embolism 507
21.12 Follow-Up of Patients After Acute Pulmonary Embolism 514
21.13 Obstructive Lung Disease 515
21.14 Quantification of Differential Pulmonary Function Before Pulmonary Surgery 516
21.15 Interstitial Lung Disease 516
References 519
Learning Objectives • Acquire information on anatomy and physiology of the

• Understand the basic notions on pathophysiology and lungs as they relate to the distribution of diagnostic radio-
clinical presentation of the most common benign pulmo- pharmaceuticals that can be utilized for imaging in
nary diseases that can be managed with the help of radio- patients with benign pulmonary diseases.
nuclide imaging. • Learn the protocols utilized for lung perfusion scintigra-
phy and lung ventilation scintigraphy.
F. Guidoccio (*) · G. Mariani • Learn the protocols utilized for 67Ga-citrate
Regional Center of Nuclear Medicine, Department of Translational scintigraphy.
University of Pisa, Pisa, Italy • Understand the clinical indications for employing radio-
nuclide imaging in patients with benign pulmonary dis-
E. Airò
Fondazione CNR/Regione Toscana Gabriele Monastero, eases, including acute pulmonary embolism, obstructive
Pisa, Italy lung disease, and interstitial lung disease.

500 F. Guidoccio et al.
21.1 G
eneral Background on the Role death in 13% of autopsies [5]. In the absence of risk factors,
of Radionuclide Imaging in Benign PE is rare in children under 15 years of age (<5 per 100,000),
Pulmonary Diseases whereas it increases dramatically after the age of 60 years
[3]. The major recognized risk factor for PE is venous throm-
Molecular imaging, which is based on the ability to image boembolism (VTE), whose incidence is similar in men as in
alterations underlying diseases at the subcellular, molecular women. In turn, risk factors associated with the development
level, has the potential not only to detect disease early on— of VTE include inherited and acquired factors [6]; the most
before anatomic abnormalities become obvious at morpho- important acquired factors include underlying malignancy
logic imaging—but also to monitor the efficacy of and recent immobilization or surgery. A more recently rec-
treatment(s) before anatomic changes take place in response ognized risk factor is long-distance flights, even in healthy
to therapy. These features imply the possibility to detect/ individuals without other risk factors [7].
diagnose disease early and accurately and to predict response The term “lung scintigraphy” is primarily used to indicate
to therapy early on during treatment, therefore to facilitate imaging obtained after systemic intravenous administration
implementation in the medical practice of a true personal- of 99mTc-labeled macroaggregates of human serum albumin
ized medicine tailored to the needs/characteristics of the (99mTc-MAA), to determine regional lung perfusion. Their dis-
individual patient. tribution of perfusion must be compared to the distribution of
Disease-specific imaging probes are necessary for actual ventilation, since hypoxia causes autoregulation of perfusion,
implementation of molecular imaging with radionuclides reducing blood flow to regions of decreased oxygen. In patients
(with either positron-emitting or single-photon emitting with acute PE, ventilation is maintained in regions of dimin-
agents, as well as with ultrasound or optical imaging). ished perfusion - at least until several h after the acute event. To
Concerning radionuclide imaging per se, when employed as visualize regional lung ventilation, either an inert radioactive
stand-alone techniques, PET and SPECT suffer from gas (e.g., 81mKr, whose use is limited due to high cost and rela-
important limitations, mostly in terms of spatial resolution.tively short half-life of the 81Rb parent) or 133Xe or 127Xe or), a
Many of these limitations can be overcome (or considerably radioactive aerosol (obtained with, e.g., 99mTc-DTPA or 99mTc-
reduced) by adopting multimodal imaging platforms, such as colloidal preparations), or an ultrafine dispersion of 99mTc-
PET/CT, SPECT/CT, and PET/MR. Similar considerations labeled carbon particles are currently recommended, whereas
apply to the development of newer integrated imaging current guidelines of the European Association of Nuclear
modalities involving fluorescence-mediated tomography and Medicine do not longer recommend use of the radioactive gas
fluorescence-mediated tomography combined with MR. 133
Xe—mostly because of limited availability and high burden
The focus of this chapter is radionuclide imaging of of radiation to patients [8].
benign pulmonary diseases, which includes evaluation of
pulmonary embolic disease, pathophysiology of respiratory
function, and infection/inflammation. Key Learning Points
Pulmonary embolism (PE) is an important, life- • Pulmonary embolism is the potentially fatal benign
threatening but treatable illness caused by migration of pulmonary disease where radionuclide imaging is
thrombi to the pulmonary circulation, most commonly from most commonly employed.
deep venous circulation of the lower extremities. If left • Early diagnosis of pulmonary embolism is crucial
untreated, PE can cause death [1] or lead to chronic thrombo- for timely effective treatment.
embolic pulmonary hypertension [2]. Appropriate treatment • Drugs used for therapy of pulmonary embolism
can prevent recurrence of PE and facilitate resolution of the entail the risk of important side effects.
existing thrombi, thus facilitating full recovery. However, • Accurate diagnosis of pulmonary embolism is
effective treatments based on heparin, oral anticoagulants, important also to avoid unnecessary overtreatment
and thrombolytic agents have well-documented side effects. in patients with suspected pulmonary embolism.
Therefore, it is imperative that early diagnosis of PE is made
and that treatment is instituted when appropriate—avoiding
unnecessary overtreatment.
Although the incidence of PE is difficult to establish due
to inaccuracies in hospital discharge records, it is estimated 21.2 Anatomy of the Lungs
to be around 100 cases per 100,000 persons per year [3]. This
estimate has remained quite stable over time, since the inci- The respiratory system includes two major components, the
dence of PE was the same in the 1980s as in the 1990s in the airways and the pulmonary parenchyma. The airways are
United States [4]. One study based on autopsy findings of anatomically subdivided into two segments: the extra-tho-
hospitalized patients has found the prevalence of PE to be racic (superior) and the intrathoracic (inferior) airways,
18%, PE being considered the main or contributory cause of whose main function is to conduct air to the alveolar surface.
21 Radionuclide Imaging of Benign Pulmonary Diseases 501
Starting from the trachea (that has an average diameter of includes a respiratory bronchiole and its branches (alveolar
1.8 cm, corresponding to a total cross-sectional area of ducts, alveolar sacs, and alveoli). The lung parenchyma is
2.54 cm2 in adults), the airways divide by dichotomous further subdivided into lobes and segments, which reflect
branching, with approximately 23 generations of branches approximately the distribution of bronchi. There are three
from the trachea to the alveoli; the average diameter of the lobes on the right (upper, middle, and lower) and two lobes
23rd generation branches (the alveolar sac) is 40 μm, corre- on the left (upper and lower); in turn, each lobe is subdivided
sponding total cross-sectional area of 10.000 cm2. into segments. The right lung comprises ten segments: three
Considering that there are about 300 million pulmonary in the right upper lobe (apical, anterior, and medial), two in
alveoli, the overall alveolar surface area is about 34 m2 dur- the right middle lobe (medial and lateral), and five in the
ing tidal breathing—but expanding to as much as 51–56 m2 right lower lobe (superior, medial, anterior, lateral, and pos-
during breath-in and even 72 m2 during forced respiration. terior). The left lung comprises eight segments: four in the
There are two mainstem bronchi (right and left) and three left upper lobe (apicoposterior, anterior, superior lingula, and
lobar bronchi on the right giving rise to a total of ten segmen- inferior lingula) and four in the left lower lobe (superior,
tal bronchi; two lobar bronchi are found on the left, giving anteromedial, lateral, and posterior). Topographic projection
rise to a total of eight segmental bronchi. of each segment of the two lungs is shown in Fig. 21.1, as
Vascular supply of the trachea and bronchial tree originates reference during interpretation of lung scintigraphy.
from branches from the inferior thyroid arteries, intercostal The vascular system of the lungs starts in the right ven-
arteries, and bronchial arteries. These arteries form a peribron- tricle with the main pulmonary artery, which divides into the
chial plexus that follows the bronchial tree deep into the lung right and left pulmonary arteries, which carry blood into
parenchyma, supplying blood also to the visceral pleura and to each lung—right and left, respectively. The pulmonary
the walls of the pulmonary arteries and veins (vasa vasorum). branches of the pulmonary arteries follow bronchial distribu-
Lung parenchyma includes part of the conduction system tion until the alveolar plexus. After exchange of respiratory
but is mainly involved in gas exchange at the alveolar level. gases at the alveolar level, alveolar capillaries carry the oxy-
The elementary functional unit of the lung is the pulmonary genated blood back to the heart (left atrium) via the pulmo-
lobule; each lobule is independent and autonomous and nary veins.
Fig. 21.1 Diagrammatic
representation according to
orthogonal views (anterior,
posterior, right lateral, left
lateral) of the lung segments,
indicated with different color
codes, originating from
branching of the
tracheobronchial tree; each
segment receives its blood
supply through a specific
terminal branch of the
pulmonary artery (modified
from: Volterrani D, Erba PA,
di Medicina Nucleare – Anterior
Tecniche e Applicazioni. Posterior
Milan: Springer-Verlag Italy;
2010)
Apical of upper lobe
Posterior
Anterior
Apical of lower lobe
Anterobasal
Laterobasal
Posterobasal
Lateral of middle lobe
Medial of middle lobe
Superior of lingula
Right Left
Inferior of lingula
21.3 R
adiopharmaceuticals Used in Benign In addition to its predominant use in oncology, the PET
Pulmonary Diseases tracer 2-[18F]fluoro-2-deoxy-d-glucose ([18F]FDG) can be
used to investigate benign pulmonary diseases such as sar-
Single-photon and positron-emitting radiopharmaceuticals coidosis. In fact, the glucose transporter GLUT-1 is overex-
used for imaging benign pulmonary diseases are presented in pressed also on the surface of activated granulocytes,
details in Chaps. 2 and 3 of this book, respectively. We sum- lymphocytes, and monocytes. Thus, [18F]FDG is useful for
marize here the main features that justify the use of such PET imaging of sarcoidosis, tuberculosis, occult infection,
radiopharmaceuticals in patients with benign pulmonary autoimmune fibrosis, pneumonia, cystic fibrosis, and acute
diseases. lung injury or acute respiratory distress syndrome (although
99m
Tc-Macroaggregated albumin (99mTc-MAA) is adminis- not with the same level of clinical validation for all these
tered i.v. for performing lung perfusion scintigraphy, based applications).
on a microembolization mechanism. The usual adult- Similar to the PET tracer [18F]FDG, the single-photon
administered activity is 40–150 MBq in adults, reduced to emitting agent 67Ga-citrate is accumulated by cells that have
1.11 MBq/kg in children (with a minimum of 14.8 MBq) if an increased metabolic rate—including both malignant cells
no 99mTc ventilation study is performed or 2.59 MBq/kg if a and inflammatory cells. The iron analog gallium in the ionic
99m
Tc ventilation study is performed. In adults the number of form Ga++ (that forms at physiological pH following injec-
particles administered should be in the range of 200,000– tion of 67Ga-citrate) binds to transferrin to form a gallium-
700,000 which in children is reduced as a function of age; transferrin complex that in turn binds to transferrin receptors
this number corresponds in general to <0.3% of the overall and is incorporated intracellularly after interacting with lac-
capillary bed in the lungs; therefore it is negligible from the toferrin receptors and other siderophores. Although the use
point of view of lung perfusion and respiratory gas exchanges. of 67Ga-citrate has been recently overshadowed by newer
Freshly prepared 99mTc-MAA with reduced numbers of par- molecular agents such as [18F]FDG and by high-resolution
ticles should be considered for patients with pulmonary CT, there are still instances where 67Ga-citrate remains the
hypertension or right-to-left shunting. In adults, the number radiotracer of choice. In particular, 67Ga-citrate scintigraphy
may be reduced to 100,000–200,000 particles without alter- is a highly sensitive diagnostic technique for fever of
ing the quality of the images for detection of perfusion unknown origin, sarcoidosis, idiopathic pulmonary fibrosis,
defects, whereas inhomogeneous distribution of activity may Pneumocystis carinii pneumonia (PCP, also known as PJP,
result when reducing the number of particles below 100,000. after Pneumocystis jirovecii pneumonia), and drug-induced
Since the 99mTc-MAA particles settle in the vial with time, pulmonary toxicity.
the vial containing the radiolabeled preparation should be
swirled before withdrawing a diagnostic aliquot, and the
syringe should be repeatedly inverted before injection. Key Learning Points
Lung ventilation agents include a variety of preparations • Lung perfusion scintigraphy is most commonly
that can be subdivided into fully diffusible radiopharma- achieved by microembolization with 99mTc- macro-
ceuticals (the inert radioactive gases Kr and Xe) and
81m 133
aggregates of human albumin (99mTc-MAA).
radioactive aerosols (such as those obtained with 99mTc- • For ventilation studies, the inert radioactive gas
DTPA and 99m
Tc-colloids) or ultrafine dispersions (i.e., 81m
Kr and radio-aerosols with 99mTc-diethylene-
99m
Tc-Technegas). triaminepentaacetate (99mTc-DTPA) or with radio-
Whereas diffusible gases mix with the gas in the entire colloids or finally with an ultrafine dispersion of
ventilation space, aerosols are composed of small particles 99m
Tc-labeled carbon (99mTc-Technegas) are cur-
that reach the alveoli. In patients with substantial airway rently recommended.
disease, there is turbulent airflow in the central airways, • Other radiopharmaceuticals used for imaging in
resulting in aerosol deposition in the major bronchi, thus benign pulmonary diseases include [18F]FDG and
limiting the evaluation of the airflow to the alveoli. In addi- 67
Ga-citrate.
tion to imaging regional distribution of ventilation as a
diagnostic aid in patients with suspected pulmonary embo-
lism, lung ventilation scintigraphy is useful also for charac-
terizing patients with other conditions, e.g., chronic
obstructive pulmonary disease. Furthermore, not only is 21.4 Lung Perfusion Imaging
regional distribution of ventilation clinically relevant, but
also the rate of clearance from the lungs provides useful In addition to a complex tuning of the ventilation/perfusion
information on permeability of the alveolar epithelium or match determined by intrapulmonary reflex mechanisms
on the integrity of the mucociliary transport mechanism of (see further below), distribution of blood within the lungs is
pulmonary secretions. influenced by gravity. Therefore, in healthy conditions there
is a physiologic gradient of lung perfusion from base to apex so dictated by clinical conditions. A possible disadvantage of
in the upright position and from back to front in the supine radioaerosol imaging is that aerosol deposition of radioactive
position. This physiologic gradient can be altered in certain droplets/particles can be altered by turbulent flow, and cen-
conditions, for instance, in case of pulmonary hyperten- tral deposition can result in a suboptimal study. SPECT
sion—thus constituting an important diagnostic criterion in imaging (preferably SPECT/CT) is recommended by some
the interpretation of a lung perfusion scintigraphy. For this investigators in addition to planar imaging or as a de novo
reason, it is important to record the position of the patient’s imaging modality.
body at the moment of injecting 99mTc-MAA.
While some authors prefer to administer the perfusion
agent in the supine position, other authors prefer the upright 21.6 Lung Ventilation Scintigraphy
position. In either case, it is recommended that the patient with Inert Radioactive Gases
keep the position chosen (supine or erect) for at least 4–5 min
before administering the pulmonary perfusion agent. The 21.6.1 Ventilation Imaging with 133Xe
patient is also instructed to cough and to take several deep
breaths before injection. 99mTc-MAA is injected slowly dur- An advantage of 133Xe ventilation imaging is that single-
ing 3–5 respiratory cycles. A well-flushed indwelling line breath, wash-in or equilibrium, and washout images can be
can be used if venous access is difficult. The radiopharma- obtained, thus allowing for a more complete characterization
ceutical should not be administered in the distal port of a of ventilation and a more sensitive test for obstructive airway
Swan-Ganz catheter or any indwelling line or port that con- disease. Physiologic information about ventilation can best
tains a filter, for example, a chemotherapy line. be obtained with 133Xe imaging. The imaging room should
Imaging is preferably performed with the patient upright be equipped with an appropriate exhaust/trapping system for
to increase chest cavity size and to minimize diaphragmatic radioactive gas. Regulations for safe handling of radioactive
motion—although this is rarely employed in the clinical rou- gas should be followed. The patient is positioned upright in
tine. If necessary, images can be obtained with the patient in front of the scintillation camera. If necessary, the patient can
the supine or decubitus position. Planar images should be be positioned supine. When performed after performing a
obtained in multiple views including anterior, posterior, both perfusion lung scan, the view that best shows the defects on
posterior oblique, both anterior oblique, and both lateral pro- perfusion scintigraphy is used for the ventilation scan.
jections. Either the anterior oblique or the lateral views can Otherwise, the posterior view is generally acquired. When
be omitted (see Fig. 21.2). In some patients it may be p ossible possible, posterior oblique images should be obtained during
to obtain only limited views. SPECT (preferably SPECT/ the washout phase (and during the wash-in phase if continu-
CT) can be used for a three-dimensional evaluation of perfu- ous imaging is performed during this phase).
sion and is actually recommended by some investigators as a If ventilation scintigraphy is performed after perfusion
routine procedure. Imaging of high-blood-flow systemic scintigraphy, a 99mTc background image should be obtained
organs can be used to detect right-to-left shunting; for using the 133Xe energy window (80 keV versus 140 keV of
instance, images of the brain and kidneys may be obtained to 99m
Tc). A facemask or mouthpiece (with nose clip) should be
distinguish right-to-left shunting from systemic distribution connected via a bacterial filter to the 133Xe delivery system.
of radiopharmaceutical components too small to be trapped Single-breath, equilibrium, and washout images are obtained.
in the capillary network of the lung. SPECT/CT is particu- Equilibrium is obtained by having the patient breathe in a
larly useful because the CT provides detailed information closed 133Xe delivery system for 3–4 min as tolerated. It
about the lyng parenchyma, offering additional information should be noted that use of 133Xe implies a certain radiation
to compare to the perfusion and ventilation images. burden to patients because of the associated β− emission of
this radionuclide; this imaging procedure is therefore not
recommended when alternative approaches to ventilation
21.5 L
ung Ventilation Scintigraphy with scintigraphy are available.
Radioaerosol
The radioaerosol (or the dispersion of 99mTc-labeled ultrafine 21.6.2 Ventilation Imaging with 81mKr
carbon particles) is administered through a mouthpiece, the
nose being occluded while the patient is engaging in tidal The main advantage of 81mKr is that images can be obtained
breathing. The radioaerosol images can be obtained in mul- in all views without interference from prior perfusion imag-
tiple planar views (see Fig. 21.3) or with SPECT (or SPECT/ ing with a 99mTc-labeled agent, due to the emission of γ-rays
CT if possible) to match those obtained for perfusion imag- having 190 keV peak energy. A disadvantage of 81mKr is that
ing. It is preferable for the patient to be upright while inhal- the short half-life of the parent radionuclide, 81Rb (4.57 h),
ing the aerosol, although the supine position is acceptable, if decreases availability and increases the overall cost of the
Fig. 21.2 Example of a normal pattern of lung perfusion scintigraphy right lateral and left lateral in middle row, right posterior oblique and
as visualized after the i.v. injection of 99mTc-MAA, showing a rather left posterior oblique in lower row (the two oblique anterior views are
homogeneous distribution of radioactivity throughout both lung fields, more rarely recorded) (modified from: Volterrani D, Erba PA, Mariani
with a physiologic gravity-dependent perfusion gradient from base to G, Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
apex (and partly also from back to front). The six views most com- Milan: Springer-Verlag Italy; 2010)
monly acquired are reported here: anterior and posterior in upper row,
Fig. 21.3 Example of a normal pattern of lung ventilation scintigraphy radioactivity accumulation in the central airways (trachea and the two
as visualized after inhalation of a radioaerosol prepared with 99mTc- main bronchi). The six views reported here mirror those most com-
nanocolloidal albumin, showing a rather homogeneous distribution of monly acquired for as lung perfusion scan: anterior and posterior in
radioactivity throughout both lung fields, with a physiologic gravity- upper row, right lateral and left lateral in middle row, right posterior
dependent ventilation gradient from base to apex (and partly also from oblique and left posterior oblique in lower row (the two oblique anterior
back to front). The focus of radioactivity accumulation detectable in the views are more rarely recorded). (modified from: Volterrani D, Erba PA,
anterior view below the lungs corresponds to the stomach and is due to Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
ingestion of some of the radioaerosol particles. In some instances Applicazioni. Milan: Springer-Verlag Italy; 2010)
(depending also on the type of radioaerosol employed), there is some
procedure. Alternating 99mTc-MAA and 81mKr imaging allows acquired during a deep breath-holding (as customarily done
perfusion and ventilation images to be obtained without for a diagnostic CT examination) but rather during tidal
patient repositioning between paired 99mTc-MAA and 81mKr breathing. Nevertheless, no definite consensus has been
views. The patient breathes continuously from the 81Rb/81mKr reached on the utility of breathing maneuvers or gating dur-
generator so that, because of the short physical half-life of ing SPECT/CT acquisition [9].
81m
Kr (13 s), changes in the distribution of radioactivity
within the lungs approximate changes in regional ventilation
rate on a time scale of minutes. A collimator with low septal 21.9 67Ga-Citrate Scintigraphy
penetration for the 190 keV γ-rays of 81mKr should be used.
SPECT (preferably SPECT/CT) can be used to evaluate ven- Gallium-67 scintigraphy may include regional, whole-body,
tilation with three-dimensional imaging. planar, and SPECT acquisitions or any combination obtained
after i.v. injection of 67Ga-citrate. About 10–25% of the
injected activity is excreted through the kidneys during the
21.7 Sequence of Imaging first 24 h after injection, the main route of excretion remain-
ing afterward the gastrointestinal tract. At 48 h postinjection,
A chest radiograph should be obtained and reviewed before radioactivity remaining in the body is equally distributed
performing lung scintigraphy. Ventilation scintigraphy with among the liver and spleen, bone and bone marrow, and soft
133
Xe is usually performed before performing the perfusion tissues (with variable physiologic distribution in the naso-
scan. On the other hand, perfusion scintigraphy can be per- pharynx, lacrimal glands, thymus, and breasts). For whole-
formed first and ventilation scintigraphy omitted if deemed body scintigraphy, anterior and posterior views are acquired,
not necessary on the basis of a completely normal perfu- cumulating 1.5–2.0 million counts/whole body, or over
sion scan. 25–35 min, whichever comes first. For adults, this corre-
The disadvantage of performing perfusion imaging before sponds to a minimum scan speed of 6–8 cm/min. Spot views
ventilation imaging with 133Xe is that the perfusion image of the chest are obtained acquiring 250,000–1,000,000 total
contributes substantial background activity (from the down- counts (over 5–20 min). Regional scintigrams of the remain-
scatter of the 140 keV photons of 99mTc) to the ventilation der of the body should be obtained for the same time. The
image (acquired with the 80 keV energy window of 133Xe), wide range in counts and time per image indicated here are
and a decision to perform or not to perform the ventilation justified by wide variability in (1) time elapsed between
study must be taken in a timely manner. On the other hand, tracer injection and imaging and (2) the ability of the patient
the advantage of performing perfusion imaging before venti- to cooperate during image acquisition.
lation imaging with 133Xe is that if the perfusion study is nor-
mal or matches the chest X-ray findings, the ventilation
study can be omitted; if instead the ventilation scan is deemed Key Learning Points
to be necessary, single-view images can be obtained in the • Perfusion scintigraphy is most commonly per-
view(s) that best shows the perfusion defect(s). formed on the basis of microembolization obtained
Because of the higher energy of the γ-rays emission and with systemic intravenous injection of 99mTc-
the short physical half-life of 81mKr, images obtained with macroaggregates of human albumin (99mTc-MAA).
this gas can be alternated with those obtained with • For ventilation studies, the inert radioactive gas
99m
Tc-MAA. 81m
Kr, radio-aerosols based on 99mTc-DTPA (or
When 99mTc-labeled aerosol imaging is performed before 99m
Tc-colloidal preparations), and an ultrafine dis-
performing the 99mTc-MAA perfusion scan, smaller amounts persion of 99mTc-labeled carbon (99mTc-Technegas)
(20–40 MBq) of the 99mTc-labeled aerosol should be are currently recommended.
administered. • Use of the inert radioactive gas 133Xe is currently
not recommended, because of radiation burden
higher than those involved by the use of other lung
21.8 SPECT/Low-Dose CT (SPECT/CT) ventilation imaging agents.
• 67Ga-citrate scintigraphy has traditionally been used
Lung scintigraphy for pulmonary embolism may be per- to characterize patients with interstitial lung dis-
formed using SPECT/CT. The low-dose CT component of ease, primarily sarcoidosis.
the study provides data for attenuation correction and also • More recently, the use of 67Ga-citrate scintigraphy
provides better morphologic information than standard chest in patients with sarcoidosis is being replaced with
X-ray. In order to match the images obtained during SPECT [18F]FDG PETC/CT.
acquisition, the CT portion of the study should not be
21.10 Clinical Indications suspected acute PE, especially because in many instances,
the clinically overt episode of acute PE is preceded by clini-
The most common clinical indication for lung scintigraphy is cally occult episodes.
to confirm or rule out the clinical suspicion of pulmonary Right ventricular failure due to pressure overload is con-
embolism. Less common clinical indications include a vari- sidered the primary cause of death in severe PE. The death
ety of conditions, as follows: rate from PE can be as high as 10% within the first hour. The
overall mortality in untreated patients with PE is 30%, rising
• Evaluate recovery of lung perfusion after pulmonary to 58% in hemodynamically unstable patients. The main
embolism. challenge in the diagnostic workup of patients with clinically
• Quantify differential pulmonary function before pulmo- suspected PE is to accurately and rapidly distinguish patients
nary surgery (usually for lung cancer) [10, 11]. who have the disease and require anticoagulant therapy from
• Evaluate the cause of pulmonary hypertension [12]. patients who do not.
• Evaluate lung transplants [13]. Diagnosis of PE based on clinical data alone is unreliable
• Evaluate congenital heart or lung disease such as cardiac [24], since signs and symptoms such as dyspnea, tachypnea,
shunts, pulmonary arterial stenosis, and arteriovenous fis- tachycardia, pleuritic chest pain, cough, and hemoptysis are
tulae, both at baseline and after surgical correction [14]. nonspecific and common and can be encountered in a wide
• Confirm the presence of bronchopleural fistula [15]. variety of respiratory diseases.
• Evaluate chronic pulmonary parenchymal disorders, Arterial blood gas analysis reveals hypoxemia.
e.g., chronic obstructive pulmonary disease and cystic Nevertheless, although considered a typical finding in acute
fibrosis [16]. PE, up to 40% of the patients with PE have normal arterial
oxygen saturation and 20% a normal alveolar-arterial oxygen
gradient. Hypocapnia is also often present—due to the hyper-
21.11 Acute Pulmonary Embolism ventilation that arises in response to hypoxemia. This clini-
cal pattern was confirmed in both the PIOPED (“Prospective
Venous thromboembolism (VTE) is a frequent but often Investigation of Pulmonary Embolism Diagnosis”) [25–27]
undetected condition that can present with acute pulmonary and the PISA-PED (“Prospective Investigation Study of
embolism (PE) as its most serious clinical manifestation. Acute Pulmonary Embolism Diagnosis”) [28, 29] trials.
Major trauma, surgery, lower limb fractures, joint replace- However, these conditions are nonspecific for PE.
ments, and spinal cord injury are major causes of VTE. Cancer Although there are no highly specific biomarkers for
is also a well-recognized predisposing factor for VTE—most acute PE, the plasma levels of D-dimer (a degradation prod-
likely due to procoagulant factors released in the circulation uct of cross-linked fibrin) are virtually invariably assessed in
by malignant cells. patients with suspected acute PE. In fact, the D-dimer plasma
Acute PE interferes with an optimal status of gas exchange levels are elevated in the presence of acute thrombosis—due
in the lungs by acting not only on perfusion but also on ven- to activation of fibrinolysis. The negative predictive value of
tilation. In fact, there are multiple intrapulmonary adaptation D-dimer testing is high, so that a normal D-dimer level ren-
mechanisms that, within a few hours of blood flow obstruc- ders the diagnosis of acute PE or VTE unlikely. On the other
tion, divert ventilation from non-perfused to perfused lung hand, the production of fibrin (with the consequent associ-
zones and thus avoid ventilating areas where respiratory gas ated fibrinolysis) is enhanced also in a wide variety of condi-
exchanges cannot occur [17, 18]. Conversely, based on tions such as cancer, inflammation, bleeding, trauma, surgery,
reversed adaptation mechanisms, perfusion is diverted from and necrosis. Therefore, the positive predictive value of ele-
poorly ventilated to better ventilated spaces, in order to opti- vated D-dimer levels is low, and D-dimer testing is not useful
mize the ventilation/perfusion ratio and therefore to maxi- for confirmation of PE.
mize respiratory gas exchanges. Early after embolization Chest X-ray is frequently abnormal; although the findings
(within minutes to a few hours), the mechanisms that divert (that include amputation of hilar artery, focal oligemia,
ventilation from non-perfused to normally perfused lung pleural-based consolidation) are usually nonspecific for PE,
areas are linked to local bronchoconstriction associated with they are useful for excluding other causes of dyspnea or
hypocapnia [19–21]. Whereas, a slower but more important chest pain [30, 31]. A preliminary chest X-ray (or preferably
mechanism takes place within 18–24 h after embolization, a chest CT) is important in all patients for excluding alterna-
i.e., reduced production of surfactant distal to embolization tive readily diagnosable conditions (pulmonary edema,
that causes shrinkage of the alveolar space in the non- pneumonia, trauma, pneumothorax, etc.) and as an aid in the
perfused region [22, 23]. It is important to keep in mind this interpretation of subsequent imaging tests.
sequence of events when interpreting lung scintigraphy Electrocardiographic changes indicative of right ventricle
(especially the ventilation/perfusion scan) in patients with strain, such as inversion of T waves in leads V1–V4, a QR
pattern in V1, S1Q3T3 pattern, and incomplete or complete MAA. If this radiopharmaceutical is homogeneously mixed
right bundle-branch block, may be helpful in the diagnostic with circulating blood, its distribution within the lungs mirrors
pathway. These EKG changes are usually found in more distribution of blood flow. Therefore, if blood flow originating
severe cases of PE; in milder cases, the only abnormality from the pulmonary artery is arrested in some segments of the
may be sinus tachycardia, present in 40% of patients. Finally, lungs because of embolization, these segments (or subseg-
atrial arrhythmias, most frequently atrial fibrillation, may be ments) appear as “cold” areas in the scan (perfusion defects)
associated with acute PE. (see Fig. 21.4) [42, 43]. With very few exceptions (such as
A number of systems have been developed to aid clini- central, non- obstructing PE causing an evenly distributed
cians in establishing the likelihood of pulmonary thrombo- reduction in whole lung perfusion, or minimal perfusion
embolic disease [32]. The main challenge in diagnostic defects below the resolution power of scintigraphy), a normal
workup of patients with suspected acute PE is to accurately perfusion scan excludes the diagnosis of PE.
and rapidly distinguish the average 25% of patients who Such extremely high sensitivity and negative predictive
have the disease (and require thrombolytic therapy) from value of lung perfusion scintigraphy are however associ-
those 75% who do not (and should be spared the potential ated with low specificity. In fact, defects in lung perfusion
adverse side effects of such unnecessary therapy). can be observed not only in the presence of pulmonary
Major hospitals have developed sequential diagnostic embolism but also in a wide variety of conditions, e.g., in
algorithms for patients with suspected acute PE. The first step prior unresolved PE, that can affect as many as 35% of
in such algorithms is represented by accurate clinical history, patients with acute PE [44]. Focal perfusion defects can be
physical examination, and some basic instrumental evalua- observed also in other conditions, including compression
tions (such as EKG, arterial blood gas analysis, etc.). There or invasion of pulmonary vessels by tumors or granulo-
are at least three internationally recognized systems that have mata, emphysema (especially in bullous disease), intersti-
been developed with the purpose of assessing, before employ- tial fibrosis, bronchiectasis, pneumonic consolidation and
ing more complex imaging procedures, the probability that a atelectasis, localized bronchial obstruction, vasculitis, arte-
given patient with suspected acute PE actually has the disease riovenous fistulae, etc. [45, 46]. Most of these conditions
(pretest probability) [33]. Of these, the most frequently used primarily affect ventilation and subsequently reduce local
prediction score is the one offered by Wells et al. [34]. This perfusion because of the intrapulmonary adaptation mecha-
prediction model has been validated extensively using both a nisms that attempt to maximize the efficiency of pulmonary
three-category score (low, moderate, or high clinical proba- gas exchanges by diverting blood flow from underventi-
bility of PE) and a two-category score (PE likely or unlikely). lated to better ventilated areas.
It is simple and based on information that is easy to obtain; on The perfusion defect observed in case of acute PE is typi-
the other hand, the weight of one subjective item (“alternative cally wedge-shaped (segmental or subsegmental) with the
diagnosis less likely than PE”) may reduce interobserver base of the wedge facing the pleura and the apex toward the
reproducibility of the Wells model. The Geneva score [35] is hilum. In proximity of this cold area, it is possible to recog-
also simple and standardized. More recently, both the Wells nize some areas of forced perfusion due to the redistribution
and the revised Geneva models were further simplified in an of blood flow in these non-embolized vessels. Such perfu-
attempt to increase their adoption into clinical practice, and sion defects typically have a segmental type of distribution.
the simplified versions were externally validated [36, 37]. In addition to perfusion defects, intrapulmonary distribution
Whichever is used, the proportion of patients with confirmed of blood as visualized by the 99mTc-MAA particles most fre-
PE can be expected to be around 10% in the low-probability quently shows loss (and actually even inversion) of the phys-
category, 30% in the moderate-probability category, and 65% iologic distribution gradient, either base to apex or back to
in the high-clinical- probability category when using the front—depending on the position of the patient at the time of
three-level classification. When the two-category classifica- injecting 99mTc-MAA.
tion is used, the proportion of patients with confirmed PE in Interpretation of the lung scans is standardized, and lung
the PE-unlikely category is around 12%. The “Pisa score” scan findings are frequently classified according to a three-tier
[38, 39] also constitutes a validated model for clinical predic- classification: normal scan (excluding PE), high-probability
tion of pulmonary embolism [40, 41]. Based on the above scan (considered diagnostic of PE in most patients), and non-
considerations, when clinical presentation raises the suspi- diagnostic scan. A nondiagnostic scan indicates the necessity
cion of PE in an individual patient, it is really important to for further diagnostic testing [47].
perform diagnostic test(s) in order to confirm or rule out the According to the revised PIOPED II criteria (Table 21.1), a
actual occurrence of PE. perfusion lung scan in a patient without acute PE is character-
The fundamental approach to radionuclide imaging in ized by non-wedge-shaped perfusion defects or a contour defect
patients with suspected PE has been lung perfusion scintigra- caused by enlarged heart, mediastinum, or diaphragm or near-
phy based on the systemic intravenous injection of 99mTc- normal or normal perfusion. A high-probability scan is charac-
a b
c d
e f
Fig. 21.4 Identification of perfusion defects in specific lung segments in middle row, right lateral and left lateral in lower row). There are
after the i.v. injection of 99mTc-MAA in a patient with acute pulmonary multiple perfusion defects with clear segmental or subsegmental distri-
embolism. Each planar scan is associated with a diagram indicating the bution, associated with some areas of forced perfusion due to the redis-
average projection of each lung segment for each view (anterior and tribution of blood flow in these non-embolized vessels. This pattern is
posterior in upper row, right posterior oblique and left posterior oblique consistent with pulmonary embolism
Table 21.1 Interpretation of lung ventilation/perfusion scintigraphy (V/Q) according to the revised PIOPED II criteria for diagnosing acute PE
Classification Scintigraphic pattern
Very low probability of PE No detectable perfusion defects
Nonsegmental perfusion abnormalities attributable to X-ray abnormalities (enlargement of the heart or
hilum, elevated hemidiaphragm, costophrenic angle effusion, linear atelectasis), without other perfusion
defects
Perfusion defect smaller than corresponding X-ray lesion ≥2 matched V/Q defects without associated
X-ray lesion 1–3 small subsegmental perfusion defects (<25% of each segment)
Solitary matched segmental V/Q defect with associated matching X-ray opacity in the mid or upper lung
zone
Scintigraphic stripe sign, i.e., a stripe of perfused lung tissue between a perfusion defect and the adjacent
pleural surface (best seen on tangential oblique views)
X-ray detectable pleural effusion of >1/3 of pleural cavity, without other perfusion defects in either lung
High probability of PE ≥2 segmental perfusion defects with V/Q mismatch
Low or intermediate probability All other findings
of PE (nondiagnostic scan)
Modified from: Sostman HD, Stein PD, Gottschalk A, Matta F, Hull R, Goodman L. Acute pulmonary embolism: sensitivity and specificity of
ventilation-perfusion scintigraphy in PIOPED II study. Radiology 2008;246:941–6
terized by one or more wedge-shaped perfusion defects in the PISA-PED protocol; according to the stringent PISA-PED
presence of normal lung parenchyma in this region on chest criteria, the results of lung perfusion scintigraphy are classi-
X-ray. A nondiagnostic scan is a scan that cannot be classified in fied as either positive or negative for acute PE—without
any of the other categories. therefore the need of introducing a nondiagnostic category
The ventilation scan has been added to the perfusion scan (Table 21.2) (see Figs. 21.5 and 21.6).
in the diagnostic pathway for acute PE with the attempt to Recent studies suggest that data acquisition in the tomo-
increase specificity for the identification of perfusion defects graphic mode as stand-alone SPECT or (preferably) hybrid
due to PE [48]. The rationale underlying use of the ventila- SPECT/CT may reduce the frequency of nondiagnostic
tion/perfusion (V/Q) scan in patients with suspected PE is scans. In this regard, SPECT/CT is emerging as an imaging
that lung areas with reduced perfusion and ventilation (V/Q procedure yielding fewer indeterminate results and better
matching pattern) indicate the presence of some preexisting diagnostic accuracy than planar imaging (see Figs. 21.7 and
parenchymal condition affecting both perfusion and ventila- 21.8). The diagnostic performance of hybrid SPECT/CT
tion—an occurrence that makes PE unlikely. Whereas, the imaging is satisfactory both for combined V/Q imaging and
presence of underperfused but normally ventilated areas for perfusion imaging alone. In particular, Gutte et al.
(V/Q mismatch pattern) is consistent with acute PE. reported 93% sensitivity (with 51% specificity and only 17%
Performing only the perfusion lung scan is acceptable in of nondiagnostic scans) for perfusion SPECT/CT, sensitivity
patients with a normal chest X-ray and is the rule for the increasing only marginally (97%) with combined V/Q
Table 21.2 Interpretation of lung perfusion scintigraphy according to the PISA-PED criteria for diagnosing acute PE
Classification Scintigraphic pattern
Normal scan No perfusion defects of any kind
Near-normal Perfusion defects smaller ≤ in size and shape to extrapulmonary chest
scan X-ray abnormalities (enlargement of the heart or hilum or mediastinum or aorta, elevated diaphragm, costophrenic angle
effusion or blunting, pleural thickening, intrafissural effusion)
Abnormal Single or multiple non-wedged-shape perfusion defects, with or without matching pulmonary X-ray abnormalities;
scan without wedge-shaped areas of overperfusion are usually not seen
PE
Abnormal Single or multiple wedge-shaped (segmental or subsegmental) perfusion defects with or without matching pulmonary X-ray
scan with PE abnormalities; wedge-shaped areas of overperfusion usually coexist
Modified from: Miniati M, Pistolesi M, Marini C, Di Ricco G, Formichi B, Prediletto R, et al. Value of perfusion lung scan in the diagnosis of
pulmonary embolism: results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISA-PED). Am J Respir Crit Care
Med. 1996;154:1387–93
Fig. 21.5 Patient with pulmonary embolism arising about 10 days oblique and left posterior oblique in lower row): multiple wedge-shaped
after removal of infected prosthetic graft of the abdominal aorta, segmental and subsegmental perfusion defects associated with areas of
replaced with cryopreserved human allograft; the patient also has a forced perfusion. This pattern is consistent with pulmonary embolism
large aneurism of the aortic arch. (a) Standard chest X-rays in the two (modified from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di
orthogonal projections. (b) Lung perfusion scintigraphy obtained in six Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer-Verlag
views after the i.v. injection of 99mTc-MAA (anterior and posterior in Italy; 2010)
upper row, right lateral and left lateral in middle row, right posterior
SPECT/CT—although associated with much better specific- The above considerations explain why nuclear medicine
ity (88%) [49]. In a more recent study involving a larger centers where a hybrid SPECT/CT gamma camera is avail-
patients’ population, Mazurek et al. obtained 100% sensitiv- able have adopted (or are considering to adopt) SPECT/CT
ity and 83% specificity (without any nondiagnostic scan) imaging de novo rather than planar imaging when investigat-
with perfusion only SPECT/CT imaging [50]; hybrid ing patients with suspected PE [53–56]; according to local
SPECT/CT imaging had significantly better diagnostic accu- experience and confidence in diagnostic protocols, either
racy than both planar imaging and stand-alone SPECT imag- perfusion-only SPECT/CT or V/Q SPECT/CT imaging is
ing (P < 0.001 in either case). recommended.
The main advantage of combined SPECT/CT imaging is Finally, investigations have been carried out on the use of
that the CT component of the investigation (even in the PET/CT imaging for diagnosing acute PE, although no solid
modality of a low-dose, non-contrast-enhanced scan) is par- data have so far been acquired to justify use of this modality
ticularly useful to identify alternative causes for the symp- in the clinical routine. Two different rationales have been fol-
toms raising the clinical suspicion of PE and/or for the lowed in this approach; the first approach is based on the use
perfusion defects detected in the radionuclide component of of PET tracers for imaging both perfusion and ventilation,
the scan [51, 52]. such as [13N]NH3 for perfusion, associated with gaseous
Fig. 21.6 Patient with

massive pulmonary embolism. a
(a) Standard chest X-ray in
the posteroanterior projection
(the patient was bedridden
and could not be correctly
positioned for the lateral
projection): enlargement of
the right cardiac chambers
and mild effusion in the left
costophrenic space. (b) Lung
perfusion scintigraphy
obtained in six views after the
i.v. injection of 99mTc-MAA
(anterior and posterior in
upper row, right lateral and
left lateral in middle row,
right posterior oblique and
left posterior oblique in lower
row): multiple wedge-shaped
segmental and subsegmental
perfusion defects (more
severe in left lung) associated b
with areas of forced perfusion
(modified from: Volterrani D,
Springer-Verlag Italy; 2010)
Fig. 21.7 Ventilation/perfusion (V/P) SPECT/CT imaging in a sections. No abnormalities justifying the perfusion defect can be
patient with acute pulmonary embolism: ventilation images in left detected on the normal ventilation scan and on the low-dose CT com-
column, perfusion images in middle column, CT images in right ponent of the scan (reproduced with permission from: Mortensen J,
column. A wedge- shaped pleural-based large mismatched perfu- Gutte H. SPECT/CT and pulmonary embolism. Eur J Nucl Med Mol
sion defect is seen anteriorly in the left lung (yellow arrows) on the Imaging. 2014;41(Suppl 1):S81–90)
axial (upper row), coronal (middle row), and sagittal (lower row)
Fig. 21.8 Selected axial section from a ventilation/perfusion (V/P) These findings are compatible with inflammation rather than with acute
SPECT/CT scan: ventilation image in left column, perfusion image in pulmonary embolism, regardless of the mismatch ventilation/perfusion
middle column, CT image in right column. Very small mismatched per- pattern (reproduced with permission from: Mortensen J, Gutte
fusion defects (yellow arrows) are seen peripherally in both lungs, cor- H. SPECT/CT and pulmonary embolism. Eur J Nucl Med Mol Imaging.
responding to widespread pleural-based interstitial ground-glass 2014;41(Suppl 1):S81–90)
abnormalities detected on the CT component of the scan (blue arrows).
[13N]N2 for ventilation [57], or 68Ga-MAA for perfusion, Furthermore, recurring PE still causes >30% of deaths within
associated with 68Ga-aerosol for ventilation [58]. The other 2 years from an acute episode [61]. Adequate short-term and
approach is based on the use of an agent labeled with 18F for long-term follow-up is therefore important after the diagnosis
specific binding to the activated platelets that constitute a and primary treatment of acute PE. Lung perfusion scintigra-
fresh thrombus [59]. phy is the optimal imaging modality for monitoring recovery
of pulmonary perfusion in the short term after acute PE (to
monitor the efficacy of therapy) as well as for long-term fol-
Key Learning Points low-up of patients. In fact, perfusion scintigraphy is much
• The diagnosis of pulmonary embolism based on more feasible, less expensive, and burdened with fewer bio-
clinical data only is considered unreliable. logical risks and lower radiation dosimetry to patients than
• Correct diagnosis of PE is mandatory prior to start- CT contrast angiography.
ing any anticoagulation therapy. Irrespective of the imaging modality that has ascertained
• Lung perfusion scintigraphy with 99mTc-MAA is the the diagnosis of acute PE (i.e., either lung scintigraphy or
cornerstone of radionuclide imaging in patients CT angiography), a baseline lung perfusion scan should be
with suspected PE. obtained in all patients immediately (or within 24 h of the
• A normal perfusion scan virtually excludes the diagnosis), to provide a reference imaging set for subse-
diagnosis of acute PE. quent follow-up scans to be performed for assessing recov-
• The extremely high negative predictive value of ery of pulmonary perfusion [62]. Timing of follow-up
lung perfusion scintigraphy for acute PE is associ- imaging varies among different clinical practices; nonethe-
ated with low specificity. less, it appears the optimal monitoring for the risk of devel-
• The addition of ventilation scintigraphy as a com- oping chronic thromboembolic pulmonary hypertension is
bined ventilation/perfusion (V/P) imaging proce- based on perfusion lung scans performed soon after acute
dure aims at increasing specificity of radionuclide PTE (i.e., at 1 and 4 weeks) (see Fig. 21.9) and then at 3, 6,
imaging for the diagnosis of acute PE. and 12 months [63, 64].
It is also important to emphasize the role of lung perfu-
sion as an integral component of diagnostic screening in all
patients with pulmonary hypertension; in fact, undiagnosed
chronic thromboembolic disease is frequently the condition
21.12 F
ollow-Up of Patients After Acute underlying this disease, even in patients without a prior clini-
Pulmonary Embolism cally obvious episode of acute PE [47, 48, 65]. The superior
performance of V/Q scintigraphy with respect to contrast CT
Untreated or poorly treated acute PE may lead to a condition pulmonary angiography in diagnosing chronic thromboem-
known as chronic thromboembolic pulmonary hypertension, bolic disease as a treatable cause of pulmonary hypertension
which is severely debilitating and potentially fatal [60]. has been clearly demonstrated, with an extremely high nega-
Acute PE at diagnosis Week 1 after therapy Week 4 after therapy
Fig. 21.9 Example of monitoring of recovery of perfusion in the short- ciated with areas of forced perfusion corresponding to shifting of blood
term follow-up of a patient with acute pulmonary embolism; for clarity, flow to non-embolized vessels. The lung perfusion scan obtained
the figure reports only the orthogonal views (anterior, posterior, right 1 week after starting heparin therapy (middle panel) shows already
lateral, left lateral) obtained after the i.v. injection of 99mTc-MAA. The clear recovery of perfusion, which is much more obvious in the scan
baseline lung perfusion scintigraphy (left panel) shows multiple bilat- obtained at 4 weeks—showing almost complete recovery of perfusion
eral wedge-shaped perfusion defects with segmental distribution asso-
tive predictive value (98.5–99.9% versus 79.7% for CT pul- more difficult to perform because of poor patient’s compli-
monary angiography) [66]. ance. Intrapulmonary distribution of ventilation is driven by
two different mechanisms, i.e., the pressure-driven move-
ment of air (convective ventilation) and movement of the
Key Learning Points
gas molecules in less concentrated areas (diffusive ventila-
• Adequate short-term and long-term follow-up is tion). Different radionuclide agents have different distribu-
important after the diagnosis and primary treatment tion patterns and visualize different components of
of acute PE, in order to recognize chronic thrombo- ventilation. Convective ventilation is prevalent from the
embolic pulmonary hypertension, a severely debili- upper airways to the bronchioles, whereas all remaining
tating and potentially fatal condition. lung volume is dominated by diffusive ventilation. In
• Lung perfusion scintigraphy is the optimal imaging obstructive lung disease, both components of ventilation,
modality for monitoring recovery of pulmonary convective and diffusive, are impaired. These changes can
perfusion in the short term after acute PE (to moni- be visualized by ventilation scintigraphy, which frequently
tor the efficacy of therapy) as well as for long-term shows pathognomonic scintigraphic findings.
follow-up of patients. Reduced diameter of the first bronchial generation, as it
• Lung perfusion scintigraphy is important also for occurs in case of bronchial obstruction due to asthma,
diagnostic screening in all patients with pulmonary causes deposition of the ventilation imaging agent on the
hypertension. wall of the first bronchial generation, thus resulting in visu-
alization of a central deposition (CD) pattern of the tracer.
Instead, destruction of lung parenchyma and reduced diam-
eter of the peripheral bronchial generation, as it occurs in
chronic bronchitis and emphysema, cause a reduction in the
21.13 Obstructive Lung Disease peripheral deposition of the tracer, thus resulting in an inho-
mogeneous deposition (ID) pattern or in a spotty distribu-
Asthma and chronic obstructive pulmonary disease (COPD) tion (SD) pattern of the ventilation imaging agent (see
constitute a heterogeneous group of obstructive lung diseases. Fig. 21.11).
In the diagnostic evaluation of patients affected by obstruc-
tive lung disease, there is not a clear role for lung perfusion
scintigraphy. The perfusion defects that can be observed in
these conditions do not have a “segmental” distribution, Key Learning Points
although it is possible to recognize a somewhat “inhomoge- • There is not a clear role for lung perfusion scintig-
neous distribution” of 99mTc-MAA, as well as a non-physio- raphy in the evaluation of COPD patients.
logic distribution of perfusion inside the lungs with loss of the • Ventilation lung scintigraphy can demonstrate spe-
physiologic distribution gradient (see Fig. 21.10). cific patterns of tracer distribution in different
Ventilation scintigraphy in this clinical setting is more conditions.
informative than the perfusion scan, although it is frequently
Fig. 21.10 Lung perfusion

scintigraphy obtained after the
i.v. injection of 99mTc-MAA in
a patient with chronic
obstructive pulmonary disease
(anterior and posterior views
in upper row, right lateral and
left lateral in middle row,
right posterior oblique and
left posterior oblique in lower
row), showing multiple
inhomogeneities of perfusion,
which is reduced
predominantly in the mantle,
subpleural regions of the
lungs and does not have a
segmental type of distribution
21.14 Quantification of Differential niques have been proposed by some investigators to define
Pulmonary Function Before lung regions that correspond more closely to pulmonary
Pulmonary Surgery anatomy. Radioactive counts within each region relative to
overall counts within the entire lungs are employed to calcu-
One important indication for lung scintigraphy (either the late the fraction of perfusion (or ventilation, if a lung ventila-
perfusion scan or the ventilation and perfusion scans) is for tion scintigraphy is performed) that will be lost after resection
regional quantitation of respiratory function in patients with of that region.
borderline pulmonary function tests scheduled for resection
surgery because of, e.g., cancer or other condition requiring
surgery [67]. The clinical issue that the clinicians must 21.15 Interstitial Lung Disease
address before performing surgery concerns the residual
respiratory function after surgery—which may consist of Interstitial lung disease is a general category that includes
pneumonectomy or lobectomy. For this type of evaluation, many different pulmonary conditions that share involvement
lung scintigraphy is most frequently performed by acquiring of the interstitium, all of them causing thickening of the inter-
the anterior and the posterior views after the i.v. injection of stitium that can be due to inflammation, scarring, or chronic
99m
Tc-MAA; the geometric mean of the two acquisitions is excess fluid (edema). Some forms of interstitial lung disease
then calculated. For regional quantitation, rectangular are short-lived; others are chronic and irreversible. A typical
regions of interest are drawn, dividing each lung into three interstitial lung disease is caused by sarcoidosis, a multisys-
regions: top, middle, and bottom regions; alternative tech- temic disease characterized by hyperactivity of cellular immu-
Fig. 21.11 Different patterns

of ventilation visualized after
inhalation of an aerosol
prepared with 99mTc-
nanocolloidal albumin in four
different patients with chronic
obstructive pulmonary
disease. The pattern of central
deposition is observed more
frequently in patients with
asthma, whereas spotty
deposition is more frequently
observed in patients with
pulmonary emphysema. The
pattern of inhomogeneous
deposition and mixed
deposition generally
corresponds to intermediate In homogeneous Central
severity degree of chronic deposition deposition
asthma and/or emphysema
responding to therapy
Spotty Mixed
deposition deposition
nity with formation of noncaseating granulomas in various

organ systems. The majority of deaths from sarcoidosis result
from respiratory failure, and no universal definition exists for
what constitutes “active” pulmonary disease. It is therefore
important to accurately assess pulmonary involvement.
Imaging methods play an important role in the diagnosis and
treatment strategy plan in patients with sarcoidosis at both pri-
mary staging and follow-up. Chest X-rays and CT may sup-
port the diagnosis of sarcoidosis. Bilateral pulmonary hilar
and mediastinal lymphadenopathy is the most common radio-
logical finding, often associated with pulmonary infiltrates.
In the past 67Ga-citrate scintigraphy has been largely used
primarily for evaluating disease activity but also as an aid for
the differential diagnosis of sarcoidosis, since it is more sen-
sitive for early detection of the disease than chest X-rays and
CT [68]. The classic findings on a 67Ga-citrate scan in patients Fig. 21.12 Planar anterior view of the chest acquired in a patient sar-
coidosis after administration of 67Ga-citrate. The “lambda” sign is due to
with sarcoidosis include increased tracer uptake in mediasti- bilateral symmetrical tracer uptake in hilar lymph nodes and in the azy-
nal lymph nodes (and thymus), showing up with a pattern gos lymph nodes (or lymph nodes in the Baréty space) (modified from:
known as the “lambda” sign (see Fig. 21.12). As to extra- Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
thoracic sarcoidosis, the “panda” sign refers to increased Nucleare – Tecniche e Applicazioni. Milan: Springer-Verlag Italy; 2010)
a b c
Fig. 21.13 Assessment of response to steroid treatment with 67Ga-citrate (b) Planar whole-body scan obtained in the same patient after a course
scintigraphy in a patient with lung-predominant sarcoidosis. (a) Planar of treatment with steroid drugs, showing an almost normal pattern of
whole-body scan obtained at the time of diagnosis of the disease, show- radioactivity distribution. (c) Planar X-rays of the chest (posteroanterior
ing markedly increased tracer uptake in the mediastinal lymph nodes projection) obtained at the time of diagnosis, showing marked enlarge-
(with typical “lambda” sign) associated with extra-thoracic manifesta- ment of hilar lymph nodes. (d) Planar X-rays of the chest (posteroante-
tions visualized as the “panda” sign of the head, due to involvement of rior projection) obtained after treatment with steroid drugs, showing
the lacrimal and parotid glands (more intense uptake in the left parotid some persistence of enlarged hilar lymph nodes (particularly on the right
gland is due to the presence of markedly increased uptake in a lymph side) (reproduced with permission from: Volterrani D, Erba PA, Mariani
node within this gland). Physiologic tracer uptake in the liver and radio- G, Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
activity excretion in the gastrointestinal tract (colon) can also be noticed. Milan: Springer-Verlag Italy; 2010)
tracer uptake in the bilateral lacrimal and parotid glands. In sarcoidosis; it can also monitor activity of the disease in
the chest, the lambda sign has been so denominated because response to steroid therapy (Fig. 21.13). 67Ga-citrate scintig-
of a resemblance with the Greek letter λ due to increased raphy is currently not routinely used in patients with sarcoid-
67
Ga uptake in the right paratracheal and bilateral hilar ade- osis because of the following limitations: (1) it is a
nopathy. Sensitivity of 67Ga-citrate scintigraphy for sarcoid- time-consuming method, since images are acquired at least
osis ranges from 50% to 90%, but specificity is low. A 48–72 h after tracer injection; (2) there is significant interob-
negative 67Ga-citrate scan in conjunction with a negative server variability in image interpretation; (3) the overall sen-
serum angiotensin-converting enzyme can rule out sarcoid- sitivity and specificity values are highly variable among
osis. Because of its sensitivity, the 67Ga-citrate scan can be different studies; and (4) administration of 67Ga entails a
used to detect active sites of the disease for biopsy to confirm rather high radiation exposure to patients.
Key Learning Points

• The traditional approach to radionuclide imaging in
patients with sarcoidosis is based on 67Ga-citrate
scintigraphy.
• 67Ga-citrate scintigraphy does not have optimal
diagnostic performance and entails a rather high
radiation burden to patients.
• [18F]FDG PET has better diagnostic accuracy than
67
Ga-citrate scintigraphy in patients with
sarcoidosis,
• [18F]FDG PET is useful for the assessment of
inflammatory activity in sarcoidosis.
• [18F]FDG PET is useful for staging and for identify-
ing occult sites of disease.
References
1. Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pul-
monary embolism. A controlled trial. Lancet. 1960;1:1309–12.
2. Coulden R. State-of-the-art imaging techniques in chronic
thromboembolic pulmonary hypertension. Proc Am Thorac Soc.
2006;3:577–83.
3. White RH. The epidemiology of venous thromboembolism.
Circulation. 2003;107:I4–8.
4. Heit JA. The epidemiology of venous thromboembolism in the
community: implications for prevention and management. J
Thromb Thrombolysis. 2006;21:23–9.
5. Nordstrom M, Lindblad B. Autopsy-verified venous thromboem-
Fig. 21.14 Comparison of planar whole-body scintigraphy with bolism within a defined urban population – the city of Malmö,
67
Ga-citrate (left panel) and PET with [18F]FDG (coronal section in Sweden. APMIS. 1998;106:378–84.
right panel) in a patient with sarcoidosis; image quality is obviously 6. Tapson VF. Acute pulmonary embolism. N Engl J Med.
much higher in the [18F]FDG PET scan, in terms of both spatial resolu- 2008;358:1037–52.
tion and lesion-to-background ratio (reproduced with permission from: 7. Lehmann R, Suess C, Leus M, Luxembourg B, Miesbach W,
Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina Lindhoff-Last E, et al. Incidence, clinical characteristics, and long-
Nucleare – Tecniche e Applicazioni. Milan: Springer-Verlag Italy; term prognosis of travel-associated pulmonary embolism. Eur
2010) Heart J. 2009;30:233–41.
8. Bajc M, Neilly JB, Miniati M, Schuemichen C, Meignan M,
Jonson B, EANM Committee. EANM guidelines for ventilation/
perfusion scintigraphy. Eur J Nucl Med Mol Imaging. 2009;36:
1356–70.
The current approach to radionuclide imaging in patients 9. Parker JA, Coleman RE, Grady E, Royal HD, Siegel BA, Stabin
with sarcoidosis is based on PET/CT with the glucose analog MG, et al. SNM practice guideline for lung scintigraphy 4.0. J Nucl
[18F]FDG, characterized by much higher resolution and Med Technol. 2012;40:57–65.
10. Reilly JJ. Evidence-based preoperative evaluation of candidates for
overall image quality than 67Ga-citrate scintigraphy thoracotomy. Chest. 1999;116(Suppl):474S–6S.
(Fig. 21.14). In fact, high glucose metabolism has been 11. Beckles MA, Spiro S, Colis G, Rudd R. The physiologic evalua-
shown to be useful for the assessment of inflammatory activ- tion of patients with lung cancer being considered for resectional
ity in sarcoidosis [69]. In particular, [18F]FDG PET has better surgery. Chest. 2003;123(Suppl):105S–14S.
12. Tunariu N, Gibbs JR, Win Z. Ventilation-perfusion scintigraphy
diagnostic accuracy than 67Ga-scintigraphy in patients with is more sensitive than multidetector CTPA in detecting chronic
sarcoidosis, in addition to several practical advantages (lower thromboembolic pulmonary disease as a treatable cause of pulmo-
radiation exposure, shorter time between injection and imag- nary hypertension. J Nucl Med. 2007;48:680–4.
ing). Current evidence demonstrates that [18F]FDG PET is a 13. Humplik BI, Sandrock D, Aurisch R, Richter WS, Ewert R, Munz
DL. Scintigraphic results in patients with lung transplants: a pro-
very useful molecular imaging method for assessing disease spective comparative study. Nuklearmedizin. 2005;44:62–8.
activity, for staging and identifying occult sites of disease, 14. Gates GF, Orme HW, Dore EK. Measurement of cardiac shunt-
and for monitoring treatment response in patients with ing with technetium-labeled albumin aggregates. J Nucl Med.
sarcoidosis. 1971;12:746–9.
15.
Jacobson AF, Herzog SA. Open bronchial stump post- 34. Wells PS, Anderson DR, Rodgers M, Ginsberg JS, Kearon C, Gent
pneumonectomy: findings on xenon-133 ventilation imaging. J M, et al. Derivation of a simple clinical model to categorize patients’
Nucl Med. 1993;34:462–4. probability of pulmonary embolism: increasing the model’s utility
16. Itti E, Fauroux B, Pigeot J, Isabey D, Clement A, Evangelista E, with simpliRED D-dimer. Thromb Haemost. 2000;83:416–20.
et al. Quantitative lung perfusion scan as a predictor of aerosol 35. Wicki J, Perneger TV, Junod AF, Bounameaux H, Perrier

distribution heterogeneity and disease severity in children with cys- A. Assessing clinical probability of pulmonary embolism in the
tic fibrosis. Nucl Med Commun. 2004;25:563–9. emergency ward. Arch Intern Med. 2001;161:92–7.
17. Levy SE, Simmons DH. Redistribution of alveolar ventilation fol- 36. Klok FA, Kruisman E, Spaan J, Nijkeuter M, Righini M, Aujesky
lowing pulmonary thromboembolism in the dog. J Appl Physiol. D, et al. Comparison of the revised Geneva score with the Wells
1974;36:60–8. rule for assessing clinical probability of pulmonary embolism. J
18. Santolicandro A, Prediletto R, Fornai E, Formichi B, Begliuomini Thromb Haemost. 2008;6:40–4.
E, Giannella-Neto A, et al. Mechanisms of hypoxemia and hypo- 37. Ceriani E, Combescure C, Le Gal G, Nendaz M, Perneger T,

capnia in pulmonary embolism. Am J Respir Crit Care Med. Bounameaux H, et al. Clinical prediction rules for pulmonary
1995;152:336–47. embolism: a systematic review and meta-analysis. J Thromb
19. Swenson EW, Finley TN, Guzman SV. Unilateral hypoventilation Haemost. 2010;8:957–70.
in man during temporary occlusion of one pulmonary artery. J Clin 38. Miniati M, Pistolesi M. Assessing the clinical probability of pulmo-
Invest. 1961;40:828–35. nary embolism. Q J Nucl Med. 2001;45:287–93.
20. Allgood RJ, Wolfe WG, Ebert PA, Sabiston DC J. Effects of carbon 39. Miniati M, Bottai M, Monti S, Salvadori M, Serasini L, Passera
dioxide on bronchoconstriction after pulmonary artery occlusion. M. Simple and accurate prediction of the clinical probability of pul-
Am J Phys. 1968;214:772–5. monary embolism. Am J Respir Crit Care Med. 2008;178:290–4.
21. D’Angelo E, Salvo Calderini T, Tavola M. The effects of CO2 40. Gu Y-H, Zhao Z. Role of three commonly used scoring systems in
on respiratory mechanics in anesthetized paralyzed humans. prediction of pulmonary thromboembolism in Xining area. Eur Rev
Anesthesiology. 2001;94:604–10. Med Pharmacol Sci. 2014;18:3517–20.
22. Finley TN, Tooley WH, Swenson EW, Gardner RE, Clements
41. Russell LA, Sigmund AE, Szymonifka J, Jawetz ST, Grond SE, Dey
JA. Pulmonary surface tension in experimental atelectasis. Am Rev SA, et al. Does my patient have a pulmonary embolism? The Wells
Respir Dis. 1964;89:372–7. vs. PISA 2 rule in orthopedic patients. J Thromb Thrombolysis.
23. Chernick V, Hodson WA, Greenfield LJ. Effect of chronic pulmo- 2018;45:417–22.
nary artery ligation on pulmonary mechanics and surfactant. J Appl 42. Wagner HN Jr, Sabiston DC Jr, McAfee JG, Tow D, Stern

Physiol. 1966;21:1315–20. HS. Diagnosis of massive pulmonary embolism in man by radio-
24. Miniati M, Prediletto R, Formichi B, Marini C, Di Ricco G, Tonelli isotope scanning. N Engl J Med. 1964;271:377–84.
L, et al. Accuracy of clinical assessment in the diagnosis of pulmo- 43. Taplin GV, Johnson DE, Dore EK, Kaplan HS. Lung photoscan
nary embolism. Am J Respir Crit Care Med. 1999;159:864–71. with macroaggregates of human serum radioalbumin. Experimental
25. Gottschalk A, Juni JE, Sostman HD, Coleman RE, Thrall J,
basis and initial clinical trials. Health Phys. 1964;10:1219–27.
McKusick KA, et al. Ventilation-perfusion scintigraphy in the 44. Paraskos VA, Adelstein SJ, Smith RE, Rickman FD, Grossman W,
PIOPED study. Part I. Data collection and tabulation. J Nucl Med. Dexter L, et al. Late prognosis of acute pulmonary embolism. N
1993;34:1109–18. Engl J Med. 1973;289:55–8.
26. Gottschalk A, Sostman HD, Coleman RE, Juni JE, Thrall J,
45. Sostman HD, Neumann RD. The respiratory system. In: Harbert
McKusick KA, et al. Ventilation-perfusion scintigraphy in the JC, Eckelman WC, Neumann RD, editors. Nuclear medicine –
PIOPED study. Part II. Evaluation of the scintigraphic criteria and diagnosis and therapy. New York, NY: Thieme Medical Publishers;
interpretation. J Nucl Med. 1993;34:1119–26. 1996. p. 553–84.
27. Stein PD, Woodard PK, Weg JG, Wakefield TW, Tapson VF,
46. Lowe VJ, Sostman HD. Pulmonary embolism. In: Ell PJ, Gambhir
Sostman HD, et al. Diagnostic pathways in acute pulmonary embo- SS, editors. Nuclear medicine in clinical diagnosis and treatment.
lism: recommendations of the PIOPED II investigators. Am J Med. 3rd ed. New York, NY: Churchill Livingstone; 2004. p. 29–46.
2006;119:1048–55. 47. Reid JH, Coche EE, Inoue T, Kim EE, Dondi M, Watanabe N,
28. Miniati M, Pistolesi M, Marini C, Di Ricco G, Formichi B,
et al. Is the lung scan alive and well? Facts and controversies in
Prediletto R, et al. Value of perfusion lung scan in the diagnosis defining the role of lung scintigraphy for the diagnosis of pulmo-
of pulmonary embolism: results of the Prospective Investigative nary embolism in the era of MDCT. Eur J Nucl Med Mol Imaging.
Study of Acute Pulmonary Embolism Diagnosis (PISA-PED). Am 2009;36:505–21.
J Respir Crit Care Med. 1996;154:1387–93. 48. McNeil BJ, Holman BL, Adelstein SJ. The scintigraphic definition
29. Miniati M, Sostman HD, Gottschalk A, Monti S, Pistolesi
of pulmonary embolism. JAMA. 1974;227:753–6.
M. Perfusion lung scintigraphy for the diagnosis of pulmonary 49. Gutte H, Mortensen J, Jensen CV, Johnbeck CB, von der Recke P,
embolism: a reappraisal and review of the prospective investiga- Petersen CL, et al. Detection of pulmonary embolism with com-
tive study of pulmonary embolism diagnosis methods. Semin Nucl bined ventilation-perfusion SPECT and low-dose CT: head-to-
Med. 2008;38:450–61. head comparison with multidetector CT angiography. J Nucl Med.
30. Sostman HD, Gottschalk A. Prospective validation of the stripe 2009;50:1987–92.
sign in ventilation-perfusion scintigraphy. Radiology. 1992;184: 50. Mazurek A, Dziuk M, Witkowska-Patena E, Piszczek S, Gizewska
455–9. A. The utility of hybrid SPECT/CT lung perfusion scintigraphy in
31. Worsley DF, Alavi A, Aronchick JM, Chen JT, Greenspan RH, pulmonary embolism diagnosis. Respiration. 2015;90:393–401.
Ravin CE. Chest radiographic findings in patients with acute pulmo- 51. Simanek M, Koranda P. The benefit of personalized hybrid

nary embolism: observations from the PIOPED Study. Radiology. SPECT/CT pulmonary imaging. Am J Nucl Med Mol Imaging.
1993;189:133–6. 2016;6:215–22.
32. Eisner MD. Before diagnostic testing for pulmonary embolism: esti- 52. Hess S, Frary EC, Gerke O, Madsen PH. State-of-the-art imaging in
mating the prior probability of disease. Am J Med. 2003;114:232–4. pulmonary embolism: ventilation/perfusion single-photon emission
33. Kelly J, Hunt BJ. The utility of pretest probability assessment in computed tomography versus computed tomography angiogra-
patients with clinically suspected venous thromboembolism. J phy – controversies, results, and recommendations from a system-
Thromb Haemost. 2003;1:1888–96. atic review. Semin Thromb Hemost. 2016;42:833–45.
53. Mortensen J, Gutte H. SPECT/CT and pulmonary embolism. Eur J follow-up of patients after acute pulmonary embolism. Medicine
Nucl Med Mol Imaging. 2014;41(Suppl 1):S81–90. (Baltimore). 2006;85:253–62.
54. Lu Y, Lorenzoni A, Fox JJ, Rademaker J, Vander Els N, Grewal RK, 62. Donnamaria V, Palla A, Petruzzelli S, Carrozzi L, Pugliesi O,

et al. Noncontrast perfusion single-photon emission CT/CT scan- Giuntini C. Early and late follow-up of pulmonary embolism.
ning: a new test for the expedited, high-accuracy diagnosis of acute Respiration. 1993;60:15–20.
pulmonary embolism. Chest. 2014;145:1079–88. 63. Palla A, Pazzagli M, Manganelli D, De Nitto P, Marini C,

55. Bhatia KD, Ambati C, Dhaliwal R, Paschkewitz R, Hsu E, Ho Rossi G, et al. Resolution of pulmonary embolism: effect of
B, et al. SPECT-CT/VQ versus CTPA for diagnosing pulmonary therapy and putative age of emboli. Respiration. 1997;64:
embolus and other lung pathology: pre-existing lung disease 50–3.
should not be a contraindication. J Med Imaging Radiat Oncol. 64. Wartski M, Collignon MA. Incomplete recovery of lung perfusion
2016;60:492–7. after 3 months in patients with acute pulmonary embolism treated
56. Le Roux PY, Robin P, Delluc A, Abgral R, Palard X, Tissot V, et al. with antithrombotic agents. THESEE Study Group. Tinzaparin ou
Additional value of combining low-dose computed tomography to Heparin Standard: Evaluation dans l’Embolie Pulmonaire Study. J
V/Q SPECT on a hybrid SPECT-CT camera for pulmonary embo- Nucl Med. 2000;41:1043–8.
lism diagnosis. Nucl Med Commun. 2015;36:922–30. 65. Lang IM. Chronic thromboembolic pulmonary hypertension: not so
57. Musch G, Venegas JG. Positron emission tomography imaging of rare after all. N Engl J Med. 2004;350:2236–8.
regional pulmonary perfusion and ventilation. Proc Am Thorac 66. Tunariu N, Gibbs SJR, Win Z, Gin-Sing W, Graham A, Gishen
Soc. 2005;2:522–7. P, et al. Ventilation-perfusion scintigraphy is more sensitive than
58. Bailey DL, Eslick EM, Schembri GP, Roach PJ. 68Ga PET ventila- multidetector CTPA in detecting chronic thromboembolic pulmo-
tion and perfusion lung imaging – current status and future chal- nary disease as a treatable cause of pulmonary hypertension. J Nucl
lenges. Semin Nucl Med. 2016;46:428–35. Med. 2007;48:680–4.
59. Lohrke J, Siebeneicher H, Berger M, Reinhardt M, Berndt M, 67. Win T, Tasker A, Groves A, White C, Ritchie AJ, Wells FC, et al.
Mueller A, et al. 18F-GP1, a novel PET tracer designed for high- Ventilation-perfusion scintigraphy to predict postoperative pulmo-
sensitivity, low-background detection of thrombi. J Nucl Med. nary function in lung cancer patients undergoing pneumonectomy.
2017;58:1094–9. AJR Am J Roentgenol. 2006;187:1260–5.
60. Pengo V, Lensing AW, Prins MH, Marchiori A, Davidson BL, 68. Silva M, Nunes H, Valeyre D, Sverzellati N. Imaging of sarcoid-
Tiozzo F, et al. Thromboembolic pulmonary hypertension study osis. Clin Rev Allergy Immunol. 2015;49:45–53.
group. incidence of chronic pulmonary hypertension after pulmo- 69. Mostard RL, Verschakelen JA, van Kroonenburgh MJ, Nelemans
nary embolism. N Engl J Med. 2004;350:2257–64. PJ, Wijnen PA, Vöö S, et al. Severity of pulmonary involve-
61. Miniati M, Monti S, Bottai M, Scoscia E, Bauleo C, Tonelli L, et al. ment and 18F-FDG PET activity in sarcoidosis. Respir Med.
Survival and restoration of pulmonary perfusion in a long-term 2013;107:439–47.
Hybrid Imaging for Tumours
of the Chest 22
Roberto C. Delgado Bolton
and Adriana K. Calapaquí Terán
Contents
22.1 Lung Cancer 524
22.1.1 Epidemiology of Lung Cancer 524
22.1.2 Classification of Lung Cancer 524
22.1.3 Diagnosis 525
22.1.4 Recurrent Lung Cancer 527
22.1.5 Appropriate Use Criteria 527
22.2 [18F]FDG PET/CT Procedure, Data Processing, Interpretation, and Reporting 527
22.3 Initial Staging of Lung Cancer 527
22.4 Restaging for Detection of Local Recurrence 530
22.4.1 Restaging After Initial Treatment (Surgery, Chemoradiotherapy or Radiotherapy) 530
22.4.2 Restaging After Surgery 531
22.4.3 Restaging After Stereotactic Body Radiation Therapy 531
22.4.4 Restaging After Radiofrequency Ablation or Microwave Ablation 532
22.4.5 Cost-Effectiveness 532
22.5 Restaging for Detection of Metastases 532
22.6 Treatment Response Evaluation 533
22.7 Mesothelioma 535
22.8 Other Thoracic Tumours 535
22.8.1 Thymoma 535
22.8.2 Teratomas 539
References 539
Learning Objectives
• Present the main features on epidemiology, presentation,
and clinical evolution patterns for tumours of the chest,
including lung cancer, malignant pleural mesothelioma,
and mediastinal tumours.
• Briefly discuss, for each type of tumour of the chest, the role
R. C. Delgado Bolton (*) of non-radionuclide imaging techniques (plain X-ray, CT,
Department of Diagnostic Imaging and Nuclear Medicine, CECT, MRI) in the initial diagnostic approach, as well as for
University Hospital San Pedro and Center for Biomedical Research
of La Rioja (CIBIR), Logroño, La Rioja, Spain staging, restaging, and assessment of response to therapy.
A. K. Calapaquí Terán
• Discuss the residual role of bone scintigraphy with 99mTc-
Department of Pathology, University Hospital Marqués de bisphosphonates or with [18F]fluoride PET/CT in patients
Valdecilla (HUMV), Santander, Cantabria, Spain with lung cancer.

524 R. C. Delgado Bolton and A. K. Calapaquí Terán
• Discuss the role of perfusion scintigraphy as part of 22.1.2 Classification of Lung Cancer
preoperative evaluation for patients with lung
cancer. Non-small cell lung cancer (NSCLC) represents 85–90% of
• Describe the methodology of performing [18F]FDG PET/ all LCs [2]. It includes three main histologic types: squa-
CT (with either low-dose CT and/or CECT) in patients mous cell carcinoma, adenocarcinoma, and large-cell carci-
with tumours of the chest. noma. The first two types account for around 80% of all LCs
• Indicate the methodology for interpreting and reporting worldwide. Squamous cell carcinomas are predominantly
the results of [18F]FDG PET/CT in patients with tumours associated with smoking and usually present as large tumours
of the chest. in the centre of the lung [3, 4]. In contrast, adenocarcinomas
• Discuss the issues regarding definition of appropriate use are frequently located in the periphery of the lung and are
criteria for patients with tumours of the chest. divided into acinar, papillary, solid adenocarcinoma with
• Discuss the roles and clinical impact of [18F]FDG PET/ mucin production, adenocarcinoma in situ (AIS) with three
CT during the clinical course of disease (for diagnosis, subtypes (non-mucinous, rarely mucinous, or mixed), mini-
staging, restaging, assessment of response to therapy) in mally invasive adenocarcinoma (MIA), and two invasive
patients with tumours of the chest including lung cancer, adenocarcinomas (lepidic-predominant adenocarcinoma and
malignant pleural mesothelioma, and mediastinal invasive mucinous adenocarcinoma). It is worth emphasising
tumours. that in 2011 the International Association for the Study of
• Discuss and compare the performance of PERCIST ver- Lung Cancer (IASLC) and other societies jointly revised and
sus RECIST criteria for assessing response to treatment in updated the classification for adenocarcinoma of the lung.
patients with lung cancer. Since then, the denomination bronchiolo-alveolar carcinoma
• Discuss the potential of future perspectives in the clinical (or BAC) is no longer used, being replaced by the last four
use of PET/CT with non-[18F]FDG tracers for patients entities listed (AIS, MIA, and the two invasive adenocarcino-
with tumours of the chest. mas) [5]. However, mixed histologic patterns are present in
most cases. In non-smokers, the most frequent type of lung
cancer is adenocarcinoma [4].
22.1 Lung Cancer Small cell lung cancer (SCLC) accounts for approximately
15% of all lung cancers [1]. It usually affects men, with a
22.1.1 Epidemiology of Lung Cancer median age of 60 years, and 99% occur in smokers. SCLC is
an aggressive malignancy characterised by often presenting
Lung cancer (LC) is a fairly common malignancy. Smoking at a disseminated stage, with early mediastinal lymph node
is widely recognised as the leading cause of lung cancer. involvement. It initially responds to chemotherapy in about
Between 2009 and 2013, there were 57.3 new cases per 80% of cases, although most patients develop resistant dis-
100,000 men and women per year, and the number of ease and die due to tumour progression [6]. The cure rates are
deaths was 46.0 per 100,000 men and women per year. The as low as 15–25% for limited disease, with a life expectancy
lifetime risk of developing lung cancer is approximately of 1 year. The treatment is chemotherapy with or without
6.6% of men and women, based on the 2010–2012 data. In radiation, whereas preoperative chemotherapy and surgery is
2013, there were an estimated 415,707 people living with indicated in T1–2 N0–1 M0. SCLC is apparently derived
lung and bronchus cancer in the United States. Cancer from primitive cells of basal bronchial epithelium with partial
stage at diagnosis, which refers to extent of a cancer in the differentiation towards neuroendocrine cells. It is frequently
body, determines treatment options and has a strong influ- associated with paraneoplastic syndromes due to production
ence on the length of survival. In general, if the cancer is of ACTH (Cushing’s syndrome), ADH (hyponatremia), calci-
found only in the part of the body where it started, it is tonin (hypocalcaemia), gonadotropins (gynecomastia), para-
defined as localised (sometimes referred to as stage I). If it thyroid hormone (hyperparathyroidism), serotonin (carcinoid
has spread to a different part of the body, the stage is syndrome), encephalomyelitis, Lambert-Eaton syndrome,
regional or distant. The earlier lung and bronchus cancer is and sensory neuropathy. From a genetic point of view, SCLC
detected, the better chance a person has of surviving is especially characterised by manifold chromosomal dele-
5 years after being diagnosed. For lung and bronchus can- tions with losses of whole chromosomes or chromosome
cer, 15.7% are diagnosed at the local stage. The 5-year sur- arms, associated with the inactivation of numerous tumour
vival for localised lung and bronchus cancer is 55.2%. suppressor genes. On one hand, the extensive DNA losses
Rates for new lung and bronchus cancer cases have been may explain the marked sensitivity of SCLC to antineoplastic
falling on average 1.8% each year over the last 10 years. chemotherapy or radiotherapy, whereas on the other hand, its
Death rates have been falling on average 2.2% each year considerable chromosomal instability is correlated with the
over 2004–2013 [1]. development of resistance to therapy.
22 Hybrid Imaging for Tumours of the Chest 525
22.1.3 Diagnosis A subgroup of patients with lung cancer are initially diag-
nosed as an unknown primary tumour or carcinoma of
Diagnostic imaging for patients with lung cancer includes unknown primary, that is, the patient presents a pathologically
chest X-ray, CT, [18F]FDG PET, bone scintigraphy and [18F] confirmed metastatic lesion with no known primary tumour
Fluoride PET, lung perfusion scintigraphy, and, in neuroen- after the first examinations [15]. The most frequent underlying
docrine neoplasms (NEN), somatostatin receptor scintigra- primary tumours are located in the lung. [18F]FDG PET has an
phy and other PET tracers for NENs, such as important role, as it is able to detect the primary tumour in
68
Ga-DOTA-conjugated somatostatin receptor targeting pep- around 40% of patients following negative or inconclusive
tides (68Ga-DOTA-TOC,68Ga-DOTA-NOC, and 68Ga-DOTA- prior sectional imaging and endoscopic procedures, as demon-
TATE). CT has been the gold standard imaging technique for strated in a meta-analysis [15]. Furthermore, some patients
many decades, but its limitations are due to the fact that it present with suspected paraneoplastic syndromes without
relies exclusively on morphological aspects [4]. underlying known malignancy. A meta-analysis has presented
PET provides functional or metabolic information. PET/ evidence of the high diagnostic performance of [18F]FDG
CT systems combine the advantages of PET, a high contrast PET/CT in the detection of malignancy in patients with para-
resolution, and CT, a high spatial resolution. [18F]FDG is the neoplastic neurological syndrome [16].
most commonly used PET tracer. [18F]FDG PET/CT has Timely detection of bone metastases is crucial. Bone scin-
been extensively studied in lung cancer, and there is evidence tigraphy with 99mTc-bisphosphonates has had an important
showing its utility for characterising solitary pulmonary nod- role in detecting bone lesions until the widespread imple-
ules [7], staging [8], guiding therapy [9], monitoring treat- mentation of [18F]FDG PET/CT that allows an accurate M
ment response [10], and predicting outcome [11], as well as staging including bone metastases [17]. Nowadays, bone
having demonstrated its utility from an economic point of scintigraphy with 99mTc-bisphosphonates has a residual role.
view with a cost-effectiveness analysis [4, 12]. [18F]FDG PET/CT is superior to bone scintigraphy with
Regarding guiding lung biopsies, frequently the thoracic 99m
Tc-bisphosphonates in the detection of bone metastases,
lesions present a heterogeneous metabolism. In order to with a higher overall diagnostic efficacy. Based on the types
increase the probability of obtaining a definite histological of bone lesions, [18F]FDG PET/CT is much better than bone
diagnosis, [18F]FDG PET/CT can be used to guide the image- scintigraphy with 99mTc-bisphosphonates for the detection of
guided biopsy to the most active area within the lesion, lytic lesions that are more frequent in lung cancer. However,
avoiding areas of less hypermetabolism usually associated in osteoblastic lesions with a low metabolism (and low [18F]
with necrosis (which would be inconclusive in the pathology FDG uptake), bone scintigraphy can sometimes show better
study). Figure 22.1 presents a patient in whom the metabolic detection of these lesions. Therefore, the only remaining
information could be of use to guide the lung biopsy to the important indication in lung cancer patients in cases with a
most active area. negative [18F]FDG PET/CT for bone metastases are patients
For SCLC, accurate staging is important for management in whom osteoblastic lesions are suspected but [18F]FDG
decisions, since timing of diagnosis, stage and performance PET/CT was negative or non-conclusive. In a broader sense,
status have a prognostic value. With regard to SCLC with the common clinical indications of the bone scan could be
NEN differentiation, lung NENs with high expression of classified into three distinct scenarios: (a) when a specific
somatostatin receptors (SSTRs) may be visualised using bone disease is present or suspected, (b) to explore unex-
PET/CT with 68Ga-DOTA-conjugated somatostatin receptor plained symptoms, and (c) for the metabolic assessment
targeting peptides (TOC, NOC, and TATE) [13]. Somatostatin prior to the start of therapy with bone-seeking radiopharma-
receptor scintigraphy with single-photon-emitting radio- ceuticals [17]. The procedure of bone scintigraphy with
pharmaceuticals can also be used but has a lower diagnostic 99mTc-bisphosphonates is the standard whole-body scan for
efficacy than PET tracers. In the management of NENs, patients with cancer [17]. The accuracy of bone scintigraphy
PET/CT with 68Ga-DOTA-conjugated somatostatin receptor has improved since the introduction of multimodality
targeting peptides is used for diagnosis and staging, restag- SPECT/CT systems, as they allow improving lesion localisa-
ing, prognosis, and management decisions [13]. Regarding tion and differential diagnosis, based on the combination, in
prognosis, it is used to determine SSTR status, as patients one single study, of the functional information of the bone
with SSTR-positive tumours are more likely to respond to SPECT and the morphological image provided by CT.
targeted therapy with somatostatin analogues [13]. With There is a PET tracer available that presents the best effi-
regard to management decisions, it is used to select patients cacy for bone lesions, [18F]Fluoride. The diagnostic efficacy
with metastatic disease for SSTR radionuclide therapy with of [18F]Fluoride PET/CT for bone metastases detection is
177
Lu- or 90Y-DOTA-peptides [13]. Radionuclide therapy clearly higher than that of [18F]FDG PET/CT, especially for
enables the delivery of a high radiation dose to the target osteoblastic lesions, and of bone scintigraphy with 99mTc-
while minimising the toxicity to normal tissues [14]. bisphosphonates. However, it is not available in most centres
a c
b e
Fig. 22.1 Forty-two year-old man (history of tobacco smoker) with a CECT (e). Learning point: Frequently the thoracic lesions present
clinical presentation of neurologic compression of the upper right heterogeneous metabolism. In order to increase the probability of

extremity. The diagnosis is a Pancoast tumour in the right lung. The obtaining a definite histological diagnosis, [18F]FDG PET/CT can be
[18F]FDG PET/CECT is performed for initial staging, confirming a used to guide biopsy to the most active area within the lesion, avoiding
Pancoast tumour T4 with nonspecific mediastinal lymph nodes and no areas of less hypermetabolism usually associated with necrosis (which
distant metastases (M0). Axial fused PET/CECT (a), CT (b), and [18F] would be inconclusive in the pathology study)
FDG PET (c); [18F]FDG PET volumetric image or MIP (d) and coronal
due to slow changes in the regulations and legislation, or Use Criteria (AUC). These AUC are intended to aid referring
because of logistical reasons. Therefore, only selected cen- medical practitioners in appropriate use of [18F]FDG PET/CT
tres can offer this PET tracer. If available, its indication for restaging of the most frequent tumours [24].
would be to precisely stage or restage bone disease, espe- In developing these appropriate use criteria for PET/CT,
cially in cases of non-conclusive lesions. the workgroup used the following definition of appropriate-
Finally, we should mention the role of another nuclear ness to guide their considerations and group discussions:
medicine technique in lung cancer patients who are being “The concept of appropriateness, as applied to health care,
evaluated for surgical treatment. Perfusion lung scintigraphy balances risk and benefit of a treatment, test, or procedure in
is frequently part of the preoperative evaluation for patients the context of available resources for an individual patient
with lung cancers, as it allows an estimation of the perfusion with specific characteristics” [24].
in each area of the lungs, assessing the remaining lung perfu- The workgroup scored each scenario as “appropriate”,
sion following lung resection. The procedure is similar to the “may be appropriate”, or “rarely appropriate” on a scale
perfusion component of the ventilation/perfusion scintigra- from 1 to 9. Scores 7–9 indicate that use of the procedure is
phy for pulmonary embolism. appropriate for the specific scenario and is generally consid-
ered acceptable. Scores 4–6 indicate that use of the proce-
dure may be appropriate for the specific scenario. This
22.1.4 Recurrent Lung Cancer implies that more research is needed to classify the scenario
definitively. Scores 1–3 indicate that use of the procedure is
Lung cancer recurs after surgery in 30–75% of patients [18]. rarely appropriate for the specific scenario and generally is
As in other cancers, after initial treatment (surgery, not considered acceptable [24].
radiotherapy), differentiating recurrence from post-surgical
changes is challenging if using CT alone, as many benign
processes (atelectasis, consolidations, and radiation induced 22.2 [18F]FDG PET/CT Procedure, Data
fibrosis) are very difficult to differentiate from loco-regional Processing, Interpretation,
recurrence [19, 20]. [18F]FDG PET/CT has a great advan- and Reporting
tage, as it differentiates the metabolically active areas.
However, it can yield false-positive results from active Regarding the [18F]FDG PET/CT procedure, data process-
inflammation, especially in the acute post-operative or post- ing, interpretation, and reporting, the EANM published
radiotherapy phase, although this feature of [18F]FDG is now guidelines for [18F]FDG PET/CT for tumour imaging v2.0
being used in the diagnosis of infection and inflammation [25]. These guidelines focus on all the aspects of the proce-
demonstrating a good diagnostic performance [21]. dure, including different protocols with or without contrast-
Recurrence of NSCLC may be classified as loco-regional enhanced CT. The document also includes a summary of
recurrence or distant metastases, the latter being the most topics related to interpretation and reporting with relevant
common form of NSCLC recurrence [19]. Depending on the references in the field [24].
initial stage at diagnosis and on the treatment applied, meta- For adequately interpreting [18F]FDG PET/CT, relevant
static recurrence comprises 39–65.5% of all recurrences [22], clinical information should be taken into account. When
whereas around 30% are loco-regional. Loco-regional recur- reporting lung lesions, it is especially important to analyse
rence is located within the treated hemithorax, usually pre- the evolution in time of the current lesions as compared with
senting as nodules that involve the surgically treated area or previous imaging studies (preferably PET/CT or CT). Lung
the area treated with radiofrequency or microwave ablation, as lesions that appear in a short time are more frequently related
well as in other thoracic structures (bronchial stump, pleura, to inflammatory or infectious processes, whereas lesions that
chest wall, and lymph nodes) [19, 20]. Moreover, apart from slowly increase in size are more probably malignant.
recurrences, new primary lung cancer is also reported in 1–2% Furthermore, when the CT component is done with contrast-
of NSCLC patients per year after initial radical therapy [23]. enhancement (CECT), any relevant findings in the CECT of
the [18F]FDG PET/CT should also be reported. An example
of a relevant and urgent finding is the presence of repletion
22.1.5 Appropriate Use Criteria defects in the bronchial arteries, suggestive of pulmonary
embolism. An example of this is shown in Fig. 22.2.
Recently, guidelines for the appropriate use of [18F]FDG PET/
CT in restaging of patients with cancer have been published.
Representatives from the Society of Nuclear Medicine and 22.3 Initial Staging of Lung Cancer
Molecular Imaging (SNMMI), the European Association of
Nuclear Medicine (EANM), and the American Society of Integrated PET/CT systems have greatly changed the
Clinical Oncology (ASCO) assembled under the auspices of an management algorithms of oncologic patients and have

autonomous workgroup to develop the following Appropriate completely changed the paradigm of oncologic imaging.
a b
Fig. 22.2 Eighty-one year-old woman with a solitary lung nodule bone metastases, confirmed as a stage IV (M1b bone mets) lung cancer,
(SLN) in the right upper lobe. [18F]FDG PET/CECT was performed to and a repletion deficit in a left bronchial artery suggestive of pulmonary
characterise the lung nodule. [18F]FDG PET volumetric image or MIP embolism, later confirmed with a dedicated CT for pulmonary embo-
(a), axial fused PET/CECT (b, e), CECT (c, f), and [18F]FDG PET lism evaluation. Palliative chemotherapy was initiated. Two follow-up
(d, g). Learning point: The SLN presents a high-glucose metabolism, [18F]FDG PET/CT studies performed 4 months (h) and 7 months (i)
and therefore, [18F]FDG PET is clearly positive for malignancy. after the first study evidence a progression of the bone metastatic
Additional findings of the [18F]FDG PET/CECT study are multiple disease
f g
h i
The metabolic information supplied by PET can be ana- allows us to improve our understanding of the biology of
tomically located even to small structures (primary tumours, neoplastic disease, as well as therapy planning and moni-
lymph nodes, soft tissue masses) given the high-resolution toring of response [26].
multidetector CT scanners. In many cases, these small One of the clearest examples of the change in paradigm
structures would pass undetected on CT scanners, but when is the clinical indication of [18F]FDG PET/CT in staging
the image incorporates the metabolic information of PET, of patients with lung cancer. [18F]FDG PET/CT allows an
they become evident, increasing the accuracy of the test. accurate initial pretreatment staging and restaging based
Therefore, PET/CT has become the most reliable imaging on TNM criteria [26]. Focusing on the initial staging of
method for the detection of malignant tumours, tissue char- lung cancer, [18F]FDG PET/CT has a key role in defining
acterisation, staging, and response monitoring [26]. The N and M stages where it is clearly superior to other sec-
integration of the functional and morphologic information tional imaging techniques such as CT and MRI [27]. The
CT component of the PET/CT can be contrast-enhanced 22.4 R

estaging for Detection of Local
to evaluate more accurately the T component. This way, Recurrence
an additional contrast- enhanced CT will not be
necessary. The publication on the appropriate use of [18F]FDG PET/CT
In the last few years, there has been a continuous increase assigned a rating of “appropriate” with an AUC score of 7 for
in the number of targeted or personalised therapeutic this clinical indication. Therefore, it was inferred from the
strategies in the fields of surgery, chemotherapy, and radia- evidence that this indication is appropriate [24].
tion therapy, all under the global denomination of person-
alised medicine [26]. The aim of personalised medicine is to
customise as much as possible the treatment to the character- 22.4.1 Restaging After Initial Treatment
istics of the individual patient, in order to increase the prob- (Surgery, Chemoradiotherapy or
ability of cure and at the same time reduce the probability of Radiotherapy)
unnecessary secondary effects [28, 29]. Based on this prin-
ciple, the first requisite is the need to have the most precise A recent meta-analysis analysed the diagnostic efficacy of
tumour staging and the exact classification of the patient’s PET and PET/CT with [18F]FDG compared to other imaging
tumour stage. In this clinical context, [18F]FDG PET/CT has techniques (OITs) for the detection of recurrent lung cancer
demonstrated it reduces the secondary effects, morbidity and [30]. The inclusion criteria for this study were for studies in
mortality associated to futile invasive procedures, additional the context of secondary lung cancer investigations using PET
staging procedures, and ineffective or non-indicated or PET/CT with [18F]FDG to diagnose lung cancer recurrence,
treatments. considering disease as a consequence of the originally diag-
nosed lung cancer, regardless of whether the recurrence was
local, regional, or distant. Thirteen articles and 1,035 patients
were included. They obtained high pooled/joint sensitivity and
Key Learning Points specificity for PET/CT. Pooled sensitivity for PET, PET/CT,
• Initial staging: [18F]FDG PET/CT has a key role and OITs were 0.94, 0.90, and 0.78, respectively. Pooled spec-
in defining N and M stages, where it is clearly ificity for PET, PET/CT, and OITs were 0.84, 0.90, and 0.80,
superior to other imaging techniques such as CT respectively. With regard to sensitivity, lower values were
and MRI. associated with OITs than PET (p = 0.000) and PET/CT
• The CT component of the PET/CT scan can be (p = 0.005), and there was no significant difference between
contrast-enhanced to evaluate more accurately the PET/CT and PET (p = 0.1). Regarding specificity, values for
T component; in this way, an additional contrast- PET/CT and PET were significantly higher than OITs (both
enhanced CT is not necessary. p = 0.000), with no significant difference between PET/CT
• For SCLC, accurate staging is important for man- and PET p = 0.2. The SROC curves evidenced a better diag-
agement decisions, since time of diagnosis, stage, nostic accuracy associated with PET/CT than PET and OITs.
and performance status have a prognostic value. They conclude PET/CT and PET were superior modalities for
• For SCLC with NEN differentiation, lung NENs the detection of recurrent lung cancer, and PET/CT was supe-
with high expression of somatostatin receptors may rior to CT [30]. With regard to the role of PET/CT in the detec-
be visualised using PET/CT with 68Ga-DOTA- tion of local recurrence, one of the limitations of this study is
conjugated somatostatin receptor targeting peptides that data for the disease are pooled regardless of whether the
(TOC, NOC, and TATE). recurrence was local, regional, or distant. Another limitation is
• Patients with an unknown primary tumour present a that subgroup analysis was not performed considering differ-
pathologically confirmed metastatic lesion with no ent initial treatments.
known primary tumour after the first examinations. Other studies not included in this systematic review and
Furthermore, some patients present with suspected meta-analysis [30] also found PET/CT associated with high
paraneoplastic syndromes without known underly- specificity for the detection of recurrent disease following
ing malignancy. [18F]FDG PET/CT has an impor- initial treatments, including homogeneous patient popula-
tant role in both situations. tions treated with surgery [31, 32], radiotherapy [33–37], or
• Bone scintigraphy with 99mTc-bisphosphonates has radiofrequency ablation [38, 39]. These are discussed below.
had an important role in detecting bone lesions until However, another study by Jiménez-Bonilla et al. [40], which
the widespread implantation of [18F]FDG PET/CT was not included in the meta-analysis, also evaluated a
that allows an accurate M staging including bone heterogeneous population, with patients with all stages of
metastases. NSCLC, from stage I to more advanced stages. They
analysed 59 suspicious lesions in 55 patients reporting an
overall sensitivity and specificity for [18F]FDG PET/CT of focal changes in the lung parenchyma around the treated
100% and 83%, respectively. [18F]FDG PET/CT had an tumour site, most frequently as ground glass opacities
impact on patients’ management in 42 of the 59 cases of sus- (GGO) [19, 41].
pected recurrence. The study by Pastis et al. [33] analysed the diagnostic
efficacy of [18F]FDG PET/CT for detecting local treatment
failure or intrathoracic recurrences after SBRT treatment in
22.4.2 Restaging After Surgery NSCLC patients. Eighty-eight patients were included, and
an [18F]FDG PET/CT was done 3 months after ending
The study by Toba et al. [31] retrospectively included 101 SBRT. [18F]FDG PET/CT was positive in 14% (12 of 88) of
NSCLC patients who had undergone potentially curable patients, being confirmed as true positive in 67% (8 of 12)
operations and were followed with an [18F]FDG PET/CT at of patients. [18F]FDG PET/CT was negative in 86% (76 of
least once a year (233 [18F]FDG PET/CT studies), selecting 88) of patients, being confirmed as true negative in 89% (68
patients without clinical or radiological evidence of recur- of 76) of patients. Therefore, sensitivity, specificity, positive
rence. Eighteen (18%) asymptomatic patients had recurrent predictive value, and negative predictive value were 50%,
disease and 22 recurrent sites were confirmed. [18F]FDG 94%, 67%, and 89%, respectively. They conclude that a
PET/CT correctly detected recurrence in 17 of the 18 (94%) [18F]FDG PET/CT scan 3 months after SBRT treatment of
patients and 21 of the 22 (95%) recurrent sites. Sensitivity, NSCLC was very specific but had a low sensitivity for the
specificity, positive predictive value, negative predictive detection of recurrent disease or treatment failure. They
value, and accuracy were 94.4, 97.6, 89.5, 98.8, and 97.0%, recommend CT (every 6 months for the first 2 years and
respectively. Additionally, [18F]FDG PET/CT detected other every year thereafter [42]) instead of [18F]FDG PET/CT in
not previously known diseases and allowed an early appro- this situation. Furthermore, they state [18F]FDG PET/CT
priate treatment [31]. In this study, all recurrent sites were should be reserved for cases with suspected metastatic dis-
located in intrathoracic or cervical fields. Although inciden- ease, or to evaluate new abnormalities found on CT, or for
tally all the recurrences were intrathoracic, the advantage of the subsequent follow-up when the inflammation due to the
[18F]FDG PET/CT is that it has demonstrated a high accu- radiation therapy has subsided [33].
racy for the detection of distant metastases. Another study focusing on lung cancer patients treated
Another study that analysed the performance of [18F]FDG with SBRT by Zhang et al. [34] analysed whether [18F]FDG
PET/CT for detecting recurrent disease after initial curative PET/CT’s standardised uptake values (SUVs) after SBRT
surgery, also not included in the previous meta-analysis [30], could predict local recurrence in NSCLC. They included 128
is the study by Choi et al. [32]. They included 358 patients patients with 140 biopsy-proven NSCLC, in whom 506 [18F]
having undergone complete resection of NSCLC, who were FDG PET/CT scans were performed. [18F]FDG PET/CT was
prospectively followed up with [18F]FDG PET/CT and con- done between 1 and 6 months after SBRT and subsequently
ventional methods. Recurrent disease occurred in 31% of as clinically indicated (median follow-up 31 months). They
patients. Conventional methods detected half of these recur- conclude that [18F]FDG PET/CT was helpful for distinguish-
rences. In the remaining patients, recurrent disease was ing SBRT-induced consolidation from local recurrence. High
detected both with CT and [18F]FDG PET/CT in 51% of SUVs (>5.0) obtained more than 6 months after SBRT for
patients and only with [18F]FDG PET/CT in 37%. [18F]FDG NSCLC were associated with local failure and should prompt
PET/CT was false negative in six small or hypometabolic biopsy to rule out local recurrence [34]. A similar study by
recurrent lesions. Because of this, the authors recommend an Takeda et al. [35] including 154 NSCLC patients with 214
annual screening method including [18F]FDG PET/CT and a [18F]FDG PET/CT scans done 1 year after SBRT for the
low-dose chest CT [26]. EANM guidelines include an detection of local recurrence reported a sensitivity and speci-
optional, but recommended, low-dose chest CT in the [18F] ficity of 100% and 96–98%, respectively. Whereas these
FDG PET/CT procedure, in order to better assess small lung studies analyse the performance of [18F]FDG PET/CT stud-
lesions [25]. When analysing lung lesions, it is key to take ies done 6 months to one year after SBRT, Van Loon et al.
into account the evolution of the lesions, comparing with [36] reported that an early [18F]FDG PET/CT scan 3 months
previous imaging studies. after radical (chemo-) radiotherapy with curative intent
helped detect progressive disease (PD). They prospectively
included 100 patients with NSCLC who had an [18F]FDG
22.4.3 Restaging After Stereotactic Body PET/CT scan done 3 months after initiation of radiotherapy.
Radiation Therapy PD was detected in 24 patients, only 16 of them with symp-
toms. In the subgroup of symptomatic patients, the impact on
Stereotactic body radiation therapy (SBRT) is an established management of [18F]FDG PET/CT was limited because no
treatment option for early-stage lung cancer. SBRT causes curative treatment could be offered as an alternative.
However, in the asymptomatic group, in 3/8 patients diag- strategies either with [18F]FDG PET/CT, a chest CT, or con-
nosed with PD, an option of a radical treatment could be ventional follow-up with a chest radiograph, all starting
offered. As PD in the asymptomatic patients was diagnosed 3 months after therapy. They concluded that an [18F]FDG
with [18F]FDG PET/CT but not with CT, the authors con- PET/CT 3 months after curative intent (chemo-) radiother-
clude that asymptomatic patients are probably the ones that apy is potentially cost-effective and is more cost-effective
could profit most from an early [18F]FDG PET/CT scan, than CT alone. Additionally [18F]FDG PET/CT in asymp-
although further studies are needed. tomatic patients appears to be equally effective and even
A frequent finding after radiotherapy is the presence of a more cost-effective [18, 46].
variable and persistent [18F]FDG uptake. Hoopes et al. [37]
presented a small patient population of inoperable stage I
NSCLC after SBRT treatment, reporting persistent and mod- Key Learning Points
erately intense [18F]FDG uptake up to 2 years after • Based on the available evidence, this indication is
SBRT. This uptake could be related to inflammation and appropriate.
fibrosis, which is probably more persistent after SBRT com- • Following surgery, chemoradiotherapy, radiother-
pared to conventional fractioned radiotherapy [43]. apy (stereotactic body radiation therapy, SBRT),
radiofrequency ablation (RFA), or microwave abla-
tion (MWA), [18F]FDG PET/CT offers a high accu-
22.4.4 Restaging After Radiofrequency racy and is superior to CT.
Ablation or Microwave Ablation • Following radiotherapy, [18F]FDG PET/CT should
be performed 3 months after completing the treat-
Besides surgery and SBRT, radiofrequency ablation ment to minimise interpretation errors due to the
(RFA) is another option for patients with stage I radiation-induced inflammatory component.
NSCLC. Following RFA the most frequent type of recur-
rence is loco-regional [44]. RFA, the same as SBRT, also
causes GGO in the lung parenchyma around the treated
tumour site [19, 41]. Different algorithms including [18F] 22.5 Restaging for Detection of Metastases
FDG PET/CT 3–6 months after RFA have been proposed
for a close follow-up of these patients [38, 39, 44, 45], The publication on the appropriate use of [18F]FDG PET/CT
although there are still few studies with a limited number assigned a rating of “appropriate” with an AUC score of 7 for
of patients. Yoo et al. [38] evaluated the performance of this clinical indication. Therefore, it was inferred from the
early post-ablation [18F]FDG PET/CT in assessing the evidence that this indication is appropriate [24].
success of the RFA of stage I NSCLC. They included 30 At the moment of diagnosis of NSCLC, around 18–36%
patients with medically inoperable stage I NSCLC who of patients have distant metastases. The detection of these
underwent three [18F]FDG PET/CT scans, one at baseline, metastases at initial staging is key in order to decide on the
another within 4 days after RFA, and the third one a most appropriate management option, as M staging has a
6-month post-RFA. They conclude early post-RFA [18F] direct impact on management and prognosis [47].
FDG PET/CT is not necessary, and 6-month post-RFA Furthermore, in patients apparently radically treated for
FDG PET/CT findings correlate better with the clinical NSCLC, around 20% will relapse due to the presence of
outcome at 1 year. Pou Ucha et al. [39] analysed a small undetected metastases at the time of initial staging [4, 47].
patient population, including seven patients, each with Metastases are usually located in the adrenal glands, bones,
single tumour lesion, who underwent RFA or microwave brain, or liver.
ablation (MWA). CT and [18F]FDG PET/CT were per- Regarding adrenal lesions, [18F]FDG PET has demon-
formed at baseline and follow-up, applying the dual time- strated a good performance in differentiating benign from
point technique when necessary. [18F]FDG PET/CT metastatic lesions in patients with cancer [48]. Few studies
presented a high accuracy and was superior to CT, have specifically addressed this issue in lung cancer patients
although the study had methodological limitations. [49, 50]. The study with most patients included 94 patients
with 113 adrenal masses detected on CT or MRI. [18F]FDG
PET showed a sensitivity, specificity, and accuracy for
22.4.5 Cost-Effectiveness detecting metastatic disease of 93%, 90%, and 92%, respec-
tively [50].
Up to now, Van Loon et al. have published the only cost- For bone metastases, [18F]FDG PET is more sensitive and
effectiveness study of NSCLC follow-up [46]. They included specific than bone scintigraphy [4, 51–53]. The best method
100 NSCLC patients comparing three different follow-up for liver lesions is MRI, but [18F]FDG PET is better than CT
as it detects lesions earlier and is more accurate. MRI is also CT before and after treatment, previously in two dimensions
the best method for brain metastases, as the diagnostic accu- (World Health Organisation, WHO [60]), and more recently
racy of [18F]FDG PET is limited due to the high physiologic in one dimension (response evaluation criteria in solid
[18F]FDG uptake in the normal brain. Other non-[18F]FDG tumour, RECIST [61]). [18F]FDG PET provides functional
tracers must be considered for brain metastases. information, and metabolic changes can be detected earlier
[18F]FDG PET/CT has a high diagnostic performance for than morphologic changes. In this regard, early assessment
the detection of metastases. This often changes the stage of of response can be of great value in patients with cancer, in
the disease and has an impact on management. particular in lung cancer. A large proportion of patients
As discussed in the “local recurrence” section, a meta- undergo treatments that are toxic and expensive with no
analysis analysed the diagnostic efficacy of PET/CT com- response, when there are second-line treatments available
pared to OITs for the detection of recurrent lung cancer, [62]. Early assessment of response to therapy can help tailor
considering disease as a consequence of the originally diag- treatments in order to continue treatments in responding
nosed lung cancer, regardless of whether the recurrence was patients and discontinue treatments and change to second-
local, regional, or distant. They obtained high pooled/joint line treatments in non-responders. Current evidence in this
sensitivity and specificity for PET/CT, concluding PET/CT setting shows that [18F]FDG PET/CT response is probably
and PET were superior modalities for the detection of recur- earlier and more accurate than CT response [62]. However,
rent lung cancer, and PET/CT was superior to CT [30]. an important issue to be solved is related to the need to stan-
In a meta-analysis that evaluated the performance of [18F] dardise the methodology. The EANM has recently updated
FDG PET/CT for the detection of distant malignancies in the [18F]FDG PET/CT procedure guidelines for tumour
various cancers, 41 studies and 4,305 patients were included imaging, focusing on harmonisation in order to make the
[54]. Of these, the studies with data on lung cancer were 5 methodology and results comparable worldwide [25]. In this
[55–59], comprising 578 patients. The pooled sensitivity, regard, one of the methodology aspects that need to be stan-
specificity, positive likelihood ratio, and negative likelihood dardised in the response assessment studies is timing of the
ratio were 0.91, 0.96, 25.9, and 0.09, respectively. They con- [18F]FDG PET/CT. The best timing has not been standardised
clude that [18F]FDG PET/CT has an excellent diagnostic per- yet, if done too early it might overestimate [18F]FDG uptake
formance for the detection of distant malignancies in patients because glucose metabolism might be present in cells that
with various cancers, especially in lung cancer, breast can- are lethally damaged and because of inflammatory processes
cer, and head and neck cancer [54]. in responding tissues [62]. If it is done too late, other prob-
lems might appear, such as late evaluation of response or the
risk of tumour repopulation. In summary, large-scale trials
Key Learning Points are needed, applying strict methodological standardisation.
• Based on the available evidence, this indication is In patients with locally advanced lung cancer who
appropriate. undergo multimodality treatment, correct restaging after
• [18F]FDG PET/CT has a high diagnostic perfor- induction therapy is needed [62]. In NSCLC stage IIIa-N2, a
mance for the detection of metastases. favourable outcome after surgical combined modality treat-
• The findings of an [18F]FDG PET/CT often change ment highly depends on pathological downstaging or clear-
stage of the disease and therefore have an impact on ance of all tumour lesions in the mediastinal lymph nodes
management. after the induction phase. CT has limitations when evaluat-
ing response to induction treatment because small-sized
lymph nodes can still harbour metastatic disease, whereas
large lymph nodes can be caused by inflammatory factors or
scarring [62, 63]. Several studies have analysed the role of
22.6 Treatment Response Evaluation [18F]FDG PET in this clinical setting with good results.
One fair-quality study of patients with stage IIIa non-
The publication on the appropriate use of [18F]FDG PET/CT small cell lung cancer with biopsy-proven N2 disease that
assigned a rating of “appropriate” with an AUC score of 7 for underwent neoadjuvant chemoradiotherapy and subsequent
this clinical indication. Therefore, it was inferred from the restaging (n = 93) found PET/CT associated with a sensitiv-
evidence that this indication is appropriate [24]. ity of 62% a specificity of 88% for identifying N2 disease.
Personalised medicine is based on tailoring treatments to Compared with pathological staging, the proportion of
the individual patient. For this reason, it is of utmost impor- patients with correct stage classification was greater with
tance to have tools that provide an early and precise assess- PET/CT than CT across tumour stages 0 through IV, though
ment of response to therapy [27, 28]. Traditionally, tumour differences were only statistically significant for stage 0 and
response has been assessed comparing the tumour size on stage I [64]. Other studies have shown that patients who are
downstaged following neoadjuvant therapy and then undergo had recognised the need for standardised response assess-
resection have a significantly longer 5-year survival of ment criteria and proposed the “WHO criteria” based on CT
40–50% [65–67] than those who have residual N2 disease findings [76, 77]. The “WHO criteria” were then refined to
[68]. Therefore, identifying patients who are N2 negative develop the “Response Evaluation Criteria in Solid Tumors
after completion of their neoadjuvant therapy is a critical (RECIST)”. RECIST 1.0 criteria were published in 2000
component for patient selection for thoracotomy [64]. and updated (RECIST 1.1) in 2009. The updated criteria
However, correctly identifying responding from non- further clarify which lesions should be measured to monitor
responding patients remains a challenge. Most patients with tumour response and how to measure them [76, 78]. In 2002
pathologically diagnosed N2 disease have undergone medi- the first successful application of PET to monitor response
astinoscopy. Repeat mediastinoscopy is difficult, often inac- to a targeted therapy (imatinib in patients with gastrointesti-
curate [69, 70], and potentially dangerous, in particular nal stromal tumour) was described [79]. This study showed
following radiotherapy. Furthermore, studies have evidenced the great potential of PET to assess early response to treat-
a high false-negative rate of repeat mediastinoscopy follow- ment, when the response assessment based on size criteria
ing neoadjuvant therapy, with a range of 25–42% [69, 71]. (RECIST) was starting to be seen as too inaccurate [75].
Fine needle aspiration (FNA) guided by endoscopic ultra- Since then, many studies have shown the efficacy of PET in
sound (EUS) has been used as a restaging method with a evaluation response to therapy. Also, new response criteria
reported accuracy of 83% in one study with a small patient including PET (PET Response Criteria In Solid Tumours,
population (n = 19) following neoadjuvant chemoradiother- PERCIST) were published in 2009 [75, 80]. Furthermore,
apy. The main problems of EUS-FNA are that it does not many studies present data on the repeatability and harmoni-
allow an adequate visualisation of the lower paratracheal sation of PET, proposing the application of thresholds to
nodes [72] and it is available in only a few centres. In sum- discriminate between responders and non-responders [75].
mary, the surgeon often has the clinical stage assessed only A meta-analysis published in 2016 compared RECIST and
by repeat PET/CT or CT to back up the management deci- PERCIST, including 6 studies and 268 patients [81]. The
sions. The prospective study by Cerfolio et al. concludes that results show that the agreement between RECIST and
repeat integrated [18F]FDG PET/CT is superior to repeat CT PERCIST was only moderate (κ = 0.590) with discordant
for the restaging of patients with N2 stage IIIa NSCLC after response classifications found in 38% of the patients.
neoadjuvant chemoradiotherapy [64]. Overall, the response rate by PERCIST was significantly
A meta-analysis published in 2012 analysed the value of higher than by RECIST (35% vs. 54%).
[ F]FDG PET and CT in predicting the pathological tumour
18
With regard to lung cancer, only very few studies have
response of NSCLC after neoadjuvant therapy. Pathological compared the prognostic value of RECIST and PERCIST
outcome was the gold standard. Thirteen studies and 414 response classifications. Ding et al. included a group of 44
patients were included with different neoadjuvant treat- patients with NSCLC treated by chemotherapy finding that
ments: chemoradiotherapy in 5 studies, chemotherapy in 2 tumour response by PERCIST was a better predictor for
studies, and mixed treatments in the remaining 5 studies progression- free survival than tumour response by
[73]. Pooled sensitivity, specificity, positive predictive value, RECIST. In a multivariate analysis including response by
and negative predictive value for prediction of response with RECIST and PERCIST, only response by PERCIST was a
PET were 83%, 84%, 74%, and 91%, respectively. The pre- significant prognostic factor [82]. In summary, PERCIST
dictive value of PET in NSCLC patients with pathological has a significant potential to improve response assessment in
response was significantly higher than that of CT (p < 0.05). clinical trials, although more studies are needed. These
However, these results have to be analysed taking into studies should ideally be performed as part of randomised
account the limitations of the meta-analysis, such as the controlled trials [76].
heterogeneity of the studies, the mixed pathological types, Immunotherapy is presenting a rapid development.
and their retrospective design. Taking into account the men- Because of this, there is now a special interest in criteria that
tioned limitations, they conclude that [18F]FDG PET is use- consider the specific type of responses seen in clinical trials
ful for predicting NSCLC non-responders to neoadjuvant of immune-modulating agents [76]. In these trials, tumour
therapy and it has a better predictive value than that of CT for shrinkage in patients who eventually responded occurred
evaluating pathological documented responses. later than for chemotherapy or targeted therapies. In some
To assess response to therapy in cancer, different criteria patients, there is even a temporary increase in tumour size,
have been developed. In 1999 the first standardised PET denominated pseudo-progression. Although new studies
response criteria were published by the European must assess if different immunotherapies have different
Organization for Research and Treatment of Cancer effects on tumour cells and immune cells, it seems possible
(EORTC) [74], based on scarce supporting evidence [75]. that the future criteria will have to be treatment-specific [76].
Previously, in 1979 the World Health Organisation (WHO) A recent phase II study in 138 patients with metastatic
NSCLC treated with atezolizumab demonstrated a marked not assess the viability of the tissue [85]. Molecular imaging
decrease in [18F]FDG uptake 6 weeks after initiating therapy with [18F]FDG PET provides functional information which is
in responding tumour, quite similar to previous observations much more precise when evaluating viability and monitoring
with chemotherapy or targeted therapy, reporting pseudo- response to therapy. Several studies performed in small cohorts
progression in only two cases [83]. have evaluated the role of [18F]FDG PET to assess response to
In summary, the available evidence indicates that chemotherapy after one to three cycles, measuring different
PERCIST is a better approach to assess tumour response PET parameters involving semiquantitative and quantitative
than RECIST but still has to be proven with systematic clini- analyses, decrease in maximum standardised uptake value
cal studies including randomised trials. The trials will have (SUVmax), total glycolytic volume (TGV), or total lesion gly-
to determine if the response rate by PERCIST predicts the colysis (TLG) [85]. All these studies suggest that in patients
clinical benefit of new anticancer drugs better than the with malignant pleural mesothelioma treated with chemother-
response rate by RECIST [76]. apy and early reduction in [18F]FDG uptake (after one to three
cycles) may be significantly correlated with patient outcome
and survival, indicating the predictive and prognostic value of
Key Learning Points
[18F]FDG PET in these patients [85]. Figure 22.3 presents the
• Based on the available evidence, this indication is case of a patient with mesothelioma.
appropriate.
• Personalised medicine is based on tailoring treat-
Key Learning Points
ments to the individual patient.
• Molecular imaging with [18F]FDG PET provides
• It is of utmost importance to have tools that provide
functional information which is much more precise
an early and precise assessment of response to
than anatomic imaging provided by CT and/or
therapy.
MRI.
• The best timing has not been standardised yet.
• [18F]FDG PET is especially useful when evaluating
• PERCIST is a better approach to assess tumour
tumour viability and monitoring response to
response than RECIST but still has to be proven
therapy.
with systematic clinical studies.
• In patients with malignant pleural mesothelioma
treated with chemotherapy, early reduction in [18F]
FDG uptake (after one to three cycles) may be sig-
22.7 Mesothelioma nificantly correlated with patient outcome and sur-
vival, thus indicating the high predictive and
Malignant pleural mesothelioma (MPM) is an aggressive prognostic value of [18F]FDG PET in these patients.
tumour that originates from mesothelial cells of the pleura. The
disease is mainly related to the exposure to asbestos fibres.
There is a long latency from the asbestos fibre exposure to the
onset of the disease. This is one of the reasons why there is an 22.8 Other Thoracic Tumours
increasing incidence worldwide [84, 85]. Due to the fact that
most patients present an advanced stage of the disease at the The most frequent tumours in the anterior mediastinum are
moment of initial diagnosis, radical surgery is indicated in a thymoma, teratoma, and thoracic lymphoma (the latter will
small subgroup of them, whereas most patients are candidates to be addressed in Chap. 27 of this book). Therefore, the dif-
be treated with chemotherapy during the course of the disease ferential diagnosis of anterior mediastinal lesions is impor-
[85, 86]. In order to apply the treatment efficiently and person- tant since tumours with a variety of histological types can
alised to the individual patient, a precise assessment and moni- appear in this region [88].
toring of response is essential. Although several response criteria
have been proposed for malignant pleural mesothelioma, it is
still not clear which is the optimal one. Classic criteria are based 22.8.1 Thymoma
on morphological criteria such as measurements on CT, but they
are considered complicated and particularly difficult in malig- The most anteriorly located mediastinal tumours are usually
nant pleural mesothelioma due to its diffuse pattern of growth different from malignant lymphomas. Surgery is considered
throughout the pleural surface. Modified RECIST criteria are unnecessary for many cystic lesions at initial diagnosis, radio-
now considered the best option available for malignant pleural logical follow-up being the preferred option. Nevertheless,
mesothelioma [85, 87], although they present problems such as differentiating thymic cysts from thymomas based on CT is
a high interobserver and intraobserver variability and they do sometimes difficult due to the high viscosity of intracystic
a b
d e
Fig. 22.3 Sixty-four year-old man diagnosed with epithelioid meso- ment evidenced a complete metabolic response (h). However, a third
thelioma. [18F]FDG PET/CECT performed for initial staging. [18F]FDG [18F]FDG PET/CECT performed 6 months later due to suspected
PET/CECT was performed. [18F]FDG PET volumetric image or MIP relapse showed multiple malignant lesions suggestive of relapse in the
(a), axial fused PET/CECT (b, e), CECT (c, f), and [18F]FDG PET (d, thorax (i), with pericardic implants (j–l), one of 7 cm that rubs and
g). Several pleural malignant lesions were detected. The patient was compresses the right ventricle, and abdominal progression with several
treated with pneumonectomy, chemotherapy, and radiotherapy. A sec- abdominal implants (m–o)
ond [18F]FDG PET/CECT was performed 2 years later at end of treat-
f g
h i
j k
l m
n o
materials. MRI is better than CT in detecting cystic tumours evidenced significant differences in SUVmax in four groups of
due to its improved contrast resolution, but there is still some anterior mediastinal tumours: low-grade thymomas, high-
overlap between benign and malignant tumours. [18F]FDG grade thymomas, thymic carcinomas, and malignant lympho-
PET has demonstrated its usefulness for differentiating thymic mas, with SUVmax values of 3.1 ± 0.7, 4.3 ± 1.5, 8.6 ± 3.1, and
cysts from thymomas [88]. A recent study by Kitami et al. has 10.1 ± 2.5, respectively [88]. They concluded that anterior
mediastinal tumours with a SUVmax of >6.0 are likely to be 22.9 Conclusions

thymic carcinomas or malignant lymphomas. Thus, those
tumours should be biopsied to determine the therapeutic strat- [18F]FDG PET/CT has been extensively studied in lung can-
egy. Furthermore, all anterior mediastinal tumours with a cer, and there is evidence showing its utility in patient man-
SUVmax of 2.0–4.0 are likely to be benign, and thus, a preop- agement decisions. Furthermore, personalised medicine is
erative biopsy could be unnecessary in these cases [88]. based on tailoring treatments to the individual patient. For
Thymic epithelial tumours (TETs) are rare malignancies this reason, it is of utmost importance to have tools that pro-
arising in the anterior mediastinum that show a very variable vide an early and precise assessment of response to therapy.
biologic behaviour, ranging from slow-growing benign lesions Early assessment of response to therapy can help tailor treat-
to highly aggressive carcinomas [89]. The treatment strategy ments in order to continue treatments in responding patients
for TETs is based on its resectability at initial diagnosis. and discontinue treatments and change to second-line treat-
However, most cases are unresectable or in advanced stages at ments in non-responders. Current evidence in this setting
initial diagnosis and are eligible of neoadjuvant or palliative shows that the [18F]FDG PET/CT response is probably ear-
chemotherapy, followed by surgery if they become resectable. lier and more accurate that the CT response.
Recurrent disease is frequent, and the management is similar
to the one for the initial diagnosis of TETs [89]. Contrast-
enhanced CT is routinely used at all stages of the management References
of these patients. Since 2006 several publications have shown
the utility of [18F]FDG PET/CT in these patients in different 1. http://seer.cancer.gov/statfacts/html/lungb.html. Accessed 10 Sept
clinical situations: characterisation of thymic masses, disease 2016.
staging, monitoring the efficacy of targeted therapies in 2. American Cancer Society. Detailed guide: lung cancer non-small
cell. What is non small cell lung cancer? American Cancer Society;
advanced TETs, early response to therapy, and prognostic 2010.
value [89]. However, due to the limited available evidence, it 3. Langer CJ, Besse B, Gualberto A, Brambilla E, Soria JC. The
has not yet been included in clinical guidelines. evolving role of histology in the management of advanced non-
small-cell lung cancer. J Clin Oncol. 2010;28:5311–20.
4. Ambrosini V, Nicolini S, Caroli P, Nanni C, Massaro A, Marzola
MC, et al. PET/CT imaging in different types of lung cancer: an
22.8.2 Teratomas overview. Eur J Radiol. 2012;81:988–1001.
5. Travis WD, Brambilla E, Noguchi M, Nicholson AG, Geisinger
Currently there is very limited evidence in the literature KR, Yatabe Y, et al. International Association for the Study of Lung
Cancer/American Thoracic Society/European Respiratory Society
regarding the role of [18F]FDG PET in teratomas. However, international multidisciplinary classification of lung adenocarci-
we could expect [18F]FDG PET may be useful as in other noma. J Thorac Oncol. 2011;6:244–85.
tumours with [18F]FDG avidity. We can expect that malignant 6. Kut V, Spies W, Spies S, Gooding W, Argiris A. Staging and
or undifferentiated teratomas would show high [18F]FDG monitoring of small cell lung cancer using [18F]Fluoro-2-deoxy-
D-glucose-positron emission tomography (FDG-PET). Am J Clin
uptake, whereas benign or differentiated teratomas have Oncol. 2007;30:45–50.
shown low [18F]FDG uptake [88]. Therefore, [18F]FDG PET 7. Divisi D, Di Tommaso S, Di Leonardo G, Brianzoni E, De Vico
may be a useful help to characterise suspected teratoma lesions A, Crisci R. 18-Fluorine fluorodeoxyglucose positron emission
(malignant or benign) based on the [18F]FDG uptake and help tomography with computerized tomography versus computer-
ized tomography alone for the management of solitary lung nod-
in management decisions of malignant teratomas in the usual ules with diameters inferior to 1.5 cm. Thorac Cardiovasc Surg.
situations (staging and restaging the disease and monitoring 2010;58:422–6.
therapy), although new studies are warranted to demonstrate 8. Agarwal M, Brahmanday G, Bajaj SK, Ravikrishnan KP, Wong
the efficacy of [18F]FDG PET in these clinical situations. CGO. Revisiting the prognostic value of preoperative 18F-fluoro-2-
deoxyglucose (18F-FDG) positron emission tomography (PET) in
early-stage (I & II) non-small cell lung cancers (NSCLC). Eur J
Nucl Med Mol Imaging. 2010;37:691–8.
Key Learning Points 9. Pommier P, Touboul E, Chabaud S, Dussart S, Le Pechoux C,
• Initial evidence suggests the usefulness of [18F] Giammarile F, Carrie C. Impact of 18F-FDG PET on treatment
strategy and 3D radiotherapy planning in non-small cell lung
FDG PET/CT in these patients in different clinical cancer: a prospective multicenter study. AJR Am J Roentgenol.
situations. 2010;195:350–5.
• The clinical usefulness of [18F]FDG PET/CT is 10. de Cabanyes Candela S, Detterbeck FC. A systematic review of
emerging especially for the characterisation of thy- restaging after induction therapy for stage IIIa lung cancer: predic-
tion of pathologic stage. J Thorac Oncol. 2010;5:389–98.
mic masses, disease staging, prognosis, monitoring 11. Fischer BM, Mortensen J, Langer SW, Loft A, Berthelsen AK,
the efficacy of targeted therapies in advanced TETs, Daugaard G, et al. PET/CT imaging in response evaluation of
and assessing early response to therapy. patients with small cell lung cancer. Lung Cancer. 2006;54:
41–9.
12. Schreyögg J, Weller J, Stargardt T, Herrmann K, Bluemel C,

tomography in the evaluation of response to chemotherapy. Cancer
Dechow T, et al. Cost–effectiveness of hybrid PET/CT for staging Chemother Rev. 2008;3:77–86.
of non-small cell lung cancer. J Nucl Med. 2010;51:1668–75. 29. Delgado Bolton RC, Carreras Delgado JL. Positron emission

13. Bozkurt MF, Virgolini I, Bagolova S, Beheshti M, Rubello D, tomography (PET) in the evaluation of response to therapy in non-
Decristoforo C, et al. Guideline for PET/CT imaging in neuroendo- small cell lung cancer. Curr Cancer Ther Rev. 2009;5:20–7.
crine neoplasms with 68Ga-DOTA-conjugated somatostatin recep- 30. He YQ, Gong HL, Deng YF, Li WM. Diagnostic efficacy of PET
tor targeting peptides and 18F-DOPA. Eur J Nucl Med Mol Imaging. and PET/CT for recurrent lung cancer: a meta-analysis. Acta
2017;44:1588–601. Radiol. 2014;55:309–17.
14. Giammarile F, Muylle K, Delgado Bolton R, Kunikowska J,
31. Toba H, Sakiyama S, Otsuda H, Kawakami Y, Takizawa H, Kenzaki
Haberkorn U, Oyen U. Dosimetry in clinical radionuclide therapy: K, et al. 18F-fluorodeoxyglucose positron emission tomogra-
the devil is in the detail. Eur J Nucl Med Mol Imaging. 2017; phy/computed tomography is useful in postoperative follow-up
https://doi.org/10.1007/s00259-017-3820-3. of asymptomatic non-small-cell lung cancer patients. Interact
15. Delgado-Bolton RC, Fernández-Pérez C, González-Maté A,
Cardiovasc Thorac Surg. 2012;15:859–64.
Carreras JL. Meta-analysis of the performance of 18F-FDG PET in 32. Choi SH, Kim YT, Kim SK, Kang KW, Goo JM, Kang CH, Kim
primary tumor detection in unknown primary tumors. J Nucl Med. JH. Positron emission tomography-computed tomography for post-
2003;44:1301–14. operative surveillance in non-small cell lung cancer. Ann Thorac
16. García Vicente AM, Delgado-Bolton RC, Amo-Salas M, López- Surg. 2011;92:1826–32.. discussion 1832
Fidalgo J, Caresia Aroztegui AP, García Garzón JR, et al. 33. Pastis NJ Jr, Greer TJ, Tanner NT, Wahlquist AE, Gordon LL, Sharma
18
F-fluorodeoxyglucose positron emission tomography in the diag- AK, et al. Assessing the usefulness of 18F-fluorodeoxyglucose
nosis of malignancy in patients with paraneoplastic neurological PET-CT scan after stereotactic body radiotherapy for early-stage
syndrome: a systematic review and metaanalysis. Eur J Nucl Med non-small cell lung cancer. Chest. 2014;146:406–11.
Mol Imaging. 2017;44:1575–87. 34. Zhang X, Liu H, Balter P, Allen PK, Komaki R, Pan T, et al.
17. Van den Wyngaert T, Strobel K, Kampen WU, Kuwet T, van der Positron emission tomography for assessing local failure after ste-
Bruggen W, Mohan HK, et al. The EANM practice guidelines for reotactic body radiotherapy for non-small-cell lung cancer. Int J
bone scintigraphy. Eur J Nucl Med Mol Imaging. 2016;43:1723–38. Radiat Oncol Biol Phys. 2012;83:1558–65.
18. Inoue T, Kim EE, Komaki R, Wong FC, Bassa P, Wong WH, et al. 35. Takeda A, Kunieda E, Fujii H, Yokosuka N, Aoki Y, Oooka Y,
Detecting recurrent or residual lung cancer with FDG-PET. J Nucl et al. Evaluation for local failure by 18F-FDG PET/CT in compari-
Med. 1995;36:788–93. son with CT findings after stereotactic body radiotherapy (SBRT)
19. Sudarski S, Henzler T, Schoenberg SO. Post-therapeutic positron for localized non-small-cell lung cancer. Lung Cancer. 2013;79:
emission tomography/computed tomography for early detection 248–53.
of non-small cell lung cancer recurrence. Transl Lung Cancer Res. 36. van Loon J, Grutters J, Wanders R, Boersma L, Oellers M,

2013;2:295–303. Dingemans AM, et al. Follow-up with 18FDG-PET-CT after radi-
20. Caulo A, Mirsadraee S, Maggi F, Leccisotti L, van Beek EJ,
cal radiotherapy with or without chemotherapy allows the detec-
Bonomo L. Integrated imaging of non-small cell lung cancer recur- tion of potentially curable progressive disease in non-small cell
rence: CT and PET-CT findings, possible pitfalls and risk of recur- lung cancer patients: a prospective study. Eur J Cancer. 2009;45:
rence criteria. Eur Radiol. 2012;22:588–606. 588–95.
21. Jiménez-Ballvé A, Pérez-Castejón MJ, Delgado-Bolton RC,
37. Hoopes DJ, Tann M, Fletcher JW, Forquer JA, Lin PF, Lo SS, et al.
Sánchez-Enrique C, Vilacosta I, Vivas D, et al. Assessment of the FDG-PET and stereotactic body radiotherapy (SBRT) for stage I
diagnostic accuracy of 18F-FDG PET/CT in prosthetic infective non-small-cell lung cancer. Lung Cancer. 2007;56:229–34.
endocarditis and cardiac implantable electronic device infection: 38. Yoo DC, Dupuy DE, Hillman SL, Fernando HC, Rilling WS,

comparison of different interpretation criteria. Eur J Nucl Med Mol Shepard JA, Siegel BA. Radiofrequency ablation of medically
Imaging. 2016;43:2401–12. ioneperable stage IA non-small cell lung cancer: are early post-
22. Rubins J, Unger M, Colice GL, American College of Chest
treatment PET findings predictive of treatment outcome? AJR Am J
Physicians. Follow-up and surveillance of the lung cancer patient Roentgenol. 2011;197:334–40.
following curative intent therapy: ACCP evidence-based clinical 39. Pou Ucha JL, Nogueiras Alonso JM, Alvarez Paez AM, Suárez
practice guideline (2nd edition). Chest. 2007;132:355S–67S. Arfenoni BA, Puig AS, López Lopez AM, et al. Diagnostic yield
23. Ponn RB. Lightning can strike twice: second primary lung cancers. of baseline and follow-up PET/CT studies in ablative therapy
Chest. 2000;118:1526–9. for non-small cell lung cancer. Rev Esp Med Nucl Imagen Mol.
24. Jadvar H, Colletti P, Delgado-Bolton R, Esposito G, Krause BJ, 2012;31:301–7.
Iagaru AH, et al. Appropriate use criteria for FDG PET/CT restag- 40. Jiménez-Bonilla JF, Quirce R, Martínez-Rodríguez I, Banzo I,

ing and response assessment of malignant disease. J Nucl Med. Rubio-Vassallo AS, Del Castillo-Matos R, et al. Diagnosis of recur-
2017;58:2026–2037. rence and assessment of post-recurrence survival in patients with
25. Boellaard R, Delgado-Bolton R, Oyen WJ, Giammarile F, Tatsch extracranial non-small cell lung cancer evaluated by 18F-FDG PET/
K, Eschner W, et al. FDG PET/CT: EANM procedure guidelines CT. Lung Cancer. 2013;81:71–6.
for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging. 41. Bojarski JD, Dupuy DE, Mayo-Smith WW. CT imaging findings of
2015;42:328–54. pulmonary neoplasms after treatment with radiofrequency ablation:
26. Flechsig P, Mehndiratta A, Haberkorn U, Kratochwil C, Giesel results in 32 tumors. AJR Am J Roentgenol. 2005;185:466–71.
FL. PET/MRI and PET/CT in lung lesions and thoracic malignan- 42. Colt HG, Murgu SD, Korst RJ, Slatore CG, Unger M, Quadrelli
cies. Semin Nucl Med. 2015;45:268–81. S. Follow-up and surveillance of the patient with lung cancer after
27. Bustos García de Castro A, Ferreiros Domínguez J, Delgado Bolton curative-intent therapy: diagnosis and management of lung cancer,
R, Fernández Pérez C, Cabeza Martínez B, García García-Esquinas 3rd ed: American College of Chest Physicians evidence-based clin-
M, Carreras Delgado JL. PET-CT in presurgical lymph node stag- ical practice guidelines. Chest. 2013;143(5_Suppl):e437S–54S.
ing in non-small cell lung cancer: the importance of false-negative 43. Cuaron J, Dunphy M, Rimner A. Role of FDG-PET scans in stag-
and false-positive findings. Radiologia. 2017;59:147–58. ing, response assessment, and follow-up care for non-small cell
28. Delgado Bolton RC, Izarduy LP, Carreras Delgado JL. Positron lung cancer. Front Oncol. 2012;2:208. https://doi.org/10.3389/
emission tomography and positron emission tomography/computed fonc.2012.00208.
44. Beland MD, Wasser EJ, Mayo-Smith WW, Dupuy DE. Primary States, National Cancer Institute of Canada. J Natl Cancer Inst.
non-small cell lung cancer: review of frequency, location, and 2000;92:205–16.
time of recurrence after radiofrequency ablation. Radiology. 62. Vansteenkiste J, Fischer BM, Dooms C, Mortensen J. Positron-
2010;254:301–7. emission tomography in prognostic and therapeutic assessment
45. Eradat J, Abtin F, Gutierrez A, Suh R. Evaluation of treatment of lung cancer: systematic review. Lancet Oncol. 2004;5:531–40.
response after nonoperative therapy for early-stage non-small cell https://doi.org/10.1016/S1470-2045(04)01564-5.
lung carcinoma. Cancer J. 2011;17:38–48. 63. Margaritora S, Cesario A, Galetta D, D'Andrilli A, Macis G,

46. van Loon J, Grutters JP, Wanders R, Boersma L, Dingemans AM, Mantini G, et al. Ten year experience with induction therapy
Bootsma G, et al. 18FDG-PET-CT in the follow-up of non-small cell in locally advanced non-small cell lung cancer (NSCLC): is clinical
lung cancer patients after radical radiotherapy with or without che- re-staging predictive of pathological staging? Eur J Cardiothorac
motherapy: an economic evaluation. Eur J Cancer. 2010;46:110–9. Surg. 2001;19:894–8.
47. Quint LE. Staging non-small cell lung cancer. Cancer Imaging. 64. Cerfolio RJ, Bryant AS, Ojha B. Restaging patients with N2

2007;7:148–59. (stage IIIa) non-small cell lung cancer after neoadjuvant chemo-
48. Yun M, Kim W, Alnafisi N, Lacorte L, Jang S, Alavi A. 18F-FDG therapy: a prospective study. J Thorac Cardiovasc Surg. 2006;131:
PET in characterizing adrenal lesions detected on CT or MRI. J 1229–35.
Nucl Med. 2001;42:1795–9. 65. Detterbeck F, Socinski M. C. In: Detterbeck F, Socinski M, Rivera
49. Gupta NC, Graeber GM, Tamim WJ, Rogers JS, Irisari L, Bishop M, Rosenman J, editors. Diagnosis and treatment of lung cancer.
HA. Clinical utility of PET-FDG imaging in differentiation of An evidence-based guide for the practicing clinician. Philadelphia:
benign from malignant adrenal masses in lung cancer. Clin Lung WB Saunders Co; 2001. p. 267–282.
Cancer. 2001;3:59–64. 66. Bueno R, Richards W, Swanson S, Jaklitsch MT, Lukanich JM,
50. Kumar R, Xiu Y, Yu JQ, Takalkar A, El-Haddad G, Potenta S, et al. Mentzer SJ, Sugarbaker DJ. Nodal stage after induction therapy
18
F-FDG PET in evaluation of adrenal lesions in patients with lung for stage IIIA lung cancer determines patient survival. Ann Thorac
cancer. J Nucl Med. 2004;45:2058–62. Surg. 2000;70:1826–31.
51. Liu T, Xu JY, Xu W, Bai YR, Yan WL, Yang HL. Fluorine-18 deox- 67. Voltolini L, Luzzi L, Ghiribelli C, Paladini P, Di Bisceglie M, Gotti
yglucose positron emission tomography, magnetic resonance imag- G. Results of induction chemotherapy followed by surgical resec-
ing and bone scintigraphy for the diagnosis of bone metastases in tion in patients with stage IIIA (N2) non-small cell lung cancer: the
patients with lung cancer: which one is the best? A meta-analysis. importance of the nodal down-staging after chemotherapy. Eur J
Clin Oncol (R Coll Radiol). 2011;23:350–8. Cardiothorac Surg. 2001;20:1106–12.
52. Cheran SK, Herndon JE, Patz EF. Comparison of whole-body FDG- 68. Komaki R, Cox JD, Hartz AJ, Byhardt RW, Perez-Tamayo C,
PET to bone scan for detection of bone metastases in patients with a Clowry L, et al. Characteristics of long-term survivors after treat-
new diagnosis of lung cancer. Lung Cancer. 2004;44:317–25. ment for inoperable carcinoma of the lung. Am J Clin Oncol.
53. Min JW, Um SW, Yim JJ, Yoo CG, Han SK, Shim YS, Kim YW. The 1985;8:362–70.
role of whole-body FDG PET/CT, Tc-99m MDP bone scintigra- 69. Mateu-Navarro M, Rami-Porta R, Bastus-Piulats R, Cirera-

phy, and serum alkaline phosphatase in detecting bone metastasis Nogueras L, GonzCález-Pont G. Remediastinoscopy after induc-
in patients with newly diagnosed lung cancer. J Korean Med Sci. tion chemotherapy in non-small cell lung cancer. Ann Thorac Surg.
2009;24:275–80. 2000;70:391–5.
54. Xu G, Zhao L, He Z. Performance of whole-body PET/CT for the 70. Pitz CC, Maass KW, Van Swieten HA, de la Rivière AB, Hofman
detection of distant malignancies in various cancers: a systematic P, Schramel FM. Surgery as part of combined modality treat-
review and meta-analysis. J Nucl Med. 2012;53:1847–54. ment in stage IIIB non-small cell lung cancer. Ann Thorac Surg.
55. Cerfolio RJ, Ojha B, Bryant AS, Raghuveer V, Mountz JM,
2002;74:164–9.
Bartolucci AA. The accuracy of integrated PET-CT compared with 71. Van Schil P, van der Schoot J, Poniewierski J, Pauwels M, Carp
dedicated PET alone for the staging of patients with nonsmall cell L, Germonpré P, De Backer W. Remediastinoscopy after neo-
lung cancer. Ann Thorac Surg. 2004;78:1017–23. adjuvant therapy for non-small cell lung cancer. Lung Cancer.
56. Fischer BM, Mortensen J, Langer SW, Loft A, Berthelsen AK, 2002;37:281–5.
Petersen BI, et al. A prospective study of PET/CT in initial staging 72. Wallace MB, Ravenel J, Block MI, Fraig M, Silvestri G, Wildi S,
of small-cell lung cancer: comparison with CT, bone scintigraphy et al. Endoscopic ultrasound in lung cancer patients with a nor-
and bone marrow analysis. Ann Oncol. 2007;18:338–45. mal mediastinum on computed tomography. Ann Thorac Surg.
57. Ohno Y, Koyama H, Onishi Y, Takenaka D, Nogami M, Yoshikawa T, 2004;77:1763–8.
et al. Non-small cell lung cancer: whole-body MR examination for 73. Zhang C, Liu J, Tong J, Sun X, Song S, Huang G. 18F-FDG-PET
M-stage assessment — utility for whole-body diffusion-weighted evaluation of pathological tumour response to neoadjuvant therapy
imaging compared with integrated FDG PET/CT. Radiology. in patients with NSCLC. Nucl Med Commun. 2013;34:71–7.
2008;248:643–54. 74. Young H, Baum R, Cremerius U, Herholz K, Hoekstra O,

58. Yi CA, Shin KM, Lee KS, Kim BT, Kim H, Kwon OJ, et al. Lammertsma AA, et al. Measurement of clinical and subclini-
Non-small cell lung cancer staging: efficacy comparison of inte- cal tumour response using [18F]-fluorodeoxyglucose and positron
grated PET/CT versus 3.0-T whole-body MR imaging. Radiology. emission tomography: review and 1999 EORTC recommenda-
2008;248:632–42. tions. European organization for research and treatment of cancer
59. Plathow C, Aschoff P, Lichy MP, Eschmann S, Hehr T, Brink I, et al. (EORTC) PET study group. Eur J Cancer. 1999;35:1773–82.
Positron emission tomography/computed tomography and whole- 75. Aide N, Fanti S. Therapy monitoring with PET in cancer patients:
body magnetic resonance imaging in staging of advanced non small achievements, opportunities and challenges ahead for the PET
cell lung cancer: initial results. Investig Radiol. 2008;43:290–7. community. Eur J Nucl Med Mol Imaging. 2017;44(Suppl 1):S1–3.
60. Miller AB, Hoogstraten B, Staquet M, Winkler A. Reporting results 76. Pinker K, Riedl C, Weber WA. Evaluating tumor response with
of cancer treatment. Cancer. 1981;47:207–14. FDG PET: updates on PERCIST, comparison with EORTC criteria
61. Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS, and clues to future developments. Eur J Nucl Med Mol Imaging.
Rubinstein L, et al. New guidelines to evaluate the response to 2017;44(Suppl 1):S55–66.
treatment in solid tumors. European Organization for Research 77. WHO handbook for reporting results of cancer treatment. Geneva
and Treatment of Cancer, National Cancer Institute of the United (Switzerland): World Health Organization; 1979.
78. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent 84. Robinson BW, Musk AW, Lake RA. Malignant mesothelioma.

D, Ford R, et al. New response evaluation criteria in solid Lancet. 2005;366:397–408.
tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 85. Lopci E, Zucali PA, Ceresoli GL, Perrino M, Giordano L,

2009;45:228–47. Gianoncelli L, et al. Quantitative analyses at baseline and interim
79. Van den Abbeele AD, Badawi RD. Use of positron emission tomog- PET evaluation for response assessment and outcome definition in
raphy in oncology and its potential role to assess response to ima- patients with malignant pleural mesothelioma. Eur J Nucl Med Mol
tinib mesylate therapy in gastrointestinal stromal tumors (GISTs). Imaging. 2015;42:667–75.
Eur J Cancer. 2002;38(Suppl 5):S60–5. 86. Sugarbaker DJ, Flores RM, Jaklitsch MT, Richards WG, Strauss
80. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to GM, Corson JM, et al. Resection margins, extrapleural nodal status,
PERCIST: evolving considerations for PET response criteria in and cell type determine postoperative long-term survival in trimo-
solid tumors. J Nucl Med. 2009;50(Suppl 1):122S–50S. dality therapy of malignant pleural mesothelioma: results in 183
81. Min SJ, Jang HJ, Kim JH. Comparison of the RECIST and
patients. J Thorac Cardiovasc Surg. 1999;117:54–65.. discussion
PERCIST criteria in solid tumors: a pooled analysis and review. 63–5
Oncotarget. 2016;7:27848–54. 87. Byrne MJ, Nowak AK. Modified RECIST criteria for assess-

82. Ding Q, Cheng X, Yang L, Zhang Q, Chen J, Li T, et al. PET/CT ment of response in malignant pleural mesothelioma. Ann Oncol.
evaluation of response to chemotherapy in non-small cell lung 2004;15:257–60.
cancer: PET response criteria in solid tumors (PERCIST) versus 88. Kitami A, Sano F, Ohashi S, Suzuki K, Uematsu S, Suzuki T,
response evaluation criteria in solid tumors (RECIST). J Thorac Kadokura M. The usefulness of positron emission tomography
Dis. 2014;6:677–83. findings in the management of anterior mediastinal tumors. Ann
83. Frederickson J, Callahan J, Funke R, Sanabria S, Weber W, de Thorac Cardiovasc Surg. 2017;23:26–30.
Crespigny A, et al. Utility of FDG-PET in immunotherapy: results 89. Segreto S, Fonti R, Ottaviano M, Pellegrino S, Pace L, Damiano V,
from a phase II study of NSCLC patients undergoing therapy with et al. Evaluation of metabolic response with 18F-FDG PET-CT in
the PD-L1 inhibitor, atezolizumab (MPDL3280A). J Nucl Med. patients with advanced or recurrent thymic epithelial tumors. Cancer
2016;57(Suppl 2):134. Imaging. 2017;17:10. https://doi.org/10.1186/s40644-017-0112-x.
Hybrid Imaging for Breast Malignancies
23
Federica Padovano, Giuliano Mariani,
and Marco Ferdeghini
Contents
23.1 Main Clinical and Molecular Features of Breast Cancer 544
23.2 Scintimammography 544
23.3 Radioguided Surgery in Breast Cancer Patients 549
23.3.1 Sentinel Lymph Node Mapping 549
23.3.2 Radioguided Occult Lesion Localization 555
Tc-Erythrocyte Multigated Radionuclide Angiography
23.4 99m 556
23.5 Whole-Body Bone Scan in Metastatic Disease 557
23.6 PET/CT 561
23.6.1 Clinical Use of [18F]FDG in Patients with Breast Cancer 561
23.6.2 Clinical Use of Non-[18F]FDG Agents in Patients with Breast Cancer 564
23.7 PET/MRI 566
References 566
Learning Objectives • Understand the rationale of employing scintimammography

• Acquire knowledge of the main features on epidemiol- in patients with equivocal mammographic lesions or to assess
ogy, presentation, genetic predisposition, and diagnosis response to treatment and early recurrence of disease.
of breast cancer. • Understand the rationale of sentinel lymph node mapping
• Distinguish the differences between conventional radiol- in patients with breast cancer.
ogy and nuclear medicine imaging. • Become familiar with the practice of radioguided occult
• Become familiar with the methodology of scintimam- lesion localization in patients with non-palpable breast
mography, an imaging technique in which radiopharma- lesions.
ceuticals are used to identify malignant lesions. • Understand the rationale of performing multigated radio-
• Become familiar with recent advances in scintimammog- nuclide angiography to detect chemotherapy-induced car-
raphy based on the development of breast-dedicated diac dysfunction.
gamma cameras, breast-specific gamma imaging (BSGI), • Become familiar with the role of whole-body bone scin-
or molecular breast imaging (MBI). tigraphy for staging of metastatic skeletal disease and for
assessing response to treatment.
• Understand the rationale and the roles of performing
F. Padovano · M. Ferdeghini (*) whole-body [18F]FDG PET/CT in patients with breast
Nuclear Medicine Unit and Department of Diagnostics and Public cancer and of positron emission mammography in patients
Health, University of Verona, Verona, Italy with breast lesions mammary gland tissues.
e-mail: marco.ferdeghini@univr.it
• Understand the rationale of using non-[18F]FDG PET
G. Mariani tracers, including those still under clinical validation.
Research and Advanced Technologies in Medicine and Surgery, • Acquire knowledge of the diagnostic performance of [18F]
University of Pisa, Pisa, Italy FDG PET/MRI in patients with breast cancer.

544 F. Padovano et al.
23.1 M
ain Clinical and Molecular Features ists play an integral part in the oncology team and deliver
of Breast Cancer personalized cancer care [4].
Recent developments in instrumentation [5] such as sin-
Breast cancer is the second leading cause of cancer deaths gle photon scintimammography and positron emission mam-
among women, accounting for approximately 570,000 mography, as well as combination of different imaging
deaths worldwide in 2015. Over 1.5 million women (25% of modalities such as magnetic resonance imaging (MRI) and
all women with cancer) are diagnosed with breast cancer positron emission tomography (PET), are useful for monitor-
every year throughout the world. It is estimated that 30% of ing patients with breast cancer. New biochemical biomarkers
all new cancer cases (252,710) diagnosed in women in the such as proteins, DNAs, mRNAs, and microRNAs could be
United States in 2017 are breast cancer [1]. employed for diagnosis and as a guide for personalized ther-
Many factors such as genetic predisposition, prior per- apy protocols for patients with breast cancer [6].
sonal or familial history of breast or other cancers, and
endogenous (lifetime cycle of estrogen levels) and exoge-
nous (hormone replacement therapy) hormone levels predis- Key Learning Points
pose women to a higher lifetime risk. Age of greatest risk is • Breast cancer is the second leading cause of cancer
between 40 and 59 years old. Modifiable environmental and deaths among women. Genetic predisposition, prior
lifestyle factors such as excessive alcohol consumption, obe- personal or familial history of breast or other can-
sity, and physical inactivity also contribute to increased risk cers, and endogenous and exogenous hormone lev-
of breast cancer [2]. els predispose women to a higher lifetime risk.
Genomics has improved our understanding of breast can- • Genomics has improved our understanding of
cer biology and revealed different intrinsic molecular sub- breast cancer biology and revealed different intrin-
types: luminal A (reflecting the histological phenotype: ER+, sic molecular subtypes that are important for select-
PR+, HER2−, Ki67−), luminal B (ER+, PR+, HER+/−, Ki67+), ing optimal therapy.
HER2 (ER−, PR−, HER2+), and basal-like subtype (ER−, • Imaging plays a primary role in the management of
PR−, HER2−, or triple-negative cancer). These subtypes are breast cancer, for staging and during follow-up, to
important for selecting optimal therapy. In particular, in the detect local and systemic recurrences.
treatment of the hormone receptor (HR)-positive subtypes • Combination of different imaging modalities play
(those that are positive for ER and/or PR), endocrine therapy, an integral part in the oncology team to deliver per-
including aromatase inhibitors or selective estrogen recep- sonalized cancer care.
tors mediators such as tamoxifen, plays an important role.
HER2-overexpressing tumors are generally treated with
HER2-targeting drugs such as trastuzumab and pertuzumab,
whereas triple-negative breast cancer (TNBC, largely reflect- 23.2 Scintimammography
ing the basal-like subtype) is mostly treated with standard
cytotoxic therapies. This imaging modality provides noninvasive in vivo charac-
DNA sequencing is not only useful in guiding therapy but terization for distinguishing malignant from benign pro-
may also be useful in preventing breast cancer. If a woman cesses and is best used in clinical scenarios where diagnostic
has a family history of breast cancer, it is wise to screen for performance of the mainstay anatomic modalities, mam-
hereditary cancer susceptibility genes such as BRCA1 or mography and ultrasound, is limited [7, 8]. Targeted tracer
BRCA2. The risk of breast cancer could then be evaluated uptake improves diagnostic accuracy, especially in patients
based on the screening results, and personalized prevention with dense breasts, thus reducing the need for invasive pro-
advice could be offered [3]. cedures and biopsies [9, 10].
Imaging plays a pivotal role in the management of breast To date, 99mTc-Sestamibi is the most widely used non-
cancer, both at the time of diagnosis for staging and treat- PET radiopharmaceutical for depicting breast cancer.
ment planning and during follow-up, to detect local and sys- Although originally introduced as a myocardial perfusion
temic recurrences. Breast cancer most frequently agent [11], it has been employed as a tumor-seeking agent
metastasizes to the lymph nodes, bone, liver, lung, and since 1994 for breast malignancies [12, 13].
brain. Most breast cancers have a propensity for bone metas- Whereas the recommended administered activity range
tases. Lobular breast cancer, on the other hand, has a pro- was 740–1100 MBq (20–30 mCi) when using standard
pensity for gastrointestinal and genitourinary metastases, gamma cameras, much lower activities (less than 370 MBq)
and ER-/PR-negative breast cancers metastasize to the are administered when using current state-of-the-art breast-
brain. Knowledge of the correlation between molecular phe- dedicated gamma cameras. Thus, the effective radiation dose
notype and metastatic pattern can help the imaging special- (a summed entire body exposure estimate that is weighted
23 Hybrid Imaging for Breast Malignancies 545
based on tissue radiosensitivity) declines from 5.9–9.4 mSv

estimated for the traditional procedure to even less than
2 mSv for the current imaging procedure, compared to
0.44 mSv for digital mammography [14]; nevertheless, the
radiation dose to the breast tissue is estimated at 0.07 mSv
for scintimammography with breast-dedicated instrumenta-
tion versus 0.44 mSv for each projection in diagnostic digital
mammography.
99m
Tc-Sestamibi accumulates principally within the mito-
chondria, and its diagnostic value is based on the increased
vascularity and greater cytoplasmic mitochondrial density in
breast cancer cells than in normal parenchyma [15–19].
Tumor efflux of 99mTc-Sestamibi has been correlated with
cellular expression of P-glycoprotein, a protective trans-
membrane protein pump found in cells that overexpress the
multidrug-resistant gene [16]. These findings are clinically
relevant as 99mTc-Sestamibi can provide the possibility to
predict in vivo the efficacy of anticancer drugs [16, 19]. In
fact, faster rates of 99mTc-Sestamibi efflux have been associ-
ated with overexpression of the P-glycoprotein [16].
Uptake of 99mTc-Sestamibi may be nonspecific, e.g., in
sites of prior surgical intervention or inflammation, as
increased perfusion of these sites may result in false-positive
findings [15, 18, 20–22]. False-positive uptake has also been
associated with benign diseases such as hyper-proliferative
breast disease and atypical hyperplasia [23].
For breast functional imaging using 99mTc-Sestamibi, sev-
Fig. 23.1 Example of a molecular breast imaging (MBI) device, the
eral modalities, such as conventional scintimammography dual-head CdZnTe detector GE Discovery NM 750b gamma camera,
(SMG), breast-specific gamma imaging (BSGI), molecular manufactured by GE Healthcare. The useful field of view of each detec-
breast imaging (MBI), and SPECT/CT, have been validated tor is 16 × 240 mm, and energy resolution at 140 keV is 6.5% (NEMA
[22, 24–27]. The term BSGI is usually employed to identify standards)
first-generation breast-dedicated imaging devices based on a
single scintillation crystal detector system, while the term Furthermore, images are acquired replicating the same orien-
MBI identifies modern breast-dedicated systems based on tations as during mammography, usually obtaining two
dual-head solid state detectors that allow superior quality 10-min images in the cranio-caudal and mediolateral oblique
imaging and greater counting efficiency (therefore resulting views per breast [15]; both single-head and dual-head breast-
in much lower injected 99mTc-Sestamibi activities than dedicated gamma cameras are commercially available (see
BSGI). Fig. 23.1).
For scintimammography with a standard gamma camera, Size and palpability of the breast lesion(s) is very impor-
planar (anterior, lateral, oblique) views are obtained 5–15 min tant for both diagnosis and prognosis, as small non-palpable
postinjection [7, 9]. Lateral and oblique view images are lesions often indicate early disease. Characterizing a non-
then acquired with the patient lying prone with hanging palpable and mammographically inconclusive lesion as
breasts, while supine positioning is employed for anterior, 99m
Tc-Sestamibi-negative (therefore metabolically benign)
oblique, and SPECT tomographic acquisitions. Image acqui- could obviate the need for a difficult biopsy procedure and
sition in both positions is preferred. Prone positioning better instead support the strategy of clinical observation alone.
separates breast tissue from high radiopharmaceutical uptake Conversely, given the high specificity of scintimammogra-
in the myocardium and liver, thus improving visualization of phy, a positive scintigraphic finding would recommend fur-
breast activity by reducing photon scatter and improving ther investigation with either imaging and/or an invasive
image contrast. On the other hand, supine positioning evaluation [8] (see Figs. 23.2 and 23.3).
improves visualization of primary lesion and internal mam- Several studies have confirmed that dedicated gamma
mary or axillae involvement. camera design combined with breast positioning during
Mild breast compression is applied during BSGI/MBI, in BSGI/MBI provides better detection of subcentimeter and
a similar manner as performed during X-ray mammography. non-palpable lesions compared to conventional scintimam-
Fig. 23.2 Molecular breast imaging (MBI) with 99mTc-Sestamibi in a gone unnoticed on mammography and ultrasound and was later detected
52-year-old woman in whom X-ray mammography and ultrasound by second-look, highly focused ultrasound (lower row) as an 8-mm
(upper row) had detected a highly suspicious breast lesion a small lesion (indicated by white squares). Histology confirmed that both
lesion, which was confirmed as a distinct focus on increased 99mTc- lesions were invasive breast carcinomas (images provided by courtesy
Sestamibi uptake (indicated by white circles). Moreover, MBI identi- of Dr. Serena Chiacchio, Regional Center of Nuclear Medicine,
fied an additional, deep-located 99mTc-Sestamibi-avid lesion that had University Hospital of Pisa, Pisa, Italy)
mography [27–31]. A systematic review and meta-analysis lesion-to-non-lesion ratio was associated with infiltration
published in 2013 reported overall sensitivity and specificity degree (P = 0.005), axillary lymph node status (P = 0.029),
of BSGI equal to 95% (95% CI: 93–96%) and 80% (95% CI: and tumor size (P = 0.002) [33].
78–82%), respectively, with an area under the summary Although interest in 99mTc-Sestamibi imaging is increas-
ROC curve equal to 0.9552. On the other hand, pooled sensi- ing owing to the increasing use of MBI devices in breast can-
tivity and specificity for subcentimeter lesions were 84% cer clinics, its performance for therapy monitoring is not yet
(95% CI, 80–88%) and 88% (95% CI, 81–92%), respectively clear [34]. In this regard, 99m Tc-Sestamibi scintimammogra-
[32]. A more recent study where visual analysis was associ- phy is certainly an attractive imaging modality to assess
ated with semiquantitative evaluation of the lesion-to-non- early response to neoadjuvant chemotherapy (NAC), thus
lesion 99mTc-Sestamibi uptake ratio reports 100% sensitivity possibly guiding patients’ management by assuring continu-
in patients with tumor size >2 cm, decreasing to 89.1% for ation of therapy in those who respond and instituting alterna-
tumor size ≤2 cm but still significantly higher than sensitiv- tive therapies in those who do not respond. Although based
ity of both ultrasound (84.2%, with P = 0.022) and mam- on limited experience, 99mTc-Sestamibi imaging performed
mography (84.5%, with P = 0.037); moreover, the during NAC appears to be highly sensitive to predict non-
Fig. 23.3 Molecular breast imaging (MBI) with 99mTc-Sestamibi the contralateral breast identified the lesion in the right breast (white
(upper left panel) in a 78-year-old woman with large-size breasts and squares in lower row). Histology confirmed invasive breast carcinoma
bilateral cancer. Only the lesion in the left breast (white circles in upper in the left breast and ductal carcinoma in situ (4 cm in size) in the right
row) had been detected by X-ray mammography and ultrasound and breast (images provided by courtesy of Dr. Serena Chiacchio, Regional
was highly 99mTc-Sestamibi-avid. Second-look re-evaluation after dis- Center of Nuclear Medicine, University Hospital of Pisa, Pisa, Italy)
covery of the additional focus of increased 99mTc-Sestamibi uptake in
responsiveness to treatment [35]. However, its relatively low overall availability and adequacy of integrated breast-
specificity suggests that a combination of other imaging dedicated imaging facilities.
modalities would still be needed [36]. It should also be noted that, in general, scintimammog-
The scenario of scintimammography with breast- raphy does not play a definite role in the phase of screen-
dedicated imaging devices is evolving at a rapid pace, so that ing for breast cancer, except in limited situations such as,
indications for BSGI/MBI are also continuously evolving. e.g., personal history of breast cancer, strong family his-
Although official guidelines issued by the Society of Nuclear tory of breast cancer and BRCA-positive individuals (i.e.,
Medicine and Molecular Imaging date back to 2010 [37], a in individuals with >20% lifetime risk for breast cancer),
number of indications listed at that time for this imaging pro- and augmented breast (where the implants, free silicone,
cedure are still valid. Nevertheless, it should be emphasized or paraffin injections compromise interpretation of the
that indications for BSGI/MBI may vary greatly among dif- X-ray mammogram). Continuing technological advances
ferent institutions according to local policies/protocols and (in particular the evolution from BSGI to MBI) justify
periodical revisions on the role of imaging with 99mTc-Ses- tion in the amount of superimposed breast tissue in each
tamibi for screening purposes [38]. section, thereby improving lesion conspicuity compared
Whereas some indications have better been established in with two-dimensional digital mammography while provid-
case of indeterminate or inconclusive imaging on ultrasound ing comparable in-section spatial resolution. Furthermore,
and digital mammography, in such instances, BSGI/MBI can tomosynthesis can be acquired with only mild compression
be used to: of the breast, thereby improving subject comfort during the
examination [40].
• Evaluate nipple discharge in patients with either nor- The availability of large-area, high-spatial-resolution
mal or abnormal mammography or sonographic find- digital X-ray detectors capable of rapid readout has
ings, with or without successful/unsuccessful contrast prompted manufacturers of mammography equipment to
ductography. develop commercial digital breast tomosynthesis systems.
• Evaluate lesions for patient reassurance in case of Although, X-ray tomosynthesis is likely to improve sensi-
BIRADS-3 pattern. tivity for small lesions when compared with planar digital
• Evaluate palpable or non-palpable breast lesions identi- mammography, it still provides only anatomic informa-
fied by other imaging techniques. tion [41].
• Evaluate palpable abnormalities not demonstrated on While BSGI/MBI yields a single 2D projection image of
mammography or ultrasound. 99m
Tc-Sestamibi distribution in the breast, its combination
• Evaluate multiple masses demonstrated on breast with tomosynthesis images (the so-called molecular breast
imaging. tomosynthesis, or MBT) maps tracer distribution in 3D [42,
• Aid in targeting biopsy to a suspicious lesion detected by 43]. When added to digital breast tomosynthesis in a pilot
other imaging techniques. clinical study, MBT has demonstrated significant improve-
• Evaluate diffuse or multiple clusters of microcalcifications. ments in specificity and positive predictive value compared
• Evaluate the breasts in patients with axillary lymph node to tomosynthesis alone [39].
metastases with an unknown primary.
• Evaluate unexplained architectural distortion.
• Evaluate a suggestive mammographic finding seen on one Key Learning Points
view only. • Scintimammography provides noninvasive in vivo
• Increase specificity by evaluating contrast-enhancing characterization for distinguishing malignant from
areas seen on MRI. benign processes when diagnostic performance of
• When MRI is diagnostically indicated but not possible the mainstay anatomic modalities, mammography
(e.g., claustrophobia, pacemaker or pump, risk of nephro- and ultrasound, is limited.
genic systemic response to gadolinium, large body habi- • 99mTc-Sestamibi is the most widely used non-PET
tus, breasts too large for the breast coil, etc.). radiopharmaceutical for depicting breast cancer. It
also can provide the possibility to predict in vivo
On the other hand, in patients with newly diagnosed the efficacy of anticancer drugs.
breast cancer, scintimammography can be used to: • Images are acquired replicating the same views/
projections as during mammography, usually
• Evaluate the extent of disease during initial staging. obtaining two 10-min images per breast, in the
• Detect multicentric, multifocal, or bilateral disease. cranio-caudal and mediolateral oblique views,
• Assess response to neoadjuvant chemotherapy. respectively.
• Characterizing a non-palpable and mammo-
Finally, in the post primary treatment setting, scintimam- graphically inconclusive lesion as 99mTc-Ses-
mography can be used to evaluate for residual disease if mar- tamibi-negative could obviate the need for a
gins post lumpectomy are positive or evaluate for recurrent difficult biopsy procedure. Conversely, a positive
cancer in a tissue flap (e.g., reconstruction using the trans- scintigraphic finding would recommend further
verse rectus abdominis). investigation with either imaging and/or an inva-
The recently developed imaging technique of digital sive evaluation.
X-ray breast tomosynthesis is attracting interest for the • BSGI/MBI yields a single 2D projection image of
increased potential of early detection of breast tumors [39]. 99m
Tc-Sestamibi distribution in the breast. Its com-
This technique, which is based on the acquisition of mul- bination with X-ray breast tomosynthesis images
tiple projections (views) across a restricted range of view- could significantly improve specificity and positive
ing angles rather than across the entire 360° arc, yields a predictive value.
series of section images in which there is a substantial reduc-
23.3 R
adioguided Surgery in Breast Cancer of lymphatic drainage of the arm, leading to postoperative
Patients complications, such as hemorrhage, infection, seroma, axil-
lary web syndrome, chronic pain, paresthesia caused by
23.3.1 Sentinel Lymph Node Mapping intercostobrachial nerve damage, reduced range of motion
and muscle weakness on the shoulder ipsilateral to the sur-
Staging and recurrence risk of disease in patients with breast gery, and, especially, lymphedema [54, 55].
carcinoma is determined by the extent of disease found in the Planar and/or SPECT/CT sentinel lymph node imaging
regional lymph node basin. Identification of the sentinel using radiolabeled colloids plays a fundamental role in
lymph node (SLN), or first node draining the primary tumor assisting in the surgical staging of breast carcinoma patients
site, is essential for effective management of breast carci- by mapping the drainage flow of the tumor and identifying
noma patients [44–47] (see also Chaps. 16 and 46 of this the sentinel node within the axilla. Radiolabeled colloid
book). particles are imaged as they penetrate the lymph system
Selective sentinel lymph node biopsy (SLNB) in breast from interstitial spaces and are transported to the first
cancer is a widely accepted technique for predicting the state lymph node along the path of flow in the lymph channel.
of the axillary lymph nodes. The rationale of this procedure This is a simulation of what happens to tumor cells that
is based on the concept of orderly progression of tumor spread to a lymph node. A 99mTc-nanocolloidal albumin
metastasis through the lymphatic route, from lower-tier to preparation (Nanocoll) is the most commonly used radio-
upper-tier lymph nodes (Fig. 23.4). Therefore, if the first tracer for lymph node detection in Europe. More than 95%
lymph node(s) encountered along the route of lymphatic of its particles are smaller than 80 nm [56]. This radiophar-
drainage is (are) free of metastasis, then it is highly unlikely maceutical is not approved for commercial use in the
that other lymph nodes along the same route harbor metasta- United States, where the most commonly used agent for
sis. In this case, de novo axillary lymph node dissection as a lymphatic mapping is 99mTc-sulfur colloid, which is gener-
staging modality can safely be avoided; conversely, if metas- ally filtered in order to select radiolabeled particles with
tasis is found in sentinel lymph node(s), axillary lymph node size <220 nm [57]. Other lymphatic mapping agents
dissection can be considered. (including either colloidal or noncolloidal radiopharmaceu-
Compared to axillary lymph node dissection (ALND), ticals—such as the receptor-binding agent 99mTc-Tilmano-
SLNB presents less morbidity [48–53]. Surgical injuries cept) can alternatively be used across different countries
resulting from ALND cause obstruction of the primary route worldwide [58].
higher-echelon nodes
efferent
vein
primary
tumor
SLN
afferent
artery
SLN
Fig. 23.4 The sentinel lymph node (SLN) concept is based on the interconnections can exist at higher levels, with variable directions of
assumption that lymph drains from a solid tumor in an orderly way lymph flow at intermediate levels within the general pattern of overall
from lower-echelon to higher-echelon lymph nodes. The first node(s) centripetal flow. Lymph from the tumor site can drain to more than one
encountered along such pathway, the SLN(s), is(are) the most likely site lymphatic basin, each one repeating the basic pattern represented here
where tumor cells migrating through lymphatic channels remain for a single basin (reproduced with permission from: Strauss HW,
entrapped before spreading to higher-echelon lymph nodes. There can Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
be more than one SLN, even in the same lymphatic basin. Complex Pathophysiology to Clinical Applications. New York: Springer; 2017)
a b c
d e f
Fig. 23.5 Different modalities of interstitial injection of the lymphos- Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
cintigraphic agent for SLN mapping in patients with breast cancer: Oncology – From Pathophysiology to Clinical Applications. New York:
intradermal (a), subcutaneous (b), subareolar (c), periareolar (d), peri- Springer; 2017)
tumoral (e), and intratumoral (f) (reproduced with permission from:
Identification of the sentinel node is crucial to the success umes of 0.5–1.0 mL are generally injected in the peritumoral
of SLNB, and with a detection rate between 94% and 100%, approach, while smaller volumes (0.1 mL) are used for the
preoperative sentinel node imaging is ideally suited for this subdermal and the periareolar routes. It is recommended to
purpose [59]. For patients with multicentric and multifocal avoid injecting close to the axilla or to internal mammary
disease whose lymphatic drainage patterns may vary, senti- lymph nodes, as it is easier to image and locate sentinel
nel node imaging and SPECT/CT of all tumors may provide nodes if one avoids the shine-through effect from the injected
a more accurate approach to map and biopsy sentinel nodes activity masking activity accumulated in lymph nodes.
for staging [60]. Lymphatic transport of radiocolloid particles after injec-
The most widely used injection modalities are peritu- tion is influenced mostly by particle size. Adequate lymphatic
moral, periareolar, subareolar, subdermal, intradermal, intra- migration of large colloid particles is hindered by their size,
tumoral, and subtumoral [47, 61] (Fig. 23.5). The two most often resulting in either delayed visualization or non-visu-
common injection sites are periareolar (superficial) through alization of axillary sentinel lymph nodes. Conversely, par-
the richly lymphatic subareolar plexus (which generates ticles that are very small migrate rapidly, but may not remain
quick visualization of the draining channels with high target trapped in the sentinel lymph node [62–64].
count rates) or peritumoral (deep), which can trace accessory Imaging provides a map of the distribution of foci of
drainage patterns, specifically the internal mammary chain radioactivity accumulation that the surgeon would encounter
(which is seen in ∽20–30% of patients injected with this when using the handheld gamma-detecting probe (or gamma
modality). probe) to localize the SLN(s) intraoperatively. In the nuclear
Peritumoral injections are administered at two to four medicine department, a physician or technologist uses
sites around the tumor into the interstitial tissues of the images and a probe to mark the skin projection of the radio-
breast. Activities of 4.63–9.25 MBq (125–250 μCi) in vol- active node(s). This is done at two projections, so that the
a b
Fig. 23.6 Example of body contour delineated by using two flexible sponds to the injection site. Similar results can be obtained by moving
57
Co wire sources as landmarkers around the region of the body within a radioactive point-source along the body contour (reproduced with
the field of view of the gamma camera in a patient with cancer in her permission from: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés
right breast during lymphoscintigraphy with 99mTc-nanocolloidal albu- Olmos RA, Eds. Atlas of Lymphoscintigraphy and Sentinel Node
min. (a) Anterior view; (b) right lateral view. The SLN in the axilla is Mapping – A Pictorial Case-Based Approach. Milan: Springer; 2013)
indicated by the red arrow, while the major focus of activity corre-
a b
Fig. 23.7 Technique for generating the body contour by placing a 57Co migration to at least two lymph nodes in the right internal mammary
flood source, indicated by black arrows in panel (a) beneath the patient’s chain is clearly visualized, while a SLN in the right axilla is more
body, so that the γ-rays generated by the flood source are detected by faintly visualized because of attenuation by overlying tissues (better
the gamma camera head (placed above) after having been either vari- visualization of the axillary SLN is achieved by acquiring oblique
ably attenuated by the patient’s body or minimally attenuated by air views and SPECT/CT imaging) (reproduced with permission from:
outside the body contour. Panel (b) shows the resulting scintigraphic Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, Eds.
image, where the large focus of activity in the right breast corresponds Atlas of Lymphoscintigraphy and Sentinel Node Mapping – A Pictorial
to intratumoral injection of 99mTc-nanocolloidal albumin; radiocolloid Case-Based Approach. Milan: Springer; 2013)
location of the nodes can be triangulated for depth in tissue, of the gamma camera or using a flexible 57Co wire
thus making it easier for the surgeon to retrieve the radioac- source (Fig. 23.6).
tive lymph nodes during surgery. • Outlining the body contour using a 57Co flood source
Occasionally, difficulty is encountered in imaging axil- placed such that the patient is in between the flood source
lary sentinel node(s). The steps that can be taken to minimize and the gamma camera (Fig. 23.7).
such difficulty include: • Imaging from multiple views.
• Positioning the breast to minimize attenuation from the
• Outlining the body contour by moving a point source breast tissue and to minimize masking of focal uptake by
(57Co) along the area of the body under the field of view the high activity remaining at the injection site.
a b c
d e f
Fig. 23.8 Lymphoscintigraphy with 99mTc-nanocolloidal albumin in a muscle: (c) axial fused SPECT/CT section; (d) corresponding CT sec-
patient with cancer in her right breast; the radiocolloid was injected tion where the lymph node is indicated by the yellow circle; (e) right
intratumorally. In the anterior (a) and lateral (b) planar views, no radio- lateral view and (f) anterior view of 3D volume rendering for better
colloid drainage can be detected from the injection site (body contour anatomical identification (reproduced with permission from: Mariani
delineated by placing a 57Co flood source beneath the patient’s body G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of
during gamma camera acquisition). Whereas, SPECT/CT acquisition Lymphoscintigraphy and Sentinel Node Mapping – A Pictorial Case-
allowed visualization of a radioactive SLN at the border of the pectoral Based Approach. Milan: Springer; 2013)
• Marking the skin to guide the surgeon in using a gamma- tions include the additional use of a blue dye injected peri-
detecting probe to find the SLN(s). tumorally to complement with a visual guide the radioactive
• Using SPECT or SPECT/CT for visualizing and anatomi- guide for identification of nodes receiving lymph that
cally localizing sentinel nodes that may have very small drains from the tumor site. In this case, sentinel lymph
amounts of activity (Fig. 23.8). nodes located with the combined procedure can be either
“hot” or stained blue, or (more often) both. All such lymph
Figures 23.9 and 23.10 show representative lymphoscinti- nodes are excised for histology (or for molecular analysis).
graphic patterns obtained with SPECT/CT imaging in Final exploration of the surgical bed with the gamma probe
patients with early breast cancer scheduled for radioguided confirms that all lymph nodes with radioactive content
SLNB. above a certain threshold (either 10% or 20% of the hottest
In the operating room, using a handheld gamma probe lymph node retrieved) have been harvested for analysis. If
guides the surgeon to locate and retrieve lymph nodes that the retrieved sentinel lymph node(s) harbor(s) tumor
harbor radioactivity. Protocols adopted in some institu- metastasis, the procedure is completed with axillary lymph
a c
b
d
f
e
Fig. 23.9 Lymphoscintigraphy with 99mTc-nanocolloidal albumin the higher-tier lymph nodes visualized in the anterior planar view (yel-
injected through the subareolar route in a patient with cancer in her left low arrow)., without specifying their topographic location. (e) Anterior
breast. (a) Left anterior oblique planar view shows lymphatic drainage view and (f) left lateral view of 3D volume rendering displaying the
of the radiocolloid to two areas of focal uptake of the tracer (red and axillary SLNs (red arrow) and one of the higher-tier radioactive lymph
green arrows) in the left axilla. (b) Anterior planar view confirms focal nodes, located in Berg’s anatomical level 3 (yellow arrow) (reproduced
uptake (red arrow) corresponding to the two SLNs in the left axilla, and with modifications, with permission from: Mariani G, Manca G, Orsini
also shows subsequent-tier lymph nodes (yellow arrows). (c) Axial F, Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of Lymphoscintigraphy
fused SPECT/CT image at the level corresponding to the two SLNs and Sentinel Node Mapping – A Pictorial Case-Based Approach.
visualized in the left anterior oblique planar view (green and red Milan: Springer; 2013)
arrows). (d) Axial fused SPECT/CT image at the level corresponding to
a c
b d
Fig. 23.10 Lymphoscintigraphy with 99mTc-nanocolloidal albumin tional radioactive lymph node located at Berg’s anatomical level 3 (blue
injected intradermally in a patient with cancer in her right breast. The arrow) is better displayed using the window for bone (d) (reproduced
axial (a) and coronal (b) fused SPECT/CT images visualize an axillary with modifications, with permission from: Mariani G, Manca G, Orsini
SLN located at the border of the pectoral muscle (red arrow). Right F, Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of Lymphoscintigraphy
anterior oblique view of 3D volume rendering with window for soft and Sentinel Node Mapping – A Pictorial Case-Based Approach.
tissues (c) confirms anatomical location of this SLN, while an addi- Milan: Springer; 2013)
node dissection—at least according to common practice mastectomy, who previously underwent breast and/or
largely adopted in the past. In this regard, current evidence axillary surgery, or who received preoperative/neoadju-
derived from analysis of the 10-year survival demonstrates vant chemotherapy may be offered SLNB.
that, in cases selected on the basis of certain clinical and • Patients with locally advanced invasive breast cancer (T3/
laboratory biomarkers, axillary lymph node dissection can T4), inflammatory breast cancer, or ductal carcinoma in
safely be avoided even in case of one or two axillary lymph situ (when breast-conserving surgery is planned) or who
nodes containing metastases [65–69]. are pregnant should not undergo SNLB.
Indications for SLNB in patients with breast cancer are
summarized in Table 23.1, as listed in the guideline jointly
issue in 2013 by the European Association of Nuclear Key Learning Points
Medicine and the Society of Nuclear Medicine and Molecular • Planar and/or SPECT/CT sentinel lymph node
Imaging [64]. More recent recommendations issued by the imaging using radiolabeled colloids plays a funda-
American Society of Clinical Oncology [70] can be summa- mental role in assisting surgical staging of breast
rized as follows: cancer patients, by mapping drainage of lymph flow
from the tumor and identifying the sentinel node
• Patients without SLN metastasis should not undergo within the axilla.
ALND. • A 99mTc-nanocolloidal albumin preparation is the
• Most of the patients with one to two metastatic SLNs most commonly used radiotracer for lymph node
should not undergo ALND when breast-conserving sur- detection in Europe, while in the United States the
gery with whole-breast radiotherapy is planned. most commonly used agent for lymphatic mapping
• Patients with SLN metastasis who will undergo mastec- is 99mTc-sulfur colloid.
tomy should be offered ALND. • The most widely used injection modalities are peri-
• Patients with operable breast cancer and multicentric tumoral, periareolar, subareolar, subdermal, intra-
tumors, with ductal carcinoma in situ, who will undergo
The earliest technique reported in 1966 involved bent

dermal, intratumoral, and subtumoral. Imaging wire implanted under fluoroscopic control through a needle
provides a map of the distribution of radioactive placed in the occult breast lesion. The use of needles com-
foci that the surgeon would encounter when using bined with hook wire was later proposed in 1976 for the pre-
the handheld gamma-detecting probe to localize the operative marking of impalpable breast lesions. Since then a
sentinel lymph node(s) intraoperatively. large number of localization approaches have been reported;
• Compared to axillary lymph node dissection, senti- these include positioning of a needle hook wire, dye injec-
nel lymph node biopsy presents less morbidity. tion, using radiological and clinical measurements, perforated
• Patients without sentinel lymph node metastasis grids, stereotactic methods, carbon localization, and intraop-
should not undergo axillary lymph node dissection. erative ultrasonography. In addition, the use of MRI, CT, and
radioactive seed localization has been described with vari-
able success. The majority of preoperative approaches for
Table 23.1 Indications and recommendations regarding application of non-palpable breast lesion localization are associated with
sentinel lymph node biopsy in patients with breast cancer, provided that
significant failure rates and complications, resulting in only
no other diagnostic procedure has ascertained the presence of metasta-
sis in axillary lymph nodes a handful being used routinely [74].
The current standard is wire-guided localization (WGL)
Clinical condition Use of SLNB
T1 or T2 tumor Established even though it has several disadvantages, the most important
DCISa with mastectomy Established one being the considerable proportion of patients with unsat-
Male breast cancer Established isfactory resection margin; these patients require reopera-
Older age Established tion. New radioguided surgery techniques have been
Obesity Established developed, including radioguided occult lesion localization
Before neoadjuvant chemotherapy Established (ROLL) and radioactive seed localization (RSL). RSL is a
T3 or T4 tumor Controversial very promising technique. Guided by ultrasound, a small
Multicentric or multifocal tumor Controversial titanium seed containing typically 1–10 MBq of iodine-125
DCIS without mastectomy Controversialb
is placed in the center of the non-palpable breast lesion.
Palpable axillary lymph nodes Controversial
During surgery location of the seed is identified with the help
Pregnancy Controversial
Prior nononcological breast surgery Controversial of a gamma probe. The results are either equal to or superior
Prior axillary surgery Controversial to those obtained with WGL, with regard to achieving free
Prior diagnostic/excisional biopsy Controversial margins and low reoperation rates. Additionally, the RSL
After neoadjuvant chemotherapy Controversial technique is less unpleasant for the patient and more flexible
Evaluation of internal mammary chain Controversialc regarding preoperative logistics. The seed can be placed a
Inflammatory breast cancer Not recommended few days before surgery, in contrast to the wire used in WGL,
Controversial indications are those for which sentinel lymph node which has to be placed within few hours of surgery [75].
biopsy is not universally accepted or for which the evidence behind the In 1998, Luini and colleagues at the European Institute
practice is limited or entirely missing
a
Ductal carcinoma in situ of Oncology (EIO) in Milan pioneered a new technique for
b
Except for DCIS with suspected or proven microinvasion localizing non-palpable breast lesions. The technique, later
c
Deep injection (peri- or intratumoral) recommended coined radioguided occult lesion localization (ROLL), had
its origins in the radioguided surgery approach for sentinel
23.3.2 Radioguided Occult Lesion Localization lymph node localization. The original technique described
by Luini et al. involved intratumoral injection of large-size
The diagnosis of breast cancer has undergone revolutionary particles of human albumin (Macrotec®, size 10–150 μm in
changes since the introduction of routine screening pro- diameter) labeled with technetium-99 m. The injection is
grams; this has directly resulted in a substantial increase in performed under sonographic or stereotactic guidance. A
the number of diagnosed breast cancers that are clinically scintigram is performed immediately after injection
non-palpable [71]. (Fig. 23.11) to verify correct positioning of the radioactive
More than 25% of radiologically suspicious breast lesions marker and the absence of contamination along the needle
are non-palpable; accurate localization is required for diag- track and/or on the skin. In addition, an X-ray mammogram
nostic core biopsy or surgical excision. Precise localization is classically performed to check correct position of the
of non-palpable lesions is an essential step to guarantee full tracer in the center of the lesion, aided by the addition of a
cancer clearance without compromising cosmetic results. small amount of iodinated contrast medium to the radioac-
Various modalities have been tested and proposed for accu- tive preparation. The patient undergoes definitive surgery
rate intraoperative localization of non-palpable breast within 24 h of localization; during surgery, searching and
lesions; each with some advantages and risks [72, 73]. retrieving of the lesion is facilitated by the use of a gamma
a b
Fig. 23.11 Example of scintigraphic imaging prior to surgery to verify view: contour of the upper torso is outlined using a flexible wire 57Co
correct intralesional injection of the 99mTc-MAA for radioguided occult source (reproduced with permission from: De Cicco C, Trifirò G, Intra
lesion localization (ROLL) in a patient with non-palpable lesion at the M, Marotta G, Ciprian A, Frasson A, et al. Optimised nuclear medicine
upper external quadrant of the left breast. (a) Anterior view: the two method for tumour marking and sentinel node detection in occult pri-
crosses mark the position of the nipples (right and left). (b) Left lateral mary breast lesions. Eur J Nucl Med Mol Imaging. 2002;31:349–54)
probe, the same device used for radioguided SLNB. The advances have prolonged survival, several agents may
procedure of ROLL has steadily gained popularity for have serious cardiovascular side effects. Trastuzumab
intraoperative localization of non-palpable lesions due to (Herceptin, Genentech), a humanized monoclonal anti-
its minimal side effects and technical ease [76, 77]. body against the human epidermal growth factor receptor
type 2 (HER2), is one such agent that is new and success-
ful but has been associated with cardiac dysfunction [78].
Key Learning Points Of the classic cytotoxic agents, anthracyclines are well
• Precise localization of non-palpable breast lesions known for their dose-dependent acute and chronic, irrevers-
is an essential step to guarantee diagnostic core ible and progressive cardiomyopathy [79]. Additionally,
biopsy and full cancer clearance without compro- radiation therapy and chemotherapy drugs are pericardial
mising cosmetic results. irritants and as such increase the risk of pericardial effusion
• The current standard is wire-guided localization or complications, making early diagnosis critical to decrease
carbon tracing. primary or contributory associated deaths seen in ∽86% of
• Radioguided occult lesion localization following symptomatic cancer patients [80].
the intralesional injection of 99mTc-MAA is becom- The standard method to detect cardiotoxicity and peri-
ing the new surgical standard. cardial effusions is acquiring ejection fraction measure-
• Radioguided seed localization based on placement ments performed prior to and post-chemotherapy treatment
of a small titanium seed containing 125I is an alterna- in breast cancer patients. Serial evaluation allows clini-
tive to the use of 99mTc-MAA for intraoperative cians to monitor the patient’s cardiac response to treat-
radioguidance with the use of a handheld gamma ment, thus reducing the risk of chemotherapy-induced
probe. comorbidities [81].
Over the years, several imaging modalities have been
developed that vary considerably in precision, ease of use,
availability, and costs. The left ventricular ejection fraction
(LVEF) may be measured by planar multigated radionu-
23.4 99mTc-Erythrocyte Multigated clide angiography (MUGA), quantitative gated blood-pool
Radionuclide Angiography SPECT (GBPS), two- and three-dimensional echocardiog-
raphy, radiographic contrast angiography, or cardiac
The development of new antitumor agents, especially MRI. The potential of newer echocardiographic methods,
molecular targeting agents, has significantly improved such as real-time three-dimensional echocardiography, tis-
the treatment options for cancer patients. Although these sue Doppler imaging with myocardial strain, and strain rate
imaging, may enhance early detection of cardiotoxicity

[82]. Finally, minimally invasive methods, such as mea- Key Learning Points
surement of blood and serum markers of endothelial dam- • Several antitumor agents have serious cardiovascu-
age and biochemical cardiac markers, are also promising lar side effects. The standard method to detect car-
[83, 84]. diotoxicity is based on measuring the ejection
Serial MUGA is the gold standard for assessing LVEF fraction prior to and post-chemotherapy treatment
and is still regarded as the best noninvasive method for in breast cancer patients.
identifying subclinical left ventricular dysfunction in • Serial planar multigated radionuclide angiography
adult patients treated with potentially cardiotoxic agent (MUGA) is the gold standard for assessing the left
[85]. MUGA is a noninvasive technique that makes use of ventricle ejection fraction and is still regarded as the
99m
Tc-
erythrocyte labeling, applying an in vivo, an best noninvasive method for identifying subclinical
in vitro, or a combined in vivo/in vitro approach, as left ventricular dysfunction.
described in detail in the guidelines for radionuclide • The labeling technique makes use of the binding of
imaging of cardiac function issued jointly by the European the radionuclide 99mTc to erythrocytes and enables
Association of Nuclear Medicine and the European the cardiac blood pool to be visualized with a gamma
Society of Cardiology [86]. Each of the three approaches camera to permit precise quantification of left ven-
has its advantages and disadvantages, which should be tricular volume throughout an average cardiac cycle.
noted, because optimal red blood cell labeling is a prereq-
uisite to accurate LVEF calculation. In short, the labeling
technique makes use of the binding of the radionuclide
99m
Tc to erythrocytes facilitated by stannous chloride, 23.5 W
hole-Body Bone Scan in Metastatic
which causes the 99mTc to bind to the beta chain of hemo- Disease
globin, thus trapping the 99mTc in the red cells. This tech-
nique enables the cardiac blood pool to be visualized with Metastases from breast carcinomas predominantly affect the
a gamma camera. bones. In fact, bone metastases represent the most common
Typically, gated planar gamma camera images in the complication of breast cancer, occurring in up to 70% of
anterior and “best septal” left anterior oblique (LAO) patients with advanced disease [88].
view provide sufficient data to evaluate the global and Current clinical guidelines from the National
regional left ventricular function. Adding gated SPECT Comprehensive Cancer Network (NCCN) recommend consid-
allows a more complete assessment of right ventricular eration of additional imaging with X-ray computed tomogra-
function and left atrial size in addition to LV function. phy (CT) and skeletal scintigraphy in patients with stage I–IIIC
Gating means that the heart cycle is split into 16 or 32 invasive breast cancer. CT, skeletal scintigraphy, and magnetic
time bins in synchrony with the patient’s cardiac cycle, resonance imaging (MRI) are also recommended for stage IV
like shining a strobe light on a moving object. The final invasive breast cancer, whereas PET/CT with 2-deoxy-2-[18F]
result is a series of images of the heart, one at each stage fluoro-d-glucose ([18F]FDG) has been proposed as an optional
of the cardiac cycle. Fourier analysis provides a func- adjunct for clinically high-risk cases [89–91].
tional image (a parametric map of sequential contractions, Cortical bone is a thin compact layer comprising 80% of the
or phase imaging), which permits precise quantification of skeleton. It surrounds the trabecular bone, which also encom-
left ventricular volume and dyssynchrony, with high passes the bone marrow [92]. Constant remodeling of bone
reproducibility. A major strength of this technique is that maintains a dynamic balance between bone resorption (by
the quantitative computation of ejection fraction and osteoclasts) and bone formation (by osteoblasts). Metastatic
chamber volumes does not depend on mathematic assump- bone lesions may be osteolytic, osteoblastic, or mixed. The
tions on ventricular geometry as is required for echocar- principal exam used to diagnose bone metastases is bone scin-
diography [82]. tigraphy, traditionally performed with gamma camera imaging
Established guidelines recommend that the cardiotoxic following injection of a 99mTc-labeled diphosphonate, although
agent not be started if baseline LVEF is significantly it should be noted that PET/CT with 18F-fluoride is more sensi-
impaired and advised discontinuation in patients with more tive than conventional bone scintigraphy. Nevertheless, the
than a 10% decrease in LVEF, resulting in LVEF values of total number of 99mTc-based bone scans worldwide still largely
less than 50% [87]. Early detection of a decrease in LVEF supersedes the number of 18F-fluoride PET/CT scans.
on serial imaging allows timely modification of either the Whole-body bone scanning (WBS) is a relatively inexpen-
dose of a cardiotoxic agent, or the selection of a different sive and noninvasive examination that is recommended for
drug [79, 80]. staging of disease burden and follow-up evaluations in symp-
Oct 2010 Sept 2014 Sept 2015 Sept 2016
Fig. 23.12 Whole-body scans with 99mTc-MDP sequentially acquired showed foci with abnormally increased 99mTc-MDP uptake in the spine,
in a woman who had been diagnosed in 2000 (at the age of 43 years) ribs, pelvis, and proximal portions of both femurs. Bone scintigraphy
with invasive breast cancer, successfully treated with surgery and adju- performed 1 year later (September 2015, when the tumor-associated
vant therapy. A second breast cancer was diagnosed in 2009 (at the age antigen C15.3 was only mildly increased at 37 U/mL) showed a virtu-
of 52 years) with metastasis lesions in the sacrum. The bone scan ally unchanged pattern. Whereas, in September 2016 (when serum
obtained in October 2010, several months after completion of external CA15.3 was 54 U/mL) scintigraphy with 99mTc-MDP showed clear pro-
beam radiation therapy on the sacrum, showed a virtually normal pat- gression of metastatic disease, now extensively involving the spine,
tern of 99mTc-MDP distribution. While undergoing maintenance therapy ribcage, and long bones (images provided by courtesy of Dr. Gayane
with the estrogen-receptor antagonist fulvestrant and the anti-RANKL Aghakhanyan, Post-graduate Specialty School of Nuclear Medicine,
antibody denosumab, mild increase of serum CEA was noticed (6.6 ng/ Regional Center of Nuclear Medicine, University of Pisa, Pisa, Italy)
mL) about 4 years later; the bone scan performed on September 2014
tomatic or high-risk patients (Fig. 23.12). Although bone scin- evaluation of the entire skeleton with a low radiation burden
tigraphy detects mainly osteoblastic bone lesions, it is much to patients. Skeletal scintigraphy allows visualization of
more sensitive and entails a much lower radiation dose than uptake in regions of increased bone turnover and osteoblast
X-ray-based imaging for whole-body investigations [93, 94]. activity and a resulting increase in blood perfusion. Patients
In the absence of visceral involvement, survival of breast normally undergo whole-body planar bone scintigraphy in
cancer patients with skeletal metastases may be long. For the anterior and posterior views 3 h after intravenous injec-
example, in one retrospective series of 859 patients, the tion of 99mTc-diphosphonates (740 MBq); SPECT (or better
median overall survival from the time of diagnosis of bone SPECT/CT—if available) acquisitions are usually performed
metastasis was 2.2 years in patients with bone only metasta- at selected spots that appear inconclusive/equivocal at planar
ses versus 5.5 months in those with coexistent liver disease imaging [97]. In this regard, emerging trends favor acquisi-
(P < 0.001) [89]. However, half of the patients diagnosed tion of whole-body SPECT/CT de novo, without prior planar
with bone metastases would experience skeletal-related imaging, for increased diagnostic accuracy. The adminis-
events (SREs) and require palliative radiotherapy or sur- tered activity can significantly be reduced when using dedi-
gery or both to treat pain and prevent loss of function [95]. cated softwares for fast acquisition protocols [98].
Furthermore, patients with initial bone-only disease are at Despite the potential false positive cases due to benign
the greatest risk of SREs, with rates of 80% reported in this causes of increased bone turnover (such as fracture—as shown
population. Preventing morbidity from skeletal metastases in Fig. 23.15—or degenerative changes) and false negative
constitutes therefore a relevant clinical challenge. cases (as seen in avascular lesions, in the presence of rapidly
Management of skeletal disease in patients with breast growing pure osteolytic metastases with no reactive increased
cancer increasingly incorporates systemic therapy options osteoblastic activity, or in lesions with low bone turnover),
such as bisphosphonates, endocrine therapy (Figs. 23.13 and whole-body scintigraphy with 99mTc-diphosphonates has diag-
23.14), and chemotherapy, all given with the goal of reduc- nostic sensitivity of about 87–88% [99] and satisfactory speci-
ing morbidity and preventing SREs, as well as improving ficity, especially if combined with suitable correlative imaging
overall survival [96]. (e.g., CT or MRI). Full evaluation typically requires additional
Bone scintigraphy with 99mTc-diphosphonates images the structural evaluation to guide clinical management using CT
osteoblastic response to bone destruction caused by tumor and/or MRI, depending on lesion location and clinical context.
cells and, unlike X-ray cross-sectional imaging, permits Although this multimodality approach is effective at establish-
Fig. 23.13 Whole-body
scans with 99mTc-MDP
sequentially acquired in a
woman who in February 2014
(at the age of 39 years) had
been treated surgically because
of invasive breast cancer. The
baseline scan performed in
March 2014 showed multiple
foci with abnormally increased
99m
Tc-MDP uptake in the skull,
spine, ribs, sternum, left
clavicle, scapula, and pelvis.
Hormonal therapy was started,
and the bone scan performed
for reassessment of the patient
about 2 years later (in July
2016) showed virtually
complete recovery to normal
of the pattern of 99mTc-MDP
distribution (images provided
by courtesy of Dr. Gayane
Aghakhanyan, Post-graduate
Specialty School of Nuclear
Medicine, Regional Center of
Nuclear Medicine, University
of Pisa, Pisa, Italy)
March 2014 July 2016
March 2016 Jan 2017
Fig. 23.14 Whole-body scans with 99mTc-MDP sequentially acquired The bone scan performed in January 2017 for reassessment of the
in a woman who in 2015 (at the age of 59 years) had been treated surgi- patient showed multiple foci of increased 99mTc-MDP uptake in the
cally because of invasive breast cancer. The baseline scintigraphy per- spine, ribcage, left humeral head, and pelvis, indicating progression of
formed in March 2016 showed a virtually normal pattern of 99mTc-MDP disease (images provided by courtesy of Dr. Gayane Aghakhanyan,
distribution, except for mildly increased tracer uptake at the second left Post-graduate Specialty School of Nuclear Medicine, Regional Center
costovertebral joint. Despite treatment with hormonal therapy and zole- of Nuclear Medicine, University of Pisa, Pisa, Italy)
dronic acid, progressively rising serum CA15.3 levels were noticed.
Jan 2016 Dec 2017
Fig. 23.15 Whole-body scans with 99mTc-MDP sequentially acquired in linearly at different levels of the ribs and costovertebral joints. The bone
a woman who had been treated surgically because of invasive breast can- scan performed in December 2017, about 2 years after the traumatic
cer in October 2015, at the age of 78 years; shortly after surgery (in event, showed virtually complete recovery to normal of the pattern of
December 2015), the patient fell on the floor. The bone scan performed in 99m
Tc-MDP distribution (images provided by courtesy of Dr. Gayane
January 2016 showed a pattern of 99mTc-MDP distribution consistent with Aghakhanyan, Post-graduate Specialty School of Nuclear Medicine,
the recent trauma, with multiple foci of increased tracer uptake arranged Regional Center of Nuclear Medicine, University of Pisa, Pisa, Italy)
ing fracture risk or confirming spinal cord compression, its frequently excluded from research protocols investigating
ability to differentiate between active and inactive disease in novel therapeutic agents. The system in most common usage,
those undergoing treatment remains problematic. Response Evaluation Criteria in Solid Tumors (RECIST)
Using serial bone scans, an initial increase in osteoblas- 1.1, primarily determines response by the change in the sum
tic activity in lesions responding favorably to therapy can of the measurements of the greatest longitudinal dimension
mimic disease progression on bone scans by stimulating a of each target lesion. Although use of bone scan or PET, or
flare response such that known lesions appear more active plain X-rays for diagnosis or complete response evaluations
and previously undetectable lesions become visible [100]. is permitted within RECIST criteria, only lytic bone metas-
Furthermore, in the context of predominantly lytic disease, tases with soft tissue masses >10 mm are considered to be
which accounts for the majority of skeletal lesions in breast measurable for response evaluation [101].
cancer, an apparent reduction in lesion activity may be due The International Union Against Cancer and World Health
to progression without an osteoblastic response. Therefore, Organization (WHO) bone response criteria are based on
there is a time lag of up to 6 months from the start of therapy radiographic evaluation of bone lesions for lesion resolution
to reliable assessment of bone scan response. This complexity or sclerosis. However, the time lag of 6–12 months required
of assessment has resulted in skeletal disease being generally for reliable radiographic evidence of response and difficul-
regarded as nonmeasurable for response evaluation within ties with evaluation of initially sclerotic lesions constrain
clinical trials with novel therapeutic agents. Detailed correla- their clinical utility. Both these systems predate modern
tion with plain X-ray films or CT, where sclerosis is generally cross-sectional imaging modalities, and more recently the
regarded as indicative of response, may improve the ability to MD Anderson Cancer Center has reported superior correla-
differentiate between responders and nonresponders. tion with progression-free survival in comparison with
Response criteria are the standard by which therapeutic WHO, using criteria which include quantitative and qualita-
efficacy of novel agents are judged in clinical trials. However, tive indices from CT or MRI assessments or both of the
radiological reliance on size change for evaluation of behavior of bone metastases. However, these criteria lack
response has meant that patients with bone-only disease are prospective validation to date [102].
23.6 PET/CT
Key Learning Points
• Bone metastases represent the most common com- 23.6.1 Clinical Use of [18F]FDG in Patients
plication of breast cancer. The most established with Breast Cancer
imaging modality to diagnose bone metastases is
bone scintigraphy, traditionally performed with [18F]FDG PET/CT has been widely utilized for the diagnosis,
gamma-camera imaging following administration staging, and restaging of different types of cancer. For breast
of a 99mTc-labeled diphosphonate. cancer patients, [18F]FDG PET/CT has been shown to play a
• Whole-body bone scanning (WBS) is a relatively role in the detection of distant metastases and tumor recurrence,
inexpensive and noninvasive examination that is as well as in the evaluation of treatment responses [103, 104]
recommended for staging of the skeletal disease (Fig. 23.16). Nonetheless, PET/CT may not adequately resolve
burden and follow-up evaluations in symptomatic small primary breast lesions or axillary nodal regions. Currently
or high-risk patients. under investigation to address this limitation, dedicated positron
• Patients undergo whole-body planar bone scintigra- emission mammography (PEM) devices have been designed to
phy in the anterior and posterior views 3 h after improve detection of small primary lesions by maximizing the
intravenous injection of 99mTc-diphosphonates. system spatial resolution [105, 106] (Fig. 23.17).
• WBS has diagnostic sensitivity of about 87–88% Routine use of [18F]FDG PET/CT is currently not recom-
and satisfactory specificity, although its ability to mended for primary breast cancer staging [107]. To limit its
distinguish active from inactive disease in patients overuse in primary staging and follow-up, [18F]FDG-PET is
undergoing treatment remains problematic. included as two of the five “don’ts” of the American Society
Fig. 23.16 PET/CT with [18F]FDG performed for staging in a 78-year- while the axial images (PET in left upper panel, CT in right upper
old obese woman with cancer of the right breast. The MIP image (lower panel, fused PET/CT in left lower panel) focus on a lymph node metas-
right panel) demonstrates multiple metastases throughout the skeleton, tasis in right axilla
Fig. 23.17 PEM with [18F]FDG acquired in a 66-year-old woman pre- position. In addition to a recurrent cancer in the residual left breast
viously submitted to surgery followed by electron beam radiation ther- (axial CT and fused PET/CT image in left panels), PEM imaging shows
apy (QUART) because of cancer in her left breast; acquisition is a pericentimetric, hypermetabolic (SUVmax 2.5) lesion at a higher level
performed in the prone position but displayed in the customary supine in the chest wall (right panels)
of Clinical Oncology and the American Board of Internal In 1999, the PET Study Group of the European
Medicine Foundation’s Choosing Wisely® campaign [108, Organisation for Research and Treatment of Cancer (EORTC)
109]. On the other hand, NCCN and ESMO guidelines developed criteria to evaluate tumor lesions at baseline and
advise [18F]FDG-PET as additional work-up in locally to assess treatment response using [18F]FDG-PET, based on
advanced, inflammatory (Fig. 23.18), recurrent, or metastatic hypothetical considerations as well as literature review with
disease, especially when standard imaging remains equivo- limited published and unpublished data available at that time
cal or suspicious and if lesions are inaccessible for biopsy. [113]. The change in standardized uptake value (SUV)
The diagnostic accuracy of [18F]FDG PET/CT in patients within the tumor lesion compared to a baseline scan is used
with newly diagnosed breast cancer is reported with variable for the assessment. Complete metabolic response is defined
values in the literature, with mean sensitivity of 56% and as complete resolution of the lesion’s [18F]FDG uptake
mean specificity of 96% [110]. The high specificity and high against its surrounding tissue and partial metabolic response
positive predictive value of [18F]FDG PET/CT (almost as a reduction of ≥15% SUV after one cycle or ≥25% after
invariably exceeding 90%) means that, in case of [18F]FDG- two or more cycles of chemotherapy (Fig. 23.19).
avid axillary lymph node(s), the SLNB procedure could be It must be emphasized that also standardized imaging
skipped and immediate ALND can be planned [111]. [18F] protocols are required to be able to evaluate response.
FDG PET/CT may also detect other lymph node metastatic Another attempt to introduce a standard method for PET
sites, e.g., in the internal mammary chain and periclavicular interpretation in the assessment of tumor response to treat-
chain [112]. ments is formulated in the PERCIST criteria [114].
Fig. 23.18 PEM with [18F]FDG acquired in a 58-year-old woman with increased [18F]FDG uptake in the skin corresponding to mastitis carci-
newly diagnosed cancer of the right breast. In addition to the primary nomatosa, as displayed also on the fused PET/CT images in right panels
tumor, axial CT and fused PET/CT images (left panels) show markedly (coronal section in upper panel, sagittal section in lower panel)
After an extensive literature review, response criteria

were proposed, and conclusions were reached using a Key Learning Points
Delphi-like approach. The PERCIST criteria advise to cor- • [18F]FDG PET/CT plays an important role for the
rect the SUV for lean body mass (SUL). According to these detection of distant metastases and tumor recur-
criteria, partial metabolic response is defined as a SUL peak rence, as well as for evaluation of treatment
reduction ≥30%. Furthermore, signal of the target lesion responses in breast cancer patients.
must be less than mean liver activity and indistinguishable • Dedicated positron emission mammography
from surrounding blood-pool levels to be classified as com- devices that maximize the system spatial resolution
plete metabolic response [115]. have been designed to improve detection of small
Assessment of response to therapy in breast cancer [18F]FDG-avid primary lesions.
patients is clinically relevant especially in cases with large • [18F]FDG PET/CT is recommended in locally
and locally advanced tumors undergoing neoadjuvant che- advanced, inflammatory, recurrent, or metastatic
motherapy. Another unsettled issue is the optimal time disease, especially when standard imaging remains
point for acquiring an interim [18F]FDG-PET/CT during equivocal or suspicious and if lesions are inacces-
neoadjuvant chemotherapy [116]. In this regard, a signifi- sible for biopsy.
cantly higher SUVmax reduction as early as after the first or • The change in standardized uptake value (SUV)
second chemotherapy course is observed in patients with his- within the tumor lesion compared to a baseline scan
tological evidence of favorable tumor response as compared is used to evaluate tumor response to treatment.
to chemoresistant ones [117].
Fig. 23.19 Sequential PET/CT scans with [18F]FDG performed in a markedly increased [18F]FDG uptake in the tumor located medially
40-year-old woman with locally advanced, triple-negative ductal carci- (SUVmax 13.1). The scan acquired after completion of neoadjuvant che-
noma of the right breast (axial CT images in upper panels, fused PET/ motherapy (right panels) shows reduced but persistent [18F]FDG uptake
CT images in lower panels). The baseline scan (left panels) shows (SUVmax 9.4) consistent with partial metabolic response
23.6.2 Clinical Use of Non-[18F]FDG Agents cannot be incorporated into DNA and thus becomes
in Patients with Breast Cancer trapped intracellularly.
Despite good correlation with Ki67, the signal intensity
Although [18F]FDG is the most frequently used PET radio- of 18F-FLT uptake is generally lower than with [18F]FDG,
pharmaceutical, other PET tracers are being developed and thus potentially leading to false-negative findings. This
used, not only for investigational purposes but also in clinical observation has so far limited the clinical utility of 18F-FLT
practice. PET/CT for staging, and additional studies are needed to bet-
Since tumor proliferation is a key biologic marker espe- ter define the clinical efficacy of 18F-FLT PET/CT for assess-
cially when assessing therapeutic efficacy for all types of ing early response to antitumor therapies [118, 119].
cancers, it may reasonably provide a better marker of early Amino acid transport is upregulated in many types of can-
response than glycolysis. cer, as a response to support the demands of increased pro-
3′-Deoxy-3′- 18F-fluorothymidine (18F-FLT) is the most tein synthesis and proliferation of malignant cells. A
widely investigated PET tracer for imaging prolifera- radiolabeled essential amino acid, l-methyl-[11C]-methionine
tion, although it is not approved yet by the United States ([11C]methionine), has been shown to accumulate in primary
Food and Drug Administration for use outside of clinical and metastatic breast cancer and can be effectively imaged
trials. 18F-FLT enters cells and becomes phosphorylated with PET [120, 121]. Furthermore, intensity of its uptake is
by thymidine kinase-1 as part of the thymidine salvage correlated with the fraction of cells in S phase of the cell
pathway of DNA synthesis. Phosphorylated 18F-FLT cycle and can therefore indirectly indicate the proliferative
status of the tumor. However, the logistical demands of rapid Although not a direct target per se of endocrine therapy in
synthesis and scanning of 11C-based radiopharmaceuticals, breast cancer, the progesterone receptor (PR) is a predictive
due to the 20-min physical half-life of this radionuclide, lim- marker for response to antiestrogen therapy. Several attempts
its its use to institutions with an on-site cyclotron. have been made to develop a PR-specific PET tracer,
Thus, attention has recently shifted toward 18F-labeled although with limited success. The best tracer available to
amino acids [122]. The synthetic amino acid anti-1-amino-3- date is 21-18F-fluoro-16α,17α-[((R)-1″-α-furylmethylidene)
18
F-fluorocyclobutane-1-carboxcylic acid (18F-FACBC, or dioxy]-19-norpregn-4-ene-3,20-dione (18F-FFNP). Early
18
F-fluciclovine), a leucine analog, has recently been approved results in preclinical models suggest that serial 18F-FFNP
by the FDA. While its clinical indication is for localization of PET scan may be a good readout and predict the efficacy of
biochemically recurrent prostate cancer, it may also have an hormonal therapy [129].
application for breast cancer imaging. Higher uptake of Amplification of the HER2 gene results in overexpression
18
F-FACBC has been demonstrated in patients with primary of the HER2 protein, which occurs in 20–25% of primary
and metastatic breast cancer compared with normal breast tis- breast cancers. When left untreated, HER2 overexpression is
sue and benign breast lesions [122, 123]. associated with aggressive growth and poor prognosis. The
An important factor in determining response to systemic anti-HER2 monoclonal antibody trastuzumab, which targets
therapy is tumor perfusion. Tumors that are poorly perfused the extracellular domain of HER2, is part of treatment in the
may not receive adequate delivery of systemic therapy to adjuvant as well as in the metastatic setting of HER2-positive
work effectively. 15O-labeled water ([15O]H2O) can be used breast cancer [130]. Radiolabeled trastuzumab has been used
to image tumor blood flow, which has been shown to be for molecular imaging of the HER2 status in breast cancer
increased in breast malignancies compared with normal patients using either single-photon or PET tracers. After devel-
breast tissue. Given the extremely short physical half-life of oping 111In-trastuzumab for clinical use, the next step in HER2
15
O (2 min), the [15O]H2O PET scan can be combined within imaging was labeling of trastuzumab with the positron-emitting
the same imaging session with other PET scans, such as [18F] radionuclide 89Zr for PET imaging of HER2 expression. PET
FDG, to map matched data regarding tumor blood flow and imaging has higher spatial resolution than SPECT and tumor
glucose metabolism [124]. As with [11C]methionine, the use uptake of the tracer can better be quantified. 89Zr has a half-life
of [15O]H2O is restricted to nuclear medicine centers with an of 78.4 h, which is compatible with the relative long biological
on-site cyclotron for the production of positron-emitting half-life of the trastuzumab antibody. Moreover, unknown
radionuclides. brain metastases were detected, showing that 89Zr-trastuzumab
The most relevant hormone receptor in breast cancer is can cross the blood-brain barrier in case of brain metastasis.
the estrogen receptor (ER), which is expressed in ~75% of The predictive value of PET with 89Zr-trastuzumab still
the patients. For the patients with an ER-positive tumor, deserves further studies [131]. Trastuzumab has also been
endocrine therapy can be an important treatment option. It is labeled to 64Cu for PET imaging. The half-life of 64Cu is 12.7 h.
possible to evaluate ER expression in vivo by PET imaging This leads to less radiation exposure compared to 89Zr imaging,
with 18F-fluoro-estradiol (18F-FES). 18F-FES-PET can quan- but may also have a relatively short physical half-life compared
titatively estimate tumor ER expression with approximately to the long biological half-life of trastuzumab. PET with
70–100% sensitivity and 80–100% specificity when com- 64
Cu-DOTA-trastuzumab detected primary breast cancer and
pared to in vitro assays [125]. These results support future lymph node and lung metastases in HER2-positive breast can-
trials to examine 18F-FES-PET to reevaluate ER expression cer patients on trastuzumab therapy [132].
noninvasively in patients who cannot be biopsied and thus
may support treatment decision-making [126].
The androgen receptor (AR) is a key target in prostate Key Learning Points
cancer patients, and various antiandrogens are available for • Although [18F]FDG is the most frequently used
therapy. Nevertheless, the AR is expressed in ~70% of all PET tracer, other PET tracers are being investigated
breast cancer patients and in 12–40% of the so-called triple- and used in patients with breast cancer.
negative patients [127]. Therefore, the AR is currently being • The most investigated non-[18F]FDG tracers in patients
explored also as a potential therapeutic target in breast can- with breast cancer include 18F-FLT, [11C]methionine,
cer patients. The AR can be imaged with PET using 18
F-FACBC, [15O]H2O, 18F-FES, 18F-FDHT, 18F-FFNP,
18
F-fluoro-dihydrotestosterone (18F-FDHT). Although no 89
Zr-trastuzumab, and 64Cu-DOTA-trastuzumab.
information is currently available on the diagnostic perfor- • Even if such tracers are used not only for investiga-
mance of 18F-FDHT-PET in breast cancer patients, with the tional purposes, but also in clinical practice, addi-
emerging interest for antiandrogen therapy in breast cancer tional clinical studies are needed for their full
patients [128], this PET tracer may well show its value in the validation.
near future.
23.7 PET/MRI of the dedicated breast radiofrequency receiver coils needed

for prone positioning and high image quality. If traditional
In addition to new radiopharmaceuticals, novel advanced PET radiofrequency receiver coils are used for integrated PET/MR
technologies, mainly PET/MRI, may offer new opportunities. imaging, their presence in the field of view of the PET detec-
Hybrid PET/MRI may be particularly interesting because of tors attenuate the number of photons originated within the
its superior resolution and soft tissue contrast and its lack of breast and can result in inaccurate SUV measurement. Thus,
ionizing radiation exposure for morphologic imaging. PET/ 16-channel breast radiofrequency coils and attenuation cor-
MR imaging scanners were introduced in 2010, and there are rection have been recently designed specifically for simulta-
currently approximately 70 systems in place worldwide as of neous PET/MR imaging for accurate PET quantitation and
a 2016 report, primarily located at academic centers [133]. have been tested in small patient cohorts. This hybrid func-
The most relevant benefits of PET/MR imaging over PET/ tional imaging modality may offer advantages for assessment
CT imaging can be summarized as follows: of response to therapy in patients with primary breast cancer
by combining the high sensitivity of MR imaging with the
• Saving time: compared to separate PET and MRI exami- high specificity of [18F]FDG PET [134–136].
nations, the simultaneous procedure takes about 30 min
instead of 60–90 min.
• Imaging the most complex cases: PET/MRI can be used Key Learning Points
for advanced diagnostics in oncology, neurology, and • PET/MRI is particularly interesting for use in breast
cardiology. cancer patients because of its superior resolution
• Saving space: the 2-in-1 system helps to optimize room and soft tissue contrast and the lack of exposure to
utilization within the healthcare organization, while two ionizing radiation for morphologic imaging.
separate imaging halls are needed for PET and MRI • The most relevant benefits of PET/MR imaging are
devices separately. saving time and space, imaging the most complex
• Improving co-registration for image fusion: due to the cases, improving co-registration for image fusion,
same patient position throughout the examination in and reduction of the overall radiation dose that
simultaneous PET/MRI scanning, the synergetic image patients receive during the examination.
has better quality than fusion images obtained post- • The most relevant drawbacks of PET/MR imaging
acquisition of separate PET and MR images. are attenuation correction issues, long acquisition
time, and cost.
One of the main advantages of PET/MR imaging over
PET/CT is the lack of ionizing radiation for morphologic
imaging, which allows a reduction of the overall radiation
dose that patients can get during the examination. This fea-
ture is especially important for oncologic and pediatric
References
imaging. On the other hand, there are some drawbacks to the 1. Sun YS, Zhao Z, Yang ZN, Xu F, Lu HJ, Zhu ZY, Shi W, Jiang J, Yao
wide diffusion of PET/MR, as follows: PP, Zhu HP. Risk factors and preventions of breast cancer. Int J Biol
Sci. 2017;13:1387–97.
• Attenuation correction (AC) issues: the PET/CT devices 2. Shah R, Rosso K, Nathanson SD. Pathogenesis, prevention,
diagnosis and treatment of breast cancer. World J Clin Oncol.
use X-ray-based correction, which is impossible in PET/ 2014;5:283–98.
MRI. MRI depends on bone for AC calculations, which 3. Hammerl D, Smid M, Timmermans AM, Sleijfer S, Martens JWM,
makes it difficult when it comes to calculating attenuation Debets R. Breast cancer genomics and immuno-oncological mark-
coefficients. ers to guide immune therapies. Semin Cancer Biol. 2017 Nov 6. pii:
S1044 579X(17)30186-4.
• Long acquisition time: acquisition times for PET/MRI 4. Chikarmane SA, Tirumani SH, Howard SA, Jagannathan JP, DiPiro
vary depending on MR pulse sequences. While simulta- PJ. Metastatic patterns of breast cancer subtypes: what radiolo-
neous imaging is better in comparison with PET and MRI gists should know in the era of personalized cancer medicine. Clin
separately, it may fall short to PET/CT imaging, a feature Radiol. 2015;70:1–10.
5. Greene LR, Wilkinson D. The role of general nuclear medicine in
which could impact workflow. breast cancer. J Med Radiat Sci. 2015;62:54–65.
• Investment risks: the cost of a PET/MRI scanner is approx- 6. Jafari SH, Saadatpour Z, Salmaninejad A, Momeni F, Mokhtari
imately within the range of US $ 4–five million, to which M, Nahand JS, et al. Breast cancer diagnosis: imaging techniques
another 10% must be added for annual maintenance [133]. and biochemical markers. J Cell Physiol. 2017 Dec 8; https://doi.
org/10.1002/jcp.26379.. [Epub ahead of print]
7. Oude Munnink TH, Nagengast WB, Brouwers AH, Schröder CP,
Investigations on simultaneous breast PET/MR imaging Hospers GA, Lub-de Hooge MN, et al. Molecular imaging of breast
for primary breast cancer have been delayed by the initial lack cancer. Breast. 2009;18:S66–73.
8. Brem R, Rechtman L. Nuclear medicine imaging of the breast: a 30. Spanu A, Chessa F, Sanna D, Cottu P, Manca A, Nuvoli S, et al.
novel, physiological approach to breast cancer detection and diag- Scintimammography with a high resolution dedicated breast cam-
nosis. Radiol Clin N Am. 2010;48:1055–74. era in comparison with SPECT/CT in primary breast cancer detec-
9. Ferrara A. Nuclear imaging in breast cancer. Radiol Technol. tion. Q J Nucl Med Mol Imaging. 2009;53:271–80.
2010;81:233–46. 31. Xu HB, Li L, Xu Q. Tc-99m sestamibi scintimammography for
10. DeCesare A, De Vincentis G, Gervasi S, Crescentini G, Fiori E, the diagnosis of breast cancer: meta-analysis and meta-regression.
Bonomi M, et al. Single photon emission computed tomography Nucl Med Commun. 2011;32:980–8.
(SPECT) with Technetium-99m sestamibi in the diagnosis of 32. Sun Y, Wei W, Yang H-W, Liu J-L. Clinical usefulness of breast-
small breast cancer and axillary lymph node involvement. World J specific gamma imaging as an adjunct modality to mammog-
Surg. 2011;35:2668–72. raphy for diagnosis of breast cancer: a systemic review and
11. Wackers FJ, Berman DS, Maddahi J, et al. Technetium-99m- meta-analysis. Eur J Nucl Med Mol Imaging. 2013;40:450–63.
hexakis 2-methoxyisobutyl isonitrile: human biodistribution, 33. Tan H, Zhang H, Yang W, Fu Y, Gu Y, Du M, et al. Breast-specific
dosimetry, safety and preliminary comparison to thallium-201 for gamma imaging with Tc-99m-sestamibi in the diagnosis of breast
myocardial perfusion imaging. J Nucl Med. 1998;30:301–11. cancer and its semiquantitative index correlation with tumor bio-
12. Kao CH, Wang SJ, Liu TJ. The use of technetium-99m methoxy- logic markers, subtypes, and clinicopahologic characteristics.
isobutylisonitrile breast scintigraphy to evaluate palpable breast Nucl Med Commun. 2016;37:792–9.
masses. Eur J Nucl Med. 1994;21:432–6. 34. Collarino A, de Koster EJ, Valdés Olmos RA, de Geus-Oei LF,
13. Burak Z, Argon M, Memiş A, Erdem S, Balkan Z, Duman Y, et al. Pereira Arias-Bouda LM. Is technetium-99m sestamibi imaging
Evaluation of palpable breast masses with 99Tcm-MIBI: a comparable to predict pathologic nonresponse to neoadjuvant chemother-
ative study with mammography and ultrasonography. Nucl Med apy in breast cancer? A meta-analysis evaluating current use and
Commun. 1994;15:604–12. shortcomings. Clin Breast Cancer. 2018;8:9–18.
14. Hendrick RE. Radiation doses and cancer risks from breast imag- 35. Trehan R, Seam RK, Gupta MK, et al. Role of scintimammog-
ing studies. Radiology. 2010;257:246–53. raphy in assessing the response of neoadjuvant chemotherapy
15. Specht JM, Mankoff DA. Advances in molecular imaging for in locally advanced breast cancer. World J Nucl Med. 2014;13:
breast cancer detection and characterization. Breast Cancer Res. 163–9.
2012;14:206–17. 36. Guo C, Zhang C, Liu J, et al. Is Tc-99m sestamibi scintimam-
16. Taillefer R. Clinical applications of 99mTc-sestamibi scintimammography useful in the prediction of neoadjuvant chemotherapy
mography. Semin Nucl Med. 2005;35:100–15. responses in breast cancer? A systematic review and meta-
17. Lee J, Rosen E, Mankoff D. The role of radiotracer imaging in analysis. Nucl Med Commun. 2016;37:675–88.
the diagnosis and management of patients with breast cancer: part 37. Goldsmith SJ, Parsons W, Guibertau MJ, Stern LH, Lanzkowsky
1 – overview, detection and staging. J Nucl Med. 2009;50:569–81. L, Weigert J, et al. SNM practice guideline for breast scintigra-
18. Jacobsson H. Single-photon-emission computed tomography phy with breast-specific γ-cameras 1.0. J Nucl Med Technol.
(SPECT) with 99mTechnetium sestamibi in the diagnosis of small 2010;38:219–24.
breast cancer and axillary node involvement. World J Surg. 38. Covington MF, Rhodes DJ, Pizzitola VJ. Molecular breast imag-
2011;35:2673–4. ing and the 2016 update of the ACR appropriateness criteria
19. Kong FL, Kim E, Yang D. Targeted nuclear imaging of breast can- for breast cancer screening. J Am Coll Radiol. 2016;13(12. Pt
cer: status of radiotracer development and clinical applications. A):1408.
Cancer Biother Radiopharm. 2012;27:105–12. 39. Williams MB, Judy PG, Gunn S, Majewski S. Dual-modality
20. Kim SJ, Kim IJ, Bae YT, Kim YK, Kim DS. Comparison of breast tomosynthesis. Radiology. 2010;255:191–8.
quantitative and visual analysis of Tc-99m MIBI scintimam- 40. Saunders RS Jr, Samei E, Lo JY, Baker JA. Can compression be
mography for detection of primary breast cancer. Eur J Radiol. reduced for breast tomosynthesis? Monte Carlo study on mass
2005;53:192–8. and microcalcification conspicuity in tomosynthesis. Radiology.
21. Nguyen BD, Roarke MC, Karstaedt PJ, Ingui CJ, Ram 2009;251:673–82.
PC. Practical applications of nuclear medicine in imaging breast 41. Gong Z, Williams MB. Comparison of breast specific gamma
cancer. Curr Probl Diagn Radiol. 2009;38:68–83. imaging and molecular breast tomosynthesis in breast cancer
22. Prekeges J. Breast imaging devices for nuclear medicine. J Nucl detection: evaluation in phantoms. Med Phys. 2015;42:4250–9.
Med Technol. 2012;40:71–8. 42. Li H, Zheng Y, Moré MJ, Goodale PJ, Williams MB. Lesion quan-
23. Waxman A. The role of 99mTc methoxyisobutylisonitrile in imag- tification in dual-modality mammography. IEEE Trans Nucl Sci.
ing breast cancer. Semin Nucl Med. 1997;27:40–54. 2007;54:107–15.
24. Moadel RM. Breast cancer imaging devices. Semin Nucl Med. 43. Moré MJ, Heng L, Goodale PJ, Zheng Y, Majewski S, Popov V,
2011;41:229–41. et al. Limited angle dual modality breast imaging. IEEE Trans
25. Fowler A. A molecular approach to breast imaging. J Nucl Med. Nucl Sci. 2007;54:504–13.
2014;55:177–80. 44. Chagpar AB, Kehdy F, Scoggins CR, Martin RC 2nd, Carlson DJ,
26. Surti S. Radionuclide methods and instrumentation for breast can- Laidley AL, et al. Effect of lymphoscintigraphy drainage patterns
cer detection and diagnosis. Semin Nucl Med. 2013;43:271–80. on sentinel lymph node biopsy in patients with breast cancer. Am
27. Schillaci O, Danieli R, Filippi L, Romano P, Cossu E, Manni C, J Surg. 2005;190:557–62.
et al. Scintimammography with a hybrid SPECT/CT imaging sys- 45. Orsini F, Guidoccio F, Vidal-Sicart S, Renato A. Valdés Olmos
tem. Anticancer Res. 2007;27:557–62. RA, Mariani G. General concepts on radioguided sentinel lymph
28. Spanu A, Cottu P, Manca A, Chessa F, Sanna D, Madeddu node biopsy: preoperative imaging, intra-operative gamma-probe
G. Scintimammography with dedicated breast camera in unifo- guidance, intraoperative imaging, and multimodality imaging. In:
cal and multifocal/multicentric primary breast cancer detection: a Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA,
comparative study with SPECT. Int J Oncol. 2007;31:369–77. eds. Atlas of lymphoscintigraphy and sentinel node mapping – a
29. Spanu A, Farris A, Chessa F, Sanna D, Pittalis M, Manca A, et al. pictorial case-based approach. Milan: Springer; 2013: pp 95–110.
Planar scintimammography and PSECT in neoadjuvant chemo or 46. Orsini F, Mariani G. Clinical applications of radioguided sentinel
hormotherapy response evaluation in locally advanced primary lymph node biopsy. In: Duatti A, editor. Radiopharmaceuticals for
breast cancer. Int J Oncol. 2008;32:1275–83. sentinel lymph node detection: status and trends. Vienna: IAEA
Radioisotopes and Radiopharmaceuticals Series No. 6; 2015. of sentinel lymph nodes in breast cancer and redefine lymphatic
p. 33–71. drainage from the breast to the axilla. Breast. 2012;21:480–6.
47. Giammarile F, Orsini F, Valdés Olmos RA, Vidal-Sicart S, 64. Giammarile F, Alazraki N, Aarsvold JN, Audisio RA, Glass E,
Giuliano AE, Mariani G. Radioguided surgery for breast cancer. Grant SF, et al. The EANM and SNMMI practice guideline for
In: Strauss HW, Mariani G, Volterrani D, Larson SM, editors. lymphoscintigraphy and sentinel node localization in breast can-
Nuclear oncology – from pathophysiology to clinical applications. cer. Eur J Nucl Med Mol Imaging. 2013;40:1932–47.
2nd ed. New York: Springer; 2017. p. 1363–400. 65. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW,
48. Bafford A, Gadd M, Gu X, Lipsitz S, Golshan M. Diminishing Blumencranz PW, et al. Axillary dissection vs no axillary dis-
morbidity with the increased use of sentinel node biopsy in breast section in women with invasive breast cancer and sentinel node
carcinoma. Am J Surg. 2010;200:374–7. metastasis: a randomized clinical trial. JAMA. 2011;305:569–75.
49. Janni W, Kühn T, Schwentner L, Kreienberg R, Fehm T, Wöckel 66. Ram R, Singh J, McCaig E. Sentinel node biopsy alone versus
A. Sentinel node biopsy and axillary dissection in breast can- completion axillary node dissection in node positive breast can-
cer: the evidence and its limits. Dtsch Arztebl Int. 2014;111: cer: systematic review and meta-analysis. Int J Breast Cancer.
244–9. 2014;2014:513780. https://doi.org/10.1155/2014/513780.
50. Lopez Penha TR, van Roozendaal LM, Smidt ML, Boersma LJ, 67. FitzSullivan E, Bassett RL, Kuerer HM, Mittendorf EA, Yi M,
von Meyenfeldt MF, Voogd AC, et al. The changing role of axil- Hunt KK, et al. Outcomes of sentinel lymph node-positive breast
lary treatment in breast cancer: who will remain at risk for devel- cancer patients treated with mastectomy without axillary therapy.
oping arm morbidity in the future? Breast. 2015;24:543–7. Ann Surg Oncol. 2017;24:652–9.
51. Bromham N, Schmidt-Hansen M, Astin M, Hasler E, Reed 68. Giuliano AE, Ballman KV, McCall L, Beitsch PD, Brennan MB,
MW. Axillary treatment for operable primary breast cancer. Kelemen PR, et al. Effect of axillary dissection vs no axillary dis-
Cochrane Database Syst Rev. 2017;1:CD004561. https://doi. section on 10-year overall survival among women with invasive
org/10.1002/14651858.CD004561.pub3. breast cancer and sentinel node metastasis: the ACOSOG Z0011
52. Mamounas EP, Kuehn T, Rutgers EJT, von Minckwitz (Alliance) randomized clinical trial. JAMA. 2017;318:918–26.
G. Current approach of the axilla in patients with early-stage 69. Ozcan LC, Giuliano AE. Is axillary lymph node dissection nec-
breast cancer. Lancet. 2017 Aug 14; https://doi.org/10.1016/ essary after a positive sentinel lymph node biopsy? Adv Surg.
S0140-6736(17)31451-4.. [Epub ahead of print] 2017;51:165–78.
53. Ballal H, Hunt C, Bharat C, Murray K, Kamyab R, Saunders 70. Lyman GH, Somerfield MR, Bosserman LD, Perkins CL, Weaver
C. Arm morbidity of axillary dissection with sentinel node biopsy DL, Giuliano AE. Sentinel lymph node biopsy for patients with
versus delayed axillary dissection. ANZ J Surg. 2018 Feb 2; early-stage breast cancer: American Society of Clinical Oncology
https://doi.org/10.1111/ans.14382.. [Epub ahead of print] clinical practice guideline update. J Clin Oncol. 2017;35:561–4.
54. Rebegea L, Firescu D, Dumitru M, Anghel R. The incidence and 71. van der Ploeg IM, Hobbelink M, van den Bosch MA, Borel Rinkes
risk factors for occurrence of arm lymphedema after treatment of IH, van Hillegersberg R. Radioguided occult lesion localisation
breast cancer. Chirurgia (Bucur). 2015;110:33–7. (ROLL) for non-palpable breast lesions: a review of the relevant
55. Cho Y, Do J, Jung S, Kwon O, Jeon JY. Effects of a physical ther- literature. Eur J Surg Oncol. 2008;34:1–5.
apy program combined with manual lymphatic drainage on shoul- 72. Buchberger W, Niehoff A, Obrist P, DeKoekkoek-Doll P, Dünser
der function, quality of life, lymphedema incidence, and pain in M. Clinically and mammographically occult breast lesions: detec-
breast cancer patients with axillary web syndrome following axil- tion and classification with high-resolution sonography. Semin
lary dissection. Support Care Cancer. 2016;24:2047–57. Ultrasound CT MR. 2000;21:325–36.
56. Nanocoll [package insert] Brondby D, GE healthcare, 2012. 73. Besic N, Zgajnar J, Hocevar M, Rener M, Frkovic-Grazio S, Snoj
57. Moncayo VM, Aarsvold JN, Grant SF, Bartley SC, Alazraki N, et al. Breast biopsy with wire localization: factors influenc-
NP. Status of sentinel lymph node for breast cancer. Semin Nucl ing complete excision of nonpalpable carcinoma. Eur Radiol.
Med. 2013;43:281–93. 2002;12:2684–9.
58. Orsini F, Puta E, Lorenzoni A, Erba PA, Mariani G. Single- 74. Sajid MS, Parampalli U, Haider Z, Bonomi R. Comparison of
photon-emitting radiopharmaceuticals for diagnostic applications. radioguided occult lesion localization (ROLL) and wire local-
In: Strauss HW, Mariani G, Volterrani D, Larson SM, editors. ization for non-palpable breast cancers: a meta-analysis. J Surg
Nuclear oncology – from pathophysiology to clinical applications. Oncol. 2012;105:852–8.
2nd ed. New York: Springer; 2017. p. 67–84. 75. Langhans L, Klausen TL, Tvedskov TF, Vejborg I, Kroman N,
59. Caruso G, Cipolla C, Costa R, Morabito A, Latteri S, Fricano Hesse B. Radioguided surgery for localization of nonpalpable
S, et al. Lymphoscintigraphy with peritumoral injection ver- breast lesions. A mini-review. Curr Radiopharm. 2016;9:114–20.
sus lymphoscintigraphy with subdermal periareolar injection of 76. Luini A, Zurrida S, Galimberti V, Paganelli G. Radioguided sur-
technetium- labeled human albumin to identify sentinel lymph gery of occult breast lesions. Eur J Cancer. 1998;34:205–6.
nodes in breast cancer patients. Acta Radiol. 2014;55:39–44. 77. Lovrics PJ, Cornacchi SD, Vora R, Goldsmith CH, Kahnamoui
60. Brouwer OR, Vermeeren L, van der Ploeg IMC, Valdés Olmos K. Systematic review of radioguided surgery for non-palpable
RA, Loo CE, Pereira-Bouda LM, et al. Lymphoscintigraphy and breast cancer. Eur J Surg Oncol. 2011;37:388–97.
SPECT/CT in multicentric and multifocal breast cancer: does 78. de Azambuja E, Bedard PL, Suter T, Piccart-Gebhart M. Cardiac
each tumour have a separate drainage pattern? Eur J Nucl Med toxicity with anti-HER-2 therapies: what have we learned so far?
Mol Imaging. 2012;39:1137–43. Target Oncol. 2009;4:77–88.
61. Mariani G, Moresco L, Viale G, Villa G, Bagnasco M, Canavese 79. de Geus-Oei LF, Mavinkurve-Groothuis AM, Bellersen L, Gotthardt
G, et al. Radioguided sentinel lymph node biopsy in breast cancer M, Oyen WJ, Kapusta L, van Laarhoven HW. Scintigraphic tech-
surgery. J Nucl Med. 2001;42:1198–215. niques for early detection of cancer treatment-induced cardiotox-
62. Aliakbarian M, Memar B, Jangjoo A, Zakavi SR, Reza Dabbagh icity. J Nucl Med Technol. 2013;41:170–81.
Kakhki V, Aryana K, et al. Factors influencing the time of senti- 80. Thorley P, Bould M, Chowdhury F. Gated cardiac SPECT: can it
nel node visualization in breast cancer patients using intradermal be used for serial assessment of left ventricular function in oncol-
injection of the radiotracer. Am J Surg. 2011;202:199–202. ogy patients? Nucl Med Commun. 2013;34:1137–40.
63. Uren RF, Howman-Giles R, Chung DKV, Spillane AJ, Noushi F, 81. de Gues-Oei LF, Mavinkurve-Groothuis AM, Bellersen L,
Gillett D, et al. SPECT/CT scans allow precise anatomical location Gotthardt M, Oyen WJ, Kapusta L, et al. Scintigraphic techniques
for early detection of cancer treatment-induced cardiotoxicity. J 99. Shie P, Cardarelli C, Brandon D, Erdman W, Abdulrahim N. Meta-
Nucl Med. 2011;52:560–71. analysis: comparison of F-18 fluorodeoxyglucose- positron
82. Marwick TH. Measurement of strain and strain rate by echo- emission tomography and bone scintigraphy in the detection of
cardiography: ready for prime time? J Am Coll Cardiol. bone metastasis in patients with breast cancer. Clin Nucl Med.
2006;47:1313–27. 2008;33:97–101.
83. Mantovani G, Madeddu C, Cadeddu C, Dessì M, Piras A, Massa 100. Glendenning J, Cook G. Imaging breast cancer bone metas-
E, et al. Persistence, up to 18 months of follow-up, of epirubicin- tases: current status and future directions. Semin Nucl Med.
induced myocardial dysfunction detected early by serial tissue 2013;43:317–23.
Doppler echocardiography: correlation with inflammatory and 101. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evalu-
oxidative stress markers. Oncologist. 2008;13:1296–305. ation criteria in solid tumours: revised RECIST guideline (version
84. Mavinkurve-Groothuis AM, Weijers G, Groot-Loonen J, Pourier 1.1). Eur J Cancer. 2009;45:228–47.
M, Feuth T, de Korte CL, et al. Interobserver, intraobserver and 102. Hamaoka T, Costelloe CM, Madewell JE, Liu P, Berry DA, Islam
intrapatient reliability scores of myocardial strain imaging with R, et al. Tumour response interpretation with new tumour response
2-d echocardiography in patients treated with anthracyclines. criteria vs the World Health Organisation criteria in patients with
Ultrasound Med Biol. 2009;35:697–704. bone-only metastatic breast cancer. Br J Cancer. 2010;102:651–7.
85. Altena R, Perik PJ, van Veldhuisen DJ, de Vries EG, Gietema 103. Hong S, Li J, Wang S. 18FDG PET-CT for diagnosis of distant
JA. Cardiovascular toxicity caused by cancer treatment: strategies metastases in breast cancer patients. A meta-analysis. Surg Oncol.
for early detection. Lancet Oncol. 2009;10:391–9. 2013;22:139–43.
86. Hesse B, Lindhardt TB, Acampa W, Anagnostopoulos C, Ballinger 104. Bitencourt AGV, Andrade WP, da Cunha RR, Conrado JLFA,
J, Bax JJ, et al. EANM/ESC guidelines for radionuclide imaging of Lima ENP, Barbosa PNVP, et al. Detection of distant metastases
cardiac function. Eur J Nucl Med Mol Imaging. 2008;35:851–85. in patients with locally advanced breast cancer: role of 18F- fluoro-
87. Panjrath GS, Jain D. Monitoring chemotherapy-induced car- deoxyglucose positron emission tomography/computed tomogra-
diotoxicity: role of cardiac nuclear imaging. J Nucl Cardiol. phy and conventional imaging with computed tomography scans.
2006;13:415–26. Radiol Bras. 2017;50:211–5.
88. Body JJ, Quinn G, Talbot S, Booth E, Demonty G, Taylor A, 105. Kalles V, Zografos GC, Provatopoulou X, Koulocheri D, Gounaris
Amelio J. Systematic review and meta-analysis on the proportion A. The current status of positron emission mammography in
of patients with breast cancer who develop bone metastases. Crit breast cancer diagnosis. Breast Cancer. 2013;20:123–30.
Rev Oncol Hematol. 2017;115:67–80. 106. Hsu DF, Freese DL, Levin CS. Breast-dedicated radionuclide
89. Du Y, Cullum I, Illidge TM, Ell PJ. Fusion of metabolic function imaging systems. J Nucl Med. 2016;57(Suppl 1):40S–5S.
and morphology: sequential [18F]fluorodeoxyglucose positron- 107. Caresia Aroztegui AP, García Vicente AM, Alvarez Ruiz S,
emission tomography/computed tomography studies yield new Delgado Bolton RC, Orcajo Rincon J, Garcia Garzon JR, et al. 18F-
insights into the natural history of bone metastases in breast can- FDG PET/CT in breast cancer: evidence-based recommendations
cer. J Clin Oncol. 2007;25:3440–7. in initial staging. Tumour Biol. 2017;39(10):1010428317728285.
90. Hahn S, Heusner T, Kümmel S, Köninger A, Nagarajah J, Müller https://doi.org/10.1177/1010428317728285.
S, et al. Comparison of FDG-PET/CT and bone scintigraphy 108. Schnipper LE, Smith TJ, Raghavan D, Blayney DW, Ganz PA,
for detection of bone metastases in breast cancer. Acta Radiol. Mulvey TM, et al. American Society of Clinical Oncology
2011;52:1009–14. identifies five key opportunities to improve care and reduce costs:
91. Jeong D, Bui M, Peterson D, Montilla-Soler J, Gage KL. FDG the top five list for oncology. J Clin Oncol. 2012;30:1715–24.
avid breast cancer bone metastases silent on CT and scintigraphy: 109. Mayo-Smith WW, Morin RL. Image wisely: the beginning, current
a case report with radiologic-pathologic correlation. Acta Radiol status, and future opportunities. J Am Coll Radiol. 2017;14:442–3.
Open. 2017 Oct 6;6(10):2058460117734243. 110. Zhang X, Wu F, Han P. The role of 18F-FDG PET/CT in the diag-
92. Hamaoka T, Madewell JE, Podoloff DA, Hortobagyi GN, Ueno nosis of breast cancer and lymph nodes metastases and microme-
NT. Bone imaging in metastatic breast cancer. J Clin Oncol. tastases may be limited. Hell J Nucl Med. 2014;17:177–83.
2004;22:2924–53. 111. Bourgeois AC, Warren LA, Chang TT, Embry S, Hudson K,
93. Plunkett TA, Smith P, Rubens RD. Risk of complications from Bradley YC. Role of positron emission tomography/computed
bone metastases in breast cancer: implications for management. tomography in breast cancer. Radiol Clin N Am. 2013;51:
Eur J Cancer. 2000;36:476–82. 781–98.
94. Pires A, Borges U, Lopes-Costa P, Gebrim L, da Silva B. Evaluation 112. Koolen BB, Valdes Olmos RA, Elkhuizen PH, Vogel WV,
of bone metastases from breast cancer by bone scintigraphy and Vrancken Peeters MJ, Rodenhuis S, et al. Locoregional lymph
positron emission tomography/computed tomography imaging. node involvement on 18F-FDG PET/CT in breast cancer patients
Eur J Obstet Gynecol Reprod Biol. 2014;180:138–41. scheduled for neoadjuvant chemotherapy. Breast Cancer Res
95. Sathiakumar N, Delzell E, Morrisey M, Falkson C, Yong M, Chia Treat. 2012;135:231–40.
V, et al. Mortality following bone metastasis and skeletal-related 113. Young H, Baum R, Cremerius U, Herholz K, Hoekstra O,
events among women with breast cancer: a population-based anal- Lammertsma AA, et al. Measurement of clinical and subclini-
ysis of U.S. Medicare beneficiaries, 1999–2006. Breast Cancer cal tumour response using [18F]-fluorodeoxyglucose and positron
Res Treat. 2012;131:231–8. emission tomography: review and 1999 EORTC recommendations.
96. Domchek SM, Younger J, Finkelstein DM, Seiden MV. Predictors European Organization for Research and Treatment of Cancer
of skeletal complications in patients with metastatic breast carci- (EORTC) PET Study Group. Eur J Cancer. 1999;35:1773–82.
noma. Cancer. 2000;89:363–8. 114. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to
97. Van den Wyngaert T, Strobel K, Kampen WU, Kuwert T, van PERCIST: evolving considerations for PET response criteria in
der Bruggen W, Mohan HK, et al. The EANM practice guide- solid tumors. J Nucl Med. 2009;50(Suppl 1):122S–50S.
lines for bone scintigraphy. Eur J Nucl Med Mol Imaging. 115. Wang Y, Zhang C, Liu J, Huang G. Is 18F-FDG PET accurate to
2016;43:1723–38. predict neoadjuvant therapy response in breast cancer? A meta-
98. Hannequin P, Mas J. Statistical and heuristic noise extraction analysis. Breast Cancer Res Treat. 2012;131:357–69.
(SHINE): a new method for processing Poisson noise in scinti- 116. Avril N, Sassen S, Roylance R. Response to therapy in breast can-
graphic images. Phys Med Biol. 2002;47:4329–44. cer. J Nucl Med. 2009;50:55S–63S.
117. Rousseau C, Devillers A, Sagan C, Ferrer L, Bridji B, Campion L, 127. Collins LC, Cole KS, Marotti JD, Hu R, Schnitt SJ, Tamimi
et al. Monitoring of early response to neoadjuvant chemotherapy RM. Androgen receptor expression in breast cancer in relation to
in stage II and III breast cancer by [18F]fluorodeoxyglucose posi- molecular phenotype: results from the Nurses’ Health Study. Mod
tron emission tomography. J Clin Oncol. 2006;24:5366–72. Pathol. 2011;24:924–31.
118. Chalkidou A, Landau DB, Odell EW, Cornelius VR, O’Doherty 128. Gucalp A, Tolaney S, Isakoff SJ, Ingle JN, Liu MC, Carey LA,
MJ, Marsden PK. Correlation between Ki-67 immunohisto- et al. Phase II trial of bicalutamide in patients with androgen
chemistry and 18F-fluorothymidine uptake in patients with receptor-positive, estrogen receptor-negative metastatic breast
cancer: a systematic review and meta-analysis. Eur J Cancer. cancer. Clin Cancer Res. 2013;19:5505–12.
2012;48:3499–513. 129. Fowler AM, Chan SR, Sharp TL, Fettig NM, Zhou D, Dence CS,
119. Kostakoglu L, Duan F, Idowu MO, Jolles PR, Bear HD, Muzi et al. Small-animal PET of steroid hormone receptors predicts
M, et al. A phase II study of 3′-deoxy-3′-18F-fluorothymidine tumor response to endocrine therapy using a preclinical model of
PET in the assessment of early response of breast cancer to neo- breast cancer. J Nucl Med. 2012;53:1119–26.
adjuvant chemotherapy: results from ACRIN 6688. J Nucl Med. 130. Slamon D, Eiermann W, Robert N, Pienkowski T, Martin M, Press
2015;56:1681–9. M, et al. Breast Cancer International Research Group. Adjuvant
120. Lindholm P, Lapela M, Någren K, Lehikoinen P, Minn H, Jyrkkiö trastuzumab in HER2-positive breast cancer. N Engl J Med.
S. Preliminary study of carbon-11 methionine PET in the evalua- 2011;365:1273–83.
tion of early response to therapy in advanced breast cancer. Nucl 131. Dijkers EC, Oude Munnink TH, Kosterink JG, Brouwers AH,
Med Commun. 2009;30:30–6. Jager PL, de Jong JR, et al. Biodistribution of 89Zr-trastuzumab
121. Linden HM, Dehdashti F. Novel methods and tracers for breast and PET imaging of HER2-positive lesions in patients with
cancer imaging. Semin Nucl Med. 2013;43:324–9. metastatic breast cancer. Clin Pharmacol Ther. 2010;87:
122. Tade FI, Cohen MA, Styblo TM, et al. Anti-3-18F-FACBC 586–92.
(18F-fluciclovine) PET/CT of breast cancer: an exploratory study. 132. Tamura K, Kurihara H, Yonemori, K, Takahashi K, Wada Y,
J Nucl Med. 2016;57:1357–63. Hasegawa K, et al. 64Cu-DOTA-trastuzumab-PET imaging in
123. Ulaner GA, Goldman DA, Gönen M, et al. Initial results of a pro- patients with HER2-positive breast cancer [abstract]. J Clin
spective clinical trial of 18F-fluciclovine PET/CT in newly diag- Oncol. 2012;Suppl:30.
nosed invasive ductal and invasive lobular breast cancers. J Nucl 133. Quick HH. Integrated PET/MR. J Magn Reson Imaging.
Med. 2016;57:1350–6. 2014;39:243–58.. Erratum in: J Magn Reson Imaging.
124. Mankoff DA, Dunnwald LK, Gralow JR, et al. Blood flow and 2014;39:1341
metabolism in locally advanced breast cancer: relationship to 134. Dregely I, Lanz T, Metz S, Mueller MF, Kuschan M, Nimbalkar
response to therapy. J Nucl Med. 2002;43:500–9. M, et al. A 16-channel MR coil for simultaneous PET/MR imag-
125. Peterson LM, Mankoff DA, Lawton T, Yagle K, Schubert EK, ing in breast cancer. Eur Radiol. 2015;25:1154–61.
Stekhova S, et al. Quantitative imaging of estrogen receptor 135. Oehmigen M, Lindemann ME, Lanz T, Kinner S, Quick
expression in breast cancer with PET and 18F-fluoroestradiol. J HH. Integrated PET/MR breast cancer imaging: attenuation cor-
Nucl Med. 2008;49:367–74. rection and implementation of a 16-channel RF coil. Med Phys.
126. van Kruchten M, Glaudemans AW, de Vries EF, Beets-Tan RG, 2016;43:4808.
Schroder CP, Dierckx RA, et al. PET imaging of estrogen recep- 136. Spick C, Herrmann K, Czernin J. 18F-FDG PET/CT and PET/MRI
tors as a diagnostic tool for breast cancer patients presenting with perform equally well in cancer: evidence from studies on more
a clinical dilemma. J Nucl Med. 2012;53:182–90. than 2,300 patients. J Nucl Med. 2016;57:420–30.
Hybrid Imaging and Radionuclide
Therapy of Musculoskeletal Diseases 24
Paola Anna Erba, Martina Sollini, Roberta Zanca,
Roberto Boni, Lesley Flynt, Elena Lazzeri,
Giuliano Mariani, and Torsten Kuwert
Contents
24.1 Structure and Functions of the Skeletal System 572
24.1.1 Bone Structure: Osteoblasts, Osteoclasts, and Hormonal Control 572
24.1.2 Bone Remodeling 573
24.1.3 Bone Development and Changes with Age 573
24.2 Radiopharmaceuticals for the Evaluation of the Skeletal System 573
24.2.1 99m
Tc-Diphosphonates 573
24.2.2 Quantitative Bone SPECT/CT 581
24.2.3 F-Fluoride PET
18
581
24.2.4 Perspectives for Radiolabeled Diphosphonates 584
24.3 Clinical Applications 584
24.3.1 Primary Bone Tumors 584
24.3.2 Fibrous Bone Dysplasia 588
24.3.3 Metastatic Bone Tumors 590
24.4 Benign Bone Disease 595
24.4.1 Workup of Bone Pain 595
24.5 Metabolic Bone Diseases 599
24.5.1 Osteoporosis 599
24.5.2 Osteomalacia 602
24.5.3 Primary and Secondary Hyperparathyroidism 603
24.5.4 Renal Osteodystrophy 604
24.5.5 Hypertrophic Osteoarthropathy 604
24.6 Trauma and Fractures 605
24.6.1 Occult Fractures or Trauma 605
24.6.2 Stress Fractures 605
P. A. Erba · R. Zanca · G. Mariani E. Lazzeri

Regional Center of Nuclear Medicine, Department of Translational Regional Center of Nuclear Medicine, University Hospital of Pisa,
Research and Advanced Technologies in Medicine and Surgery, Pisa, Italy
T. Kuwert (*)
M. Sollini Clinic of Nuclear Medicine, Friedrich-Alexander-University of
Department of Biomedical Sciences, Humanitas University, Erlangen-Nürnberg, Erlangen, Germany
Milan, Italy e-mail: Torsten.Kuwert@uk-erlangen.de
R. Boni
Nuclear Medicine Service, “Papa Giovanni XXIII” Hospital,
Bergamo, Italy
L. Flynt
Department of Nuclear Medicine, M.D. Anderson Cancer Center,
Houston, TX, USA

572 P. A. Erba et al.
24.7 Vascular Disorders 608

24.7.1 Aseptic Necrosis of the Femoral Head 608
24.7.2 Sympathetic Reflex Dystrophy 609
24.8 Inflammation 610
24.8.1 Rheumatoid Arthritis 610
24.8.2 Seronegative Spondyloarthropathies 611
24.9 Infection 613
24.9.1 Radionuclide Imaging of Infection/Inflammation 614
24.9.2 Spondylodiscitis 620
24.9.3 Infection/Inflammation of Prosthetic Joint Implants 623
24.9.4 Non-radionuclide Imaging for Infection/Inflammation of Prosthetic Joint Implants 624
24.9.5 Radionuclide Imaging for Infection/Inflammation of Prosthetic Joint Implants 624
24.10 Paget’s Disease of the Bone 628
24.11 Treatment of Skeletal Metastases with Bone-Seeking Radiopharmaceuticals 629
24.11.1 Generalities on Metastatic Disease to the Skeleton 629
24.11.2 Pathophysiologic Bases for the Use of Bone-Seeking Radiopharmaceuticals 631
24.11.3 Bone-Seeking Radiopharmaceuticals for Therapy 632
24.11.4 Clinical Use of β− Particle-Emitting Bone-Seeking Radiopharmaceuticals 633
24.11.5 Clinical Use of the α++ Emitter 223Ra-dichloride 637
References 639
Learning Objectives 24.1 Structure and Functions

• To understand the principles of bone physiology and of the Skeletal System
metabolism.
• To understand bone remodeling and bone development. Bone tissue is a specialized form of connective tissue, which
• To learn the protocols used for bone scintigraphy, includ- allows deposition of calcium and phosphate in its extracel-
ing SPECT/CT. lular matrix conferring remarkable hardness and strength. It
• To become acquainted with how a bone scintigram should fulfills three main functions:
be interpreted.
• To become familiar with quantitative skeletal SPECT/ • Support of the body and as a site for the insertion of skel-
CT. etal muscles.
• To learn the principles of 18F-fluoride PET and of poten- • Protection of major organs and bone marrow.
tial new tracers of bone metabolism. • Storage of calcium and phosphate, with a central role in
• To learn how bone scintigraphy can be used in the diag- maintaining calcium homeostasis in body fluids.
nostic workup of primary bone tumors.
• To learn how bone scintigraphy can be used to stage
malignant disease. 24.1.1 Bone Structure: Osteoblasts,
• To understand how bone scintigraphy in conjunction with Osteoclasts, and Hormonal Control
SPECT/CT can be used in the diagnostic workup of bone
pain. The cells involved in the bone formation process are called
• To learn the scintigraphic appearance of benign bone dis- osteoblasts. Their differentiation from stromal progenitor
ease such as metabolic bone disease, osteonecrosis, frac- cells (originally mesenchymal cells from the mesodermal
tures, inflammation, and infection and how bone germ line) is stimulated by factors such as parathyroid hor-
scintigraphy can be used in the diagnostic workup of mone (PTH), prostaglandin, and some growth factors. The
these conditions. main function of the osteoblasts that produce many molecules
• To understand the rationale of using bone-seeking radio- (e.g., TGF, IGFs, PDGF) is to synthesize the bone matrix—an
pharmaceuticals labeled with radionuclides emitting elec- organic backbone. After its synthesis, the bone matrix is min-
trically charged particles for treatment of painful bone eralized and constantly regenerated as a result of the continu-
metastases. ous bone remodeling process. The main constituent of the
• To learn the principles and procedures of employing bone matrix is type I collagen, although minimal amounts of
bone-seeking radiopharmaceuticals labeled with radionu- collagen type III, V, and FACIT are involved; non-collagen
clides emitting β− or α++ particles for treating patients proteins account for about 10–15% of the bone matrix. The
with metastatic bone disease. extracellular matrix, and especially type I collagen that guar-
24 Hybrid Imaging and Radionuclide Therapy of Musculoskeletal Diseases 573
antees elasticity and flexibility, determines bone structure. their structural and functional differences, cortical bone and tra-
Growth of the mineral component (initially localized in spe- becular bone are composed of the same cellular elements and
cific sites of the collagen matrix, in the spaces between fibers) extracellular matrix. The main structural difference is quantita-
is facilitated by high levels of calcium and phosphorus. tive: 80–90% of compact bone is calcified, while the calcified
Osteoclasts are plurinuclear giant cells responsible for initi- component accounts for only 15–25% in the trabecular bone.
ating the bone remodeling process. They are produced by the This difference in composition mirrors functional differences:
bone marrow, and mature cells act under the effect of PTH and cortical bone has purely protective and structural functions,
other local substances, such as tumor growth factor (TGF), while the main function of the trabecular bone is metabolic.
tumor necrosis factor (TNF), interleukin-1 (IL-1), and interleu-
kin-6 (IL-6). Vitamin D (1,25-dihydroxycholecalciferol) is a
powerful stimulator of osteoclastic activity, promoting the dif- 24.1.3 Bone Development and Changes
ferentiation of osteoclast precursors and inhibiting proliferation with Age
of T cells via interleukin-2 (IL-2); calcitonin (CT) inhibits the
action of mature osteoclasts. Other local factors are involved in The volume of cortical bone (mainly present in the long
the control of bone remodeling, IL-1 being one of the most pow- bones and the appendicular skeleton) is regulated by the for-
erful stimuli for osteoclasts (although its effect is mediated, at mation of the periosteal bone and the remodeling of both
least in part, by PGs). Another important cytokine involved in Havers channels and endosteal bone. In females, the loss of
this complex process is IL-6, which is secreted by bone cells in cortical bone (which is the main factor predisposing to verte-
response to PTH and vitamin D. Osteoprotegerin (OPG) acts as bral and femoral fractures) begins after 40 years and is accel-
an antagonist on the bone remodeling process of osteoclasts. erated 5–10 years after menopause. The loss of trabecular
The sequence of the cellular events that accompany bone tissue bone occurs only after menopause. Bone remodeling begins
loss is at the base of the bone remodeling processes that takes with the activation of the osteoclast precursors and is fol-
place both during aging and in some pathological conditions lowed by the formation of “mature” osteoclasts that are usu-
(e.g., osteoporosis, fractures, myeloma, metastasis). ally grouped in small pits (the Howship lacunae) on bone
Cortical bone and trabecular bone can be considered as surfaces; it ends with the apoptosis (programmed cell death)
two distinct entities, each characterized by specific modifica- of osteoclasts. Osteoclasts’ apoptosis causes some osteoblas-
tions at each stage of bone remodeling, depending on the tic cell modifications (chemotaxis, proliferation, and differ-
environment in which bone cells are. entiation) which promote bone mineralization that ends with
discontinuation of osteoblast activity (usually with complete
regeneration of the reabsorption lacunae).
24.1.2 Bone Remodeling
Bone remodeling is a continuous process of renewal of the

skeleton that occurs throughout life, with the aim of preserv- Key Learning Points
ing its mechanical integrity. This process involves continuous • Bone remodeling is a continuous process of renewal
removal of bone tissue (bone reabsorption) followed by the of the skeleton.
synthesis of new bone matrix and subsequent mineralization • It involves bone reabsorption and new bone forma-
(bone formation). The removal of the “old” bone by the osteo- tion by the osteoclasts and osteoblasts, respectively.
clasts implies the release of calcium and other components of • After the age of 40, loss of bone occurs, consider-
the matrix into the serum. Bone remodeling involves two ably accelerated in females post menopause.
types of cells, osteoclasts and osteoblasts, the interaction of
which is finely balanced; minimal alterations of this balance
may result in loss of bone matrix or, rarely, increase in the
bone mass. 24.2 Radiopharmaceuticals
The outer portion of the bone is a thick layer of calcified tis- for the Evaluation of the Skeletal
sue, the cortical bone (i.e., compact bone), located in the diaphy- System
sis where the hematopoietic bone marrow is contained. Toward
metaphyses and epiphyses, the cortical bone becomes thinner, 24.2.1 99mTc-Diphosphonates
and the medullary channel is occupied by trabecular structures
thin and calcified that form the spongy bone (i.e., trabecular or Bone uptake and retention of radiopharmaceuticals belong-
cancellous bone). The space circumscribed by trabeculae con- ing to this family (see Chap. 2) is related to calcium content:
tains hematopoietic elements in direct communication with soft tissues have a low calcium level (0.005%) and exhibit
those contained in the medullary cavity of the diaphysis. Despite only weak uptake, while bones have a high calcium content
(14–24%) and exhibit high uptake. Other factors determin- low-energy general-purpose collimators may be used [1]. A
ing the degree of radiopharmaceutical uptake in bones high-resolution pinhole collimator can be used to evaluate
include bone perfusion, the nature of calcium phosphate small anatomical details (e.g., the femoral head), particularly
deposits, including size, hydration status, Ca/P ratio, and the in children.
osteoblastic/osteoclastic metabolic activity. Trabecular bone Total-body imaging (see example in Fig. 24.1): Acquisition
has a higher retention index than cortical bone. Therefore, of anterior and posterior views with a matrix of 256 × 1024
the femur (with its thick cortical component) has a lower or 512 × 2048, zoom factor 1, scanning speed between 10
retention index than the ribs. Retention in metaphyseal bone and 15 cm/min (adjusted to obtain more than 1.5 million
(with 14.3% calcium content, rich vascularization, and high
metabolic activity) is 0.77% dose/g at 3 h, higher than 0.49%
dose/g in diaphyseal bone (which has 23.9% calcium content
but is less vascularized). Uptake of bone-seeking radiophar-
maceuticals is bound to adequate blood perfusion, so that
avascular necrosis appears on the bone scan as a “cold” spot
in the early stage (1–2 weeks). Only after the beginning of
bone remodeling, the previously “cold” area shows radio-
pharmaceutical uptake. Increased blood flow results in high
radiopharmaceutical uptake.
24.2.1.1 Protocols for 99mTc-Diphosphonate

Scintigraphy
Current EANM guidelines provide an overview of the proto-
cols used for radionuclide imaging of bone metabolism [1].
The average injected activity for bone scintigraphy in an
adult person is 500 MBq (13 mCi), and the skeleton receives
the highest radiation exposure. In children, the administered
activity must be based on body weight according to the
EANM/SNMMI Pediatric Dosage Harmonization Working
Group [2, 3]; however, a minimum activity of 40 MBq is
required to obtain images that can be interpreted correctly.
Images are usually acquired about 3 h after radiopharma-
ceutical injection, except for the three-phase bone scan typi-
cally performed in case of a clinical suspicion of infection or
inflammation (e.g., osteomyelitis, prosthesis loosening, or
infection). The three-phase bone scan requires dynamic
acquisition of a rapid sequence of images of a specific region
of interest, which yields a vascular (or blood flow) phase
image, performed simultaneous to the i.v. injection of the
radiopharmaceutical. This is followed by blood pool phase
images acquired about 1–10 min after injection; finally, about
3 h later, a whole-body scan and planar (and possibly SPECT/
CT) images of specific regions of interest are acquired.
In the interval between radiopharmaceutical injection and
Fig. 24.1 Standard images acquired for planar whole-body skeletal
late image acquisition, the patient should be hydrated with at scintigraphy about 3 h after the i.v. administration of 99mTc-HDP. The
least 1.5 L of water to promote urinary excretion of the frac- left panel depicts the anterior and posterior views obtained in the whole-
tion of radiopharmaceutical not adsorbed to the skeleton, and body imaging mode (with arms along the body). Tracer distribution
throughout the skeleton only shows diffuse mild inhomogeneities (e.g.,
to enhance scintigraphic contrast as well as the bone-to-soft- in lower portion of thoracic spine), with mildly increased uptake cor-
tissue ratio. Overall quality of the images can be affected by responding to the anterior arc of the second left rib. Furthermore, there
other factors, such as renal and/or heart failure, obesity, and is apparent focally increased tracer uptake at the posterior tract of the
advanced age. last ribs, due to overlapping of these bone segments with the kidneys,
site of physiologic excretion. Radioactivity accumulation in the urinary
Images are acquired with a single- or double-headed bladder is clearly recognized (modified from: Volterrani D, Erba PA,
gamma camera equipped with low-energy high-resolution Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
parallel-hole collimators (recommended). Alternatively, Applicazioni. Milan: Springer Italy; 2010)
counts/image). If available, a body contour system should be 256 × 256, zoom factor 1.33, and a predefined acquisition
used, as it automatically changes position of the gamma time (4–10 min) or predefined number of counts. When using
camera head to maintain a minimum distance between the a predefined number of counts, this should be determined
collimator and the patient. Without automatic contouring, based on the region of interest (Table 24.1) depending on the
the distance between the head of the gamma camera and the field of view (FOV) of the gamma camera: the larger the
patient should be maintained as close as possible. Total-body FOV, the greater the number of total counts required to yield
images can be obtained in two different modes (often both similar count densities over equivalent regions of the skele-
options are available with the same scanner): ton [1].
Planar imaging may be integrated with images obtained
• Sequential images: the head of the gamma camera col- using a pinhole collimator (50,000–100,000 total counts) to
lects an image, then it moves to the adjacent district, and evaluate small structures or to better visualize some details
the acquisition/processing system provides at the end a (Fig. 24.3).
complete picture of the skeleton. SPECT or SPECT/CT acquisition: The region of interest
• Continuous images (preferred in adults). should be placed in the center of the FOV of the gamma cam-
era. SPECT imaging should be performed as recommended
When a total-body acquisition system is not available,
multiple planar images may be acquired, paying attention to Table 24.1 Recommended number of counts per region of interest
overlap each spot to the next one, in order not to miss any according to the EANM practice guidelines for bone scintigraphy
area that could be the site(s) of pathological findings. Region of interest Predefined number of counts
Planar images of regions of interest (see example in Skull and large joints 250,000–400,000
Fig. 24.2): The anterior, posterior, oblique, or lateral views Thoracoabdominal region 700,000–1,000,000
are generally acquired with a matrix of 128 × 128 or Distal joints 150,000–250,000
Fig. 24.2 Planar spot images

obtained in the same patient
as in Fig. 24.1. The top panel
depicts the anterior and
posterior views of the chest
that are acquired with the
arms raised in order to resolve
in part overlapping of
anatomic structures in the
upper torso—particularly in
the posterior view. Irregular
tracer uptake in the
cartilaginous tract of ribs
adjacent to the sternum
reflects partial calcification of
the sterno-chondral tracts.
The bottom panel depicts the
right lateral and left lateral
views of the chest. There is
some scintigraphic
visualization of the kidneys,
projecting almost to overlap
the spine (modified from:
Volterrani D, Erba PA,
Milan: Springer Italy; 2010)
anterior, posterior, oblique, and/or lateral views (matrix

128 × 128, zoom factor 1.33), with SPECT (preferably
SPECT/CT) acquisitions, or with images acquired with a
pinhole collimator.
24.2.1.2 Post-acquisition Processing for

99m
Tc-Diphosphonate Scintigraphy
No specific image processing is required for whole-body and
planar images, except for setting the display contrast to visu-
alize the low-count segments. Processing is more complex
Fig. 24.3 High-resolution acquisitions using a pinhole collimator of for dynamic three-phase scintigraphy (Fig. 24.4). The blood
the two hips obtained about 3 h after administration of 99mTc-HDP in a flow phase is processed by drawing two regions of interest
16-year-old boy with recent avascular necrosis of the left femoral head.
Acquisition with a pinhole collimator results in the production of a (ROIs), respectively, over the suspected area and over a ref-
circle-shaped image; the two images acquired separately for each hip erence region—typically the contralateral, nonaffected seg-
are positioned side-by-side. The right hip (indicated by gray arrow) ment. The corresponding activity/time curves so obtained
shows a normal pattern of tracer distribution, where the growth plates show the regional blood flow, which increases in case of
are clearly visible (due to young age of the patient); in the left hip, there
is a focal area of reduced tracer uptake, corresponding to the non- acute inflammation/infection. For the blood pool images, the
perfused bone. At later stages after onset of the event, bone remodeling degree of radiopharmaceutical accumulation in the area of
in the left hip during healing will result in increased tracer uptake interest (which is compared to the contralateral one) reflects
(reproduced with permission from: Volterrani D, Erba PA, Mariani G, the degree of capillary permeability, which is abnormally
Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
Milan: Springer Italy; 2010) increased in case of inflammation/infection. In the delayed
images, any area of increased tracer uptake resulting from
bone remodeling is evaluated.
by the gamma camera manufacturer [1]. In a typical acquisi- In case of SPECT and SPECT/CT acquisitions, parame-
tion protocol for a dual-headed gamma camera with the ters for image reconstruction may vary for different
detector heads oriented in a 180°geometry using a step-and- manufacturers.
shot modality, a total of 60 or 64 frames per detector head, Images are a crucial component of the examination; there-
each with a duration of 10–30 s, are acquired over 360° into fore, they should be provided to the patient and to the refer-
a 128 × 128 matrix (pixel size 4.6 × 4.6 mm). The acquisition ring physician together with the final medical report.
time should be increased to 30–40 s per angular view when
imaging regions with low counts (e.g., the skull). An equiva-
lent total number of counts should be acquired if continuous
Key Learning Points
acquisition is used. Regarding the CT component of a
• Bone uptake of 99mTc-diphosphonates depends on
SPECT/CT acquisition, particular attention should be paid to
perfusion, bone calcium content, and the osteoblas-
patient positioning, since the region of interest must be
tic/osteoclastic metabolic activity.
entirely included in the space delimited by the specific points
• Whole-body images of bone metabolism are usu-
on the imaging table.
ally acquired about 3 h after radiopharmaceutical
Three-phase bone scan acquisition (see example in
injection.
Fig. 24.4):
• In case of infection/inflammation, the delayed
images are complemented by images of blood flow
• Vascular or blood flow phase: dynamic acquisition (30–
obtained directly after tracer injection and of blood
60 frames of 1–2 s) of the region of interest starts simul-
pool acquired about 5 min later.
taneously with the i.v. injection of the
• Planar scans may be complemented by SPECT or
SPECT/CT acquisitions.
• Blood pool phase: planar images of the region of interest
(3–5 min, matrix size 128 × 128 or 256 × 256, zoom fac-
tor 1.33) are acquired within 10 min of tracer injection,
preferably at 5 min. 24.2.1.3 Interpretation Criteria
• Delayed phase: acquisition of the whole body in anterior The most important finding of a normal bone scan is its left-
and posterior views about 3 h after tracer administration. to-right symmetry. The bone segments more exposed to phys-
ical stress (e.g., sacroiliac region, vertebrae, major joints)
Whole-body images may be implemented with additional undergo more active bone remodeling and can therefore
planar images of the specific region of interest including appear as “hot” spots relative to the remainder of the skeleton
Fig. 24.4 Summary information derived from a three-phase bone scin- acquisition at 5 min post-injection—the so-called “blood pool” phase;
tigraphy performed in a patient with avascular necrosis of left femoral there is obviously increased radioactivity accumulation at the left hip,
head evaluated in the subacute phase. The eight images in top panel are corresponding to increased capillary permeability. The two images in
selected frames from the dynamic acquisition, showing the tracer still in lower panel depict the static anterior and posterior views acquired about
the intravascular phase immediately upon its i.v. bolus injection; the 3 h post-injection; markedly enhanced bone remodeling in left hip,
aortoiliac bifurcation is clearly recognized. The activity/time curves in where the core of the femoral head still exhibits reduced metabolic
right middle panel correspond to two ROIs defined over the two femoral activity and is surrounded by active bone remodeling (modified from:
heads; the higher curve corresponds to the left femoral head, where Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
blood perfusion is now enhanced versus the right femoral head—cor- Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
responding to the recovery phase. The left middle panel depicts static
a b c d
Fig. 24.5 Planar whole-body skeletal scans in children and adoles- preadolescent child in the 10–12-year age range (radioactivity accumu-
cents. Anterior (a) and posterior (b) views obtained in a child in the lation at the right wrist is due to some residual tracer remaining at the
6–8-year age range. Anterior (c) and posterior (d) views obtained in a injection port)
(Fig. 24.1). Similarly, areas with active bone remodeling dur-• Chest: the joint between the manubrium and body of the
ing growth in children and adolescents exhibit markedly sternum may exhibit increased tracer uptake, as also the
increased uptake on the bone scan (Fig. 24.5). Particular sternoclavicular joints. In elderly people it is common to
attention should be paid to the evaluation of some districts: observe some tracer uptake in the calcified rib cartilages.
The ribs normally exhibit relatively low tracer uptake,
• Skull: counts are relatively low. Cranial suture lines are although insertion of the erector spinae muscles can cause
visualized in approximately 1% of the adults. In the ante- uptake inhomogeneity (physiological variant in about 7%
rior view, orbital and facial bones are well visualized; of subjects). The scapulae are well visualized, with
some uptake may be present at maxilla and ethmoid, as increased activity at the lower tip, where the mechanical
result of chewing-induced stress. In childhood, the suture stress associated with muscle insertion is greater.
between the sphenoid and the occipital bone may be par- • Spine: the upper thoracic vertebrae are not displayed indi-
ticularly evident. vidually, whereas the lower thoracic and lumbar vertebrae
• Neck: the cervical vertebrae can be distinguished only are well seen. The physiological lumbar lordosis often
using high-resolution acquisitions; the spinous process of makes the lumbar vertebrae to seem relatively “hot” in the
C7 is usually well visualized. In physiologic conditions, anterior view.
tracer uptake in the neck (in the absence of free 99mTc- • Pelvis: tracer uptake is usually symmetric between the
pertechnetate) can be observed in the thyroid cartilage or right and left sides; even if the patient is asked to void the
the hyoid bone. bladder immediately before acquiring the scan, it is com-
mon to observe some residual accumulation of activity in

the bladder, which may mask the pubic bone/symphysis
and sacrum.
• Upper and lower limbs: there is relatively low activity. A
relatively increased tracer uptake is typically seen in the
periarticular regions; in children the presence of increased
tracer uptake in the epiphysis of long bones (growth plate)
is common as it reflects the physiologic process of bone
growth (see Fig. 24.5).
Scintigraphic images depict the extent of bone remodel-

ing, mostly with reference to the mineral component. In fact,
the degree of tracer uptake is mainly related to increased
osteoblastic activity. Osteolytic lesions (most often caused
by bone metastases from various cancers) are almost invari-
ably associated with an osteoblastic reaction, which
(although insufficient to counterbalance osteolysis) is gener-
ally sufficient to induce increased tracer uptake on the bone
scan. However, the uptake of bone-seeking radiopharmaceu-
ticals is related to several physiological and pathological
conditions in the absence of significant osteolysis: bone frac-
ture (increased tracer uptake may last for years), osteoarthri-
tis, Paget’s disease of the bone, and so on. “Pure” osteolytic
lesions without osteoblastic reaction (such as those fre-
quently caused by myeloma and some metastases from renal
cancers) are generally negative on the bone scan. Based on
these considerations, an accurate medical history aimed at
identifying prior traumas, fractures, surgical interventions,
Fig. 24.6 Planar whole-body scan (anterior and posterior views)
or other concomitant diseases should be obtained before per- obtained in a morbidly obese woman about 3 h after i.v. administration
forming and evaluating bone scintigraphy. of 99mTc-HDP. Scintigraphic visualization of the skeleton is markedly
In certain clinical situations, the normal contrast impaired both because of low bone-to-soft-tissue ratios and because of
between healthy and pathological tissues (as well as attenuation of γ-rays by the overabundant adipose tissue, particularly in
the anterior view. Increased tracer uptake can be noticed at the knees,
between the skeleton itself and soft tissues) is reduced as a most likely due to degenerative osteoarticular disease caused by the
consequence of an increased background activity due to increased workload on the joints associated with obesity (reproduced
altered distribution and/or excretion of the tracer, for with permission from: Volterrani D, Erba PA, Mariani G, Eds.
example, in case of renal failure or obesity (Fig. 24.6). Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
Springer Italy; 2010)
Altered biodistribution of the radiopharmaceutical can
also result from an error during administration, as it occurs
in case of partial tracer extravasation out of a vein; in this In three-phase bone scintigraphy, the activity/time curve
case a focal area of uptake is observed at the injection site, of the pathological segment reaches a higher peak than the
and it may be associated with visualization of one or more contralateral side when blood flow is increased, while the
(axillary) lymph nodes on the same side. Such findings are peak is lower in case of reduced perfusion. Increased radioac-
generally well recognized as non-pathological; however, tivity accumulation in the blood pool image is a consequence
in case of nontypical sites of injection, such “hot” spot of increased capillary permeability (with possible local
areas may be a potential source of misinterpretation of edema) associated with inflammation/infection, while
images. Although rarely happening, there is the possibility increased tracer uptake observed in the delayed phase denotes
that the radiopharmaceutical is erroneously injected in an an active process of bone remodeling. Three-phase bone scin-
artery rather than in a vein; in this case, scintigraphy dif- tigraphy is generally used to confirm lesion(s) characterized
fers considerably from the normal pattern, resulting in by hyper-vascularization and increased bone remodeling. A
intense tracer accumulation in the portion of the arm distal typical example is seen in acute osteomyelitis, where increased
from the injection site, where bone/soft tissue contrast is vascularization (accumulation in the blood flow phase) coex-
remarkably low due to persisting high background activity ists with increased capillary permeability (accumulation in
(Fig. 24.7). blood pool images) and with increased osteoblastic activity
(uptake in delayed images). The early phase of aseptic (avas- In conclusion, the possible sources of error when inter-
cular) necrosis of the femoral head is characterized by hypo- preting a bone scan include focal soft tissue spot (e.g., intra-
vascularization (hypoperfusion in the blood flow phase) muscular injections, hematoma, severe renal failure,
visible as a “cold” area with an intense peripheral uptake in hypercalcemia), partial tracer extravasation, attenuation arti-
the delayed images, the latter reflecting repair and bone facts caused by metal prostheses, motion artifacts, urinary
remodeling. A three-phase bone scintigraphy is the investiga- contamination, superimposition of bladder activity, “pure”
tion of choice in patients with suspected prosthetic loosening lytic lesions, renal failure, and homogenously increased
with some probability of infection. bone activity (e.g., superscan) [1].
Table 24.2 summarizes the main clinical indications for
bone scintigraphy with 99mTc-diphosphonates [1].
Table 24.2 Summary of indications for bone scintigraphy with

99m
Tc-diphosphonates according to the EANM guidelinea
Oncology
• Solid tumors with high propensity for bone metastasis,
including prostate, breast, lung, and renal cancers
• Malignant hematological disease (lymphomas) limited to bone
• Primary bone tumors and bone dysplasia, including
osteosarcoma, osteoid osteoma, osteoblastoma, fibrous
dysplasia, giant cell tumor
• Soft tissue sarcomas, including rhabdomyosarcoma
• Paraneoplastic syndromes, including hypertrophic pulmonary
osteoarthropathy, algodystrophy, polymyalgia rheumatica,
poly-dermatomyositis, and oncogenic osteomalacia
• Distribution of osteoblastic activity before radionuclide therapy
with bone-seeking agents
Rheumatology
• Chronic inflammatory arthritis, including rheumatoid arthritis,
spondyloarthropathies and related disorders (ankylosing
spondylitis, psoriatic arthritis, Reiter’s arthritis, SAPHO
syndrome [synovitis, acne, pustulosis, hyperostosis, osteitis],
chronic recurrent multifocal osteomyelitis), and sacroiliitis
• Osteoarthritis of the lumbar facet joints and hip, femorotibial
and femoropatellar osteoarthritis, rhizarthrosis, tarsal
osteoarthritis
• Enthesopathies, including plantar fasciitis, Achilles tendinitis,
and bursitis
• (Avascular) Osteonecrosis, most frequently located at the
femoral head, femoral condyle, and tibial plateau
• Osteonecrosis of the jaw
• Complex regional pain syndrome type I of the hand, hip, knee,
and foot
• Tietze’s syndrome (costochondritis)
• Polymyositis
• Paget’s disease of bone
• Langerhans cell histiocytosis (LCH): single system LCH and
Fig. 24.7 Planar whole-body scan obtained about 3 h after administra- multisystem LCH with bone involvement
tion of 99mTc-MDP in a patient previously submitted to left nephrec- • Non-Langerhans cell diseases, such as Erdheim-Chester
tomy. The tracer had inadvertently been injected intra-arterially rather disease, Schnitzler syndrome, and Rosaï-Dorfman disease
than intravenously, resulting in markedly enhanced radioactivity accu- • Other rare osteoarticular diseases, including sarcoidosis with
mulation in the soft tissues distal to the point of intra-arterial injection, bone involvement, mastocytosis, Behçet’s disease, and familial
with a pattern resembling an “evening glove.” As an incidental finding, Mediterranean fever
the left lower limb (where lymphedema secondary to left nephrectomy Bone and joint infection
and lymphadenectomy is present) is increased in size and exhibits • Osteomyelitis (acute, subacute, or chronic, of bacterial,
enhanced radioactivity accumulation versus the right lower limb (repro- mycobacterial, or fungal origin)
duced with permission from: Volterrani D, Erba PA, Mariani G, Eds.
• Septic arthritis
• Spondylodiscitis or spondylitis
Springer Italy; 2010)
Table 24.2 (continued) relates with CT density. In principle, using quantitative

• Septic loosening or mechanical complication of internal fixation SPECT/CT, normal values of tracer uptake can be estab-
(long bones or spine) or arthroplasty (hip, knee, ankle, or lished for each skeletal region. These normal values might be
shoulder)
used to diagnose diffuse abnormalities of tracer uptake in
• Malignant (necrotizing) external otitis
disseminated bone disease. Furthermore, quantitative
Orthopedics, sports and traumas
• Periostitis, including shin splints and thigh splints SPECT/CT allows monitoring of the activity of osseous
• Enthesopathies, including plantar fasciitis, Achilles tendinitis, metastases in the course of treatment. Preliminary evidence
and bursitis indicates that this technology might also be useful to assess
• Spondylolisthesis (acute or subacute) the floridity of skeletal lesions more accurately than visual
• Radiologically occult stress-related fractures (e.g., scaphoid, evaluation of tracer uptake [5]. Future work will show how
tarsals) or nonspecific symptoms
this new potential of bone scintigraphy can be used with ben-
• Insufficiency fractures, including osteoporotic vertebral or
occult fractures, sacral fractures, femoral head or neck fractures,
efit in clinical routine.
tibial plateau fractures, and tarsal and metatarsal fractures
• Septic loosening, mechanical complication, and synovitis of
internal fixation (long bones or spine) or prosthesis (hip, knee,
ankle, or shoulder) Key Learning Points
• Pseudoarthrosis (delayed union, nonunion) • The most important finding of a normal bone scan
• Periarticular heterotopic ossification is its left-to-right symmetry.
• Viability of bone graft • The intensity of tracer uptake is mainly related to
Metabolic bone disease increased osteoblastic activity, which is present also
• Primary and secondary hyperparathyroidism
in most predominantly osteolytic metastases.
• Osteomalacia
• Purely lytic metastases, e.g., from myeloma, may
• Renal osteodystrophy
• Rare skeletal manifestations of endocrine disorders, including
escape detection at bone scintigraphy.
hyperthyroidism and acromegaly • Increased radioactivity accumulation in the blood
• Vitamin D deficiency pool image is a consequence of increased capillary
Pediatrics permeability associated with inflammation and/or
• Osteochondritis of the hip (Legg-Calvé-Perthes disease) infection.
• Transient synovitis of the hip • To date SPECT/CT cameras afford the determina-
• Osteoid osteoma tion of tissue radioactivity concentration expressed,
• Battered child syndrome
e.g., as SUV values, which is around 6 in vertebral
• Mandibular condylar hyperplasia
bone on the delayed images.
• Bone infarction (sickle cell disease, thalassemia)
Exploration of unexplained symptoms
• Subacute or chronic musculoskeletal or bone pain with normal
clinical examination and radiographs (arthralgia, monoarthritis,
oligoarthritis, polyarthritis, localized or multifocal bone pain,
backache) 24.2.3 18F-Fluoride PET
• Further exploration of abnormal biochemical (e.g., phosphate or
calcium metabolism) or radiological findings 18
F-Fluoride accumulates in bones with a mechanism similar
• Fever of unknown origin: exclusion of osteomyelitis to diphosphonates (see Chap. 3) and may therefore be used
a
It should be noted that this basically is a compilation of the skeletal to acquire PET images of the skeleton. Upon i.v. injection,
conditions that imply increased uptake of bone-seeking radiopharma- 18
F-fluoride diffuses quickly from the blood to reach the
ceuticals. Therefore, from the diagnostic point of view, virtually all
such conditions should be considered when interpreting a positive bone extracellular bone matrix. All 18F-fluoride that reaches the
scintigraphy, taking into account also the specific clinical scenario bone (about 50% of injected activity, similar to 99mTc-diphos-
based on which the referring physician ordered the bone scan phonates) is adsorbed. The exchange of 18F− ions with the
hydroxyl groups of hydroxyapatite crystals allows linking to
the surface of the bone matrix, while its retention at bone
24.2.2 Quantitative Bone SPECT/CT remodeling sites depends on subsequent migration inside the
crystalline bone matrix. Bone accumulation and blood clear-
To date SPECT/CT systems allow quantification of osseous ance are fast (with bi-exponential kinetics and half-lives of
radioactivity concentration in absolute units, such as kBq/ 0.4 h and 2.6 h, respectively), so that 1 h after administration
mL with an acceptable degree of accuracy. About 3 h after only 10% of the injected activity is still circulating. Therefore,
i.v. injection of 99mTc-diphosphonates, vertebral radioactivity a high bone/soft tissue ratio is reached within a short time,
concentration is in the range of 50 kBq/mL, translating into resulting in high-quality images as early as even <1 h after
standardized uptake values (SUV) of around 6 [4]. SUV cor- injection.
bed positions may be sufficient, while whole-body images

are required for all oncological indications [6] (see example
in Fig. 24.8). The acquisition time should be 3–5 min per bed
position, but it may be modified according to the scanner and
the injected activity; the 3D acquisition mode is generally set
as the standard operation mode in current PET scanners. In
general, the parameters used for [18F]FDG PET imaging can
be used also for 18F-fluoride PET, although sometimes it may
be necessary to use some filters to correct for the high con-
centration of activity in very small segments. Artifacts may
occur (most evident in sagittal and coronal sections) for the
thoracic spine (lungs attenuate less than soft tissues) when
no attenuation correction is applied. Attenuation correction
produces images with better uniformity than without attenu-
ation correction; the decision as to whether or not to correct
the images for attenuation varies according to the segment to
be explored. The residual bladder activity due to the urinary
excretion of 18F-fluoride may obscure the pelvis.
The overall time required for a 18F-fluoride PET scan is
much shorter than for conventional bone scintigraphy:
15–30 min waiting time after tracer injection (versus about
3 h for 99mTc-diphosphonates) and 15–30 min acquisition
time (versus at least 40 min when SPECT images are
acquired in addition to planar imaging). This faster acquisi-
tion time results in better patient’s compliance, thus reducing
artifacts due to patient’s movements.
The clinical indications for 18F-fluoride PET are the same
as for conventional bone scintigraphy. Generally, 18F-fluoride
Fig. 24.8 Patterns of normal distribution of 18F-fluoride visualized dur- PET identifies more lesions than scintigraphy with 99mTc-
ing PET, displayed as maximum intensity (MIP) attenuation-corrected
diphosphonates (especially if planar imaging only is
images. The adult skeleton (left) is characterized by well-evident tracer
uptake—relatively enhanced in the skull, ribs, spine, borders of the acquired), particularly in the axial skeleton, and reduces the
scapula, pelvis, and along the cortices of the long bones (the focal area number of doubtful/equivocal findings (see Fig. 24.9). Given
of activity accumulation in the right foot corresponds to some extrava- the extremely high sensitivity of 18F-fluoride PET in identi-
sation of the tracer at the injection site in a subject with difficult venous
fying skeletal areas with even mild changes in mineral
access). The pattern of 18F-fluoride distribution observed in the pediatric
skeleton (as in the 14-year-old girl represented here) shows, in addition, metabolism, interpretation of the scans requires caution and
linearly increased tracer uptake along the epiphyseal growth plates of close correlation with the CT counterpart of each focal area
the long bones, similar to a conventional 99mTc-MDP bone scan (image of increased tracer uptake.
of the normal adult scan kindly provided by Guofan Xu, MD, PhD,
Table 24.3 summarizes the main clinical indications for
Department of Nuclear Medicine, M.D. Anderson Cancer Center,
Houston, Texas, USA)
18
F-fluoride PET/CT according to the SNM guideline
released in 2010 [7].
Estimates of radiation burden to patients following 24.2.3.1 18F-Fluoride PET Acquisition Protocol
administration of 18F-fluoride or 99mTc-diphosphonates Uptake of 18F-NaF in the skeleton measures the amount of
(2.11 MBq/kg and 7.4 MBq/kg, respectively) indicate that in osteoblastic activity in actively mineralizing bone. This pro-
adults there are no significant differences between the two cess consists of 18F-fluoride being deposited on the hydroxy-
radiopharmaceuticals (4 mSv and 3 mSv, respectively, for a apatite surface of newly forming bone by exchanging fluoride
70 kg patient), whereas the effective dose is lower for 99mTc- with hydroxyl ions and creating fluorapatite. Standardized
diphosphonates (2 mSv) than for 18F-fluoride (about 3.5 mSv) uptake values (SUVs) are calculated on the basis of the activ-
in children weighing <20 kg. ity concentration (kBq/mL) normalized to injected activity
18
F-Fluoride PET imaging may be started 15–30 min after (MBq) and body weight (kg) of the patient. SUV of a lesion
the injection, acquiring a number of bed positions sufficient or a structure is often expressed as a maximum value
to cover the region of interest according to the clinical indi- (SUVmax) or mean value (SUVmean) [8, 9]. Alternatively, since
cation; if the region of interest is limited (e.g., the spine), 2–3 the rate-limiting step of tracer uptake is regional blood flow,
Fig. 24.9 Comparison of conventional 99mTc-

MDP bone scintigraphy (anterior and posterior
whole-body images on the left) and PET with
18
F-fluoride (MIP image on the right). Images
refer to a 69-year-old man with a history of
metastatic prostate cancer and rising serum PSA
levels. The conventional bone scintigraphy
revealed only irregularly increased tracer uptake
in the left parietal bone, with equivocal findings
in the thoracic spine. PET with 18F-fluoride
performed shortly after the 99mTc-MDP bone
revealed multifocal bone metastases, well evident
in the skull, in the thoracic spine, and in other
skeletal segments
Table 24.3 Summary of indications for PET/CT with 18

F-fluoride method calculates plasma clearance (Ki) using a three-
according to the SNM guidelinea compartment model to analyze the bone time-activity curve
Primary indication and the arterial input function [10]. Patlak graphical analysis
• Identification of skeletal metastases, including localization and is a simplification of the Hawkins method, calculating
determination of the extent of disease plasma clearance as the slope of normalized bone uptake
Indications without sufficient information, but most likely
appropriate
against normalized time [11].
• Back pain and otherwise unexplained bone pain Although plasma clearance measurements from dynamic
• Child abuse PET can accurately trace bone formation rate, practical
• Abnormal radiographic or laboratory findings issues limit its potential for routine clinical use. Patients
• Osteomyelitis must endure a 60 min image acquisition and undergo arterial
• Trauma blood sampling. The 60 min dynamic scan is also costly to
• Inflammatory and degenerative osteoarticular disease perform and can only examine one skeletal region at a time.
• Avascular necrosis Therefore, a whole-body assessment requires multiple scans
• Osteonecrosis of the mandible
over different sessions. Plasma clearance measurements
• Condylar hyperplasia
scans are frequently used, however, because they include
• Metabolic bone disease
• Paget’s disease of bone
tracer availability when calculating relative uptake among
• Bone graft viability competing skeletal sites.
• Complications of prosthetic joints
• Reflex sympathetic dystrophy
• Distribution of osteoblastic activity before radionuclide therapy Key Learning Points
with bone-seeking agent • F-Fluoride allows the visualization of bone metabo-
18
a
It should be noted that this guideline was issued in 2010. In the time lism in a manner similar as the 99mTc-diphosphonates.
elapsed since that date, a much more important body of evidence has
• 18F-Fluoride-PET offers higher signal-to-noise
been acquired on the clinical benefits of this imaging modality in a vast
variety of clinical conditions ratios and better spatial resolution than
99m
Tc-diphosphonate-SPECT.
• Waiting time after tracer injection is 15–30 min,
bone and plasma clearance measurements normalized to the much shorter for 18F-fluoride PET than for 99mTc-
ratio of uptake values to arterial tracer concentration are cal- diphosphonate SPECT.
culated as an index of new bone formation. The Hawkins
24.2.4 Perspectives for Radiolabeled 24.3 Clinical Applications

Diphosphonates
24.3.1 Primary Bone Tumors
The availability of clinical 68Ge/68Ga generators (68Ga
T1/2 = 67.7 min; high positron branching = 89%) provides new Malignant primary tumors of the bone usually feature highly
options to image bone metastatic disease with PET/CT. In increased bone metabolism as well as increased tracer uptake
fact, several compounds of biological interest can be labeled in the first two phases of a three-phase bone scintigraphy. In
with 68Ga as obtained from generators in the chemical state of malignant primary bone tumors, bone scintigraphy is usually
68
Ga(III). Since gallium is a metallic element, radiolabeling not indicated to diagnose the primary but rather to define the
requires the use of a chelator in order to bind the radionuclide extent of osseous spread. Detection of a primary bone tumor
to the molecule of interest. The macrocyclic chelators most on a bone scan performed for bone pain is in most cases inci-
frequently used to this purpose are 1,4,7,10-tetraazacyclodo- dental. In malignant as well as in the benign bone tumors,
decane-1,4,7,10-tetraacetic acid (DOTA) and 1,4,7-triazacy- planar radiography and MRI are usually performed before
clononane-1,4,7-triacetic acid (NOTA) or their derivatives. bone scintigraphy. However, in some cases X-ray and MRI
The combination of novel bisphosphonates with macrocyclic cannot identify the exact nature of an osseous lesion. Bone
chelators provides promising tracers for either diagnosis and/ scintigraphy is not very well suited for distinguishing malig-
or therapy of bone metastatic disease, according to the current nant from benign bone lesions, as many of the benign tumors
concept of theragnostics. In fact, the good target-to-back- also have highly increased uptake of the 99mTc-diphospho-
ground ratio, which all bisphosphonates have in common, is nates. However, as a rule of thumb, the absence of an increase
also advantageous for therapeutic applications due to reduced in bone metabolism usually points to a benign tumor.
radiation dose for nontarget tissue. 68Ga-NO2APBP is charac- [18F]FDG PET/CT has recently emerged as an extremely
terized by a very high uptake in bone metastases (30–60 min valuable diagnostic tool in primary bone tumors. In an exten-
after injection), combined with fast blood clearance and very sive meta-analysis, Bastiaannet et al. demonstrated that [18F]
low uptake in soft tissue. This pattern of biodistribution FDG PET/CT has high sensitivity and specificity in discrimi-
(which is comparable to 18F-fluoride) is superior to that of nating between sarcomas and benign tumors, as well as low-
99m
Tc-MDP [12]. grade from high-grade sarcomas based on the semiquantitative
The use of the DOTA-based radiopharmaceutical measurements of glucose consumption (SUV) [14]. Despite
177
Lu-BPAMD has proved valuable for therapeutic purposes, this favorable experience, the current American College of
since the low-energy β− emission of 177Lu hardly reaches the Radiology (ACR) appropriateness criteria guideline (https://
bone marrow; accordingly, only low or no hematologic tox- acsearch.acr.org/docs/69421/Narrative/) does not recom-
icity was observed in pilot clinical investigations. New 68Ga- mend the use of PET/CT with [18F]FDG the staging or char-
and 177Lu-labeled bisphosphonates possessing improved acterization of primary bone tumors, in part likely because of
pharmacological properties are being explored for possible the overlap in the range of SUVs between benign and malig-
clinical use [13]. nant lesions. Instead, PET/CT with [18F]FDG is currently
Preliminary studies suggest a high therapeutic potential included in the NCCN guidelines for primary and metastatic
for zoledronic acid, the most potent bisphosphonate cur- bone tumors [15], and it constitutes an extremely useful tool
rently employed in patients with bone metastatic disease, for many orthopedic oncologists, who are increasingly refer-
upon conjugation with DOTA (DOTAZOL) for labeling with ring patients for [18F]FDG PET/CT because of its high sensi-
177
Lu. A NODAGA-based zoledronic acid derivative tivity (well over 80%) [16, 17] for the detection of bone
(NODAGAZOL) has the advantage over the DOTA analog that metastases and recurrences.
purification is not required after labeling with 68Ga. Both
68
Ga-NODAGAZOL and 68Ga-DOTAZOL exhibit similarly high 24.3.1.1 Osteosarcoma
bone accumulation, low uptake in soft tissue, and fast renal Osteosarcoma (OS) is the most common primary malignant
clearance. Nevertheless, bone accumulation is greater for bone tumor in children and adolescents and the second most
NODAGAZOL than for DOTAZOL. Considering the potential common primary malignant bone tumor following multiple
of 177Lu-DOTAZOL for endoradiotherapy, 68Ga-NODAGAZOL myeloma in all age groups [18]. OS is more common in
might become a theranostic agent for bone metastases. males than in females and usually presents as a painful mass
rising at the metaphyseal regions, the knee being the most
common site. Histologically, osteosarcomas can be divided
Key Learning Point into a number of subtypes according to the degree of differ-
• The development of 68Ga-labeled PET tracers suit- entiation, location within the bone (intramedullary, surface/
able for investigating bone metabolism is ongoing. juxtacortical, extra-skeletal), and histological variants.
Osteosarcoma may also occur as a secondary lesion in asso-
ciation with underlying benign conditions (secondary osteo- of biopsy by directing the needle to the area of most
sarcoma). The most common subtype of OS is the intense metabolism. Using a tumor-to-background [18F]
conventional type (80%). Typical X-ray features include FDG uptake ratio cutoff level of 3.0 for malignant bone
medullary and cortical bone destruction, wide zone of transi- lesions [21], the sensitivity, specificity, and accuracy of
tion, permeative or moth-eaten appearance, and aggressive [18F]FDG PET in identifying malignant OS were 93%,
periosteal reaction (sunburst type, Codman triangle, 66.7%, and 81.7%, respectively. Furthermore, [18F]FDG
lamellated-onion skin reaction, soft tissue mass with variable PET/CT is more accurate for preoperative staging of sar-
calcified and osteoid tumor matrix). comas than conventional imaging [22, 23]. [18F]FDG
CT is generally used for staging or identification of pre- PET/CT also plays a role in assessing response to treat-
dominantly lytic lesions in which small amounts of mineral- ment and has high prognostic significance [24].
ized material may be undetected on both plain film and Considering that the amount of necrosis induced by neo-
MRI. In particular, chest CT is the most appropriate exam adjuvant chemotherapy is a significant prognostic factor
for the screening of pulmonary metastases from for survival, a post-therapeutic SUVmax <2.5 predicts
osteosarcoma. tumor necrosis in osteosarcoma [25]. After neoadjuvant
A joint-to-joint MRI is the most accurate imaging chemotherapy, SUVmax <2 correlates with good histologic
investigation for local staging, particularly in evaluating
for intraosseous tumor extension, soft tissue involvement,
or skip metastases. Assessment of the growth plate is
essential as up to 75–88% of metaphyseal tumors do cross
the growth plate into the epiphysis. On MRI imaging, the
non-mineralized soft tissue component shows intermedi-
ate T1w and high T2w signal intensity, while the mineral-
ized/ossified components show low T1 and T2 signal
intensity. Scattered regions of hemorrhage with variable
signal or of necrosis are also frequent. Evaluation for
enhancement of the solid tumor component is essential
for guiding biopsy.
For some decades scintigraphy with 99mTc-
diphosphonates has constituted one of the mainstays in the
characterization and management of patients with osteosar-
coma. The tumor lesions are characterized by avid uptake
of the bone-seeking agent, both at the primary site
(Fig. 24.10) and at possible metastatic sites within the skel-
eton (Fig. 24.11) and/or in soft tissues (e.g., pulmonary or
liver metastases). In patients who are candidates for limb
amputation (which is employed with declining frequency
in case of osteosarcoma, because of effective neoadjuvant
therapies reducing the need for such aggressive treatments)
or to other forms of less invasive surgery, bone scintigraphy
can serve as a guide to select the most appropriate level of
bone resection—based on identification of the segment of
bone showing increased tracer uptake, therefore involved
by the disease.
Although [18F]FDG PET/CT in osteosarcoma is not yet
considered appropriate by ACR criteria, multiple studies
have demonstrated that PET/CT can provide useful infor-
mation regarding the primary site, staging, and follow-up
[19, 20]. Besides assessment of locoregional extent, [18F] Fig. 24.10 Planar whole-body scintigraphy with 99mTc-HDP in a
FDG PET/CT has proven to be of value in estimating the patient with newly diagnosed osteosarcoma confined to the left 11th rib,
biological aggressiveness of the tumor, as high-grade sar- better imaged in the posterior view. The scintigraphic pattern per se
does not have distinctive features allowing a differential diagnosis ver-
comas are characterized by intense [18F]FDG avidity. sus other bone diseases (reproduced with permission from: Volterrani
Furthermore, the tumor is often metabolically nonhomo- D, Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare –
geneous, and PET/CT may increase the diagnostic yield Tecniche e Applicazioni. Milan: Springer Italy; 2010)
Fig. 24.11 Planar whole-body (a) and spot acquisitions (b) on the clearly depicted as foci with markedly increased tracer uptake (repro-
chest (orthogonal and oblique views) in a patient who had been previ- duced with permission from: Eary JF. Diagnostic applications of
ously treated for an osteosarcoma of the right femoral head, now the nuclear medicine: sarcomas. In: Strauss HW, Mariani G, Volterrani D,
site of a prosthetic hip implant. There is no sign of active disease at the Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical
treated primary site, whereas multiple metastatic sites at the ribcage are Applications. New York, NY: Springer; 2017:1047–1064)
response, while SUVmax >5 is associated with poor histo- ing study obtained. CT or MRI is used for further character-
logic response [26]. ization. CT can demonstrate the sclerotic lesion and the nidus,
Concerning PET/CT with 18F-fluoride, this investigation while MRI can better define intramedullary and soft tissue
is useful for distinguishing postoperative changes from changes.
tumor recurrence after surgery, based on the fact that, unlike Bone scintigraphy with 99mTc-diphosphonates shows
[18F]FDG, 18F-fluoride does not localize in areas of inflam- intense osteoblastic activity in the tumor region—without
mation but only in newly mineralized bone. however any specific pattern allowing clear definition of the
underlying disease. Although there is no specific role for
24.3.1.2 Osteoid Osteoma and Osteoblastoma [18F]FDG PET/CT in the characterization of OO or OB, [18F]
Osteoid osteoma (OO) and osteoblastoma (OB) usually FDG PET/CT can be used to assess response to local treat-
affect adolescents and young adults [27]. OO is comment [32], e.g., radioablation; efficacy of treatment trans-
monly diagnosed in the cortex of the long bones (50% lates into a significant decrease in [18F]FDG uptake [33].
within the femur or tibia) as an intracortical nidus associ- Tumor recurrence and/or persistence can be detected by
ated with a variable amount of mineralization, accompa- focally increased radiotracer activity.
nied by cortical thickening and reactive sclerosis in a long
bone shaft [28, 29]. 24.3.1.3 Ewing Sarcoma
OB is more common in the posterior element of the spine, Ewing sarcoma (ES) is the second most common malignant
especially in the cervical spine [30]. The differential diagno- tumor in children and young adults. The first imaging modal-
sis between OO and OB has traditionally been based on a size ity in the evaluation of suspected Ewing sarcoma is the plain
criterion (1.5 cm nidus); however, more recently OO and OB film, which usually demonstrates a lesion in the diaphysis of
have been recognized to constitute two different pathologies a long bone, most commonly the femur. Local staging is then
rather than the differential expression of a single tumor [31]. performed with MRI, while CT is used for identifying dis-
In fact, OB has a greater potential for growth, with destruc- tant metastases, particularly in the lung. The most important
tion of bone tissue or even malignant transformation, and prognostic factors in the staging of ES are size of the lesion
recurs more often than OO. X-rays are usually the first imag- and distant metastases.
Fig. 24.12 Spot image of

bone scintigraphy with a b
99m
Tc-MDP in a patient with
Ewing sarcoma of the right
ankle, showing markedly
increased tracer uptake at the
tumor site (a). Plain X-ray (b)
shows fibular involvement
with extension to surrounding
soft tissue (reproduced with
permission from: Eary
JF. Diagnostic applications of
nuclear medicine: sarcomas.
In: Strauss HW, Mariani G,
Volterrani D, Larson SM, Eds.
Nuclear Oncology – From
Pathophysiology to Clinical
Applications. New York, NY:
Springer; 2017:1047–1064)
Similarly as in patients with osteosarcoma, bone scintigra- (with highest incidence in the 20–30-year age range), and
phy with 99mTc-diphosphonates has traditionally been employed most of the cases arise in bones of the hand; it usually mani-
as a valuable aid for clinical management of patients with fests with a melting sore at pressure.
Ewing sarcoma. Also in this case, the scintigraphic pattern A rare form of chondroma is periosteal or juxtacortical
shows nonspecific increased uptake of the bone-seeking agent chondromas. They occur on the surface of the bone, affecting
both at the primary site (Fig. 24.12) and at metastatic sites. equally males and females between the ages of 30 and
[18F]FDG PET/CT has been shown to be superior to MRI 40 years. They can arise in the diaphysis of long bones of the
for skip lesions, especially in the active pediatric bone mar- limbs or in the phalanges, where they manifest with a pain-
row [34], and in the detection of metastatic lymph nodes, but less, slow-growing swelling.
inferior to chest CT in the evaluation of pulmonary metasta- Chondromas can also arise from the synovial sheaths of
ses [35]. However, with the most recent equipment where tendons or in the soft tissues adjacent to the tendons in the
PET is combined with high-performing CT components, this hand and feet of adults; in such cases, they are referred to as
is most likely no more the case. soft tissue or synovial chondromas.
Regarding the value of [18F]FDG PET/CT for prognosis Therapy consists in surgical removal of tumor tissue by
and for monitoring response to treatment in ES patients [36, scraping. Radiotherapy is not useful because this is one of
37], elevated pretreatment SUV correlates with high-grade the least radiation-sensitive tumors. Figure 24.13 shows an
tumors, and SUVmax >5.8 correlates with poor prognosis [38, example of bone scintigraphy in a patient with cervical chon-
39]; furthermore, post-neoadjuvant treatment SUVs <2.5 droma, characterized by intense uptake.
correlate with good histological response (90% necrosis) and The diagnosis of cartilaginous tumors is based on clinical
longer survival [40]. examination and imaging findings. There is a wide spectrum
of cartilaginous tumors, ranging from benign tumors without
24.3.1.4 Cartilaginous Bone Tumors metastatic potential (such as enchondroma and osteochon-
Chondromas are benign tumors composed of mature hyaline droma) to high-grade malignant tumors (such as chondrosar-
cartilage. They generally have limited growth potential and coma with aggressive behavior and early metastases).
are not locally aggressive. These tumors are called enchon- Multimodality radiologic assessment is mandatory to define
dromas when they occur in the medullary canal of the bone. the pattern of the lesion. X-ray is generally the front-line diag-
Enchondroma has the same incidence in males as in females nostic imaging modality used. However, plain X-ray is often
a b c
Fig. 24.13 Bone scintigraphy with 99mTc-HDP obtained in a patient vicular joints and between the upper third and the middle third of the
with chondroma of the cervical spine. Both in the planar whole-body sternum—a non-infrequent occurrence that is normally devoid of any
scan (a) and in the planar spot acquisitions obtained at about 3 h post- clinical relevance (reproduced with permission from: Volterrani D,
injection in the right lateral view (b) and in the left lateral view (c), the Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche
chondroma exhibits markedly enhanced tracer uptake. As an incidental e Applicazioni. Milan: Springer Italy; 2010)
finding, mildly increased tracer uptake can be noticed at the costocla-
not capable to fully characterize the lesion, and further evalu- than benign lesions [41]. An SUVmax >2 has been sug-
ation with cross-sectional imaging is warranted. CT images gested as the threshold for malignant nature [42].
can better assess the cartilaginous matrix (arc-and-ring or
stippled morphology) and the endosteal scalloping, which
might be useful to distinguish enchondroma from a low-grade 24.3.2 Fibrous Bone Dysplasia
chondrosarcoma. Both MRI and CT are accurate in distin-
guishing osteochondroma from chondrosarcoma by measur- Fibrous bone dysplasia is a non-hereditary congenital
ing the thickness of the cap, which is usually >2 cm in benign skeletal disorder in which the bone is replaced with
malignant lesions. a fibrous-like tissue with early osteogenesis; its prevalence
Bone scintigraphy can contribute to the differentiation is not easily defined, as the illness is often asymptomatic.
of benign from malignant lesions, as malignant lesions are Bone lesions are generally monostotic but in about 10–15%
characterized by markedly increased 99mTc-diphosphonate of cases are polyostotic, especially when associated with
uptake. Similarly, high [18F]FDG uptake might help in the genetic syndromes such as the McCune-Albright’s or
differential diagnosis between benign and malignant Mazabraud’s syndromes. The polyostotic form is often uni-
lesions: benign lesions have low SUVs, benign lesions lateral or monomelic. Fibrous bone dysplasia presents in
with atypical histological features have slightly higher children or young adults as a painless osseous abnormality
values, and malignant lesions have markedly higher SUVs affecting the ribs (28%), femurs (23%), or neurocranium
(20%). However, in the polyostotic form, it is more likely into sarcomas, although they rarely metastasize; there-
that lesions are associated with pain and frail bone, with fore, they are often considered as quasi-malignant
possible fractures. In some patients (or in some bone sites), lesions.
lesions are hypertrophic and can cause neurological compli- The typical appearance of GCT on X-ray is an expansile
cations. The diagnosis is based on radiological and scinti- lesion without bone matrix at the metaphysis/epiphysis of a
graphic imaging, in which hyperactivity of the bone-seeking long bone; this finding suggests further assessment with CT
radiopharmaceutical is observed, both in bone lesions and in and MRI. CT better evaluates the absence of mineralization,
the complicating fractures. The amount of woven bone and the narrow transition zone, and the cortical thinning with
fibrous tissue determines the X-ray and CT appearance, possible periosteal reaction, while MRI shows a typical
which can be sclerotic, cystic, or mixed cystic-sclerotic [43, whorled appearance on T2-weighted sequences and better
44]. Given its variable appearance, MRI is usually less use- detects the presence soft tissue extent.
ful for the diagnosis. GCTs are extremely [18F]FDG-avid, with SUVs overlap-
Bone scintigraphy demonstrates areas of reduced uptake ping those of malignant bone tumors [50]. Uptake is often
alternating with areas of increased uptake, reflecting the more prominent at the periphery with a central photopenic
complex structure of the lesion. In general, the fibrous or region (doughnut sign). The high vascularity of the tumor
cystic elements concentrate little or no radiopharmaceutical, leads to a generalized regional hyperemia, which can trans-
while bones or calcified areas exhibit increased tracer uptake. late into diffuse increase in radiotracer uptake [51]. In the
The tracer accumulates markedly in pathological fractures rare case of multicentric giant cell tumors, the role of [18F]
and moderately on the periphery, indicating the expansive FDG PET/CT is to detect occult lesions [52]. Although
nature of the disease with stimulation of the sclerotic edge. rarely, GCT can metastasize or transform into malignant sar-
Tracer uptake begins to intensify when injuries are replaced comas; in this case, PET/CT is useful for detecting metasta-
by bone tissue. sis. In the primary malignancy, PET/CT can detect foci of
[18F]FDG PET/CT is useful to assess the extent of poly- more prominent uptake, which might be suitable for biopsy.
ostotic disease; the bone abnormalities seen as ground-
glass, mixed sclerotic, and lytic expansive lesions 24.3.2.2 C ortical Fibrous Defects and
correspond to increased uptake on the PET scan [45]. [18F] Non-ossifying Fibroma
FDG PET/CT has been used to assess rare syndromes This is an asymptomatic condition characteristic of young
associated with fibrous bone dysplasia [46]. In those cases, boys, generally detected incidentally during an imaging test.
the scan is also a valuable tool to detect other abnormali- It is typically represented by a single cortical lesion of long
ties associated with the specific syndrome, such as intra- bones, which in most cases regresses spontaneously.
muscular myxomas in Mazabraud’s syndrome. Sometime the lesion persists and evolves in non-ossifying
Furthermore, [18F]FDG PET/CT can detect malignant fibrous bone marrow. Bone scintigraphy might be normal;
degeneration [47] which can occur in 1–4% of cases, such when positive, it shows an area of slightly increased radio-
as transformation into osteosarcoma, fibrosarcoma, and pharmaceutical uptake characterized by a ringlike pattern
malignant fibrous histiocytoma, especially in the polyostotic surrounding a photopenic lesion. In case of calcification or
form [48]. In case of progressively increasing [18F]FDG ossification, uniform and intense uptake is present.
uptake, malignant degeneration should be suspected, and a Acquisition with a pinhole collimator and SPECT (or prefer-
biopsy should be performed under PET/CT guidance. ably SPECT/CT) might be useful.
Fibrous dysplasia and other fibrous lesions such as non-
ossifying fibroma and fibrous cortical defect (see below) can
be misdiagnosed as metastatic cancer in the staging of other Key Learning Points
primary tumors, as all of these diseases have enhanced [18F] • Malignant primary bone tumors usually exhibit
FDG uptake. Fibrous dysplasia is a potential pitfall also with increased uptake in every phase of a three-phase
18
F-fluoride PET/CT, due to the high tracer uptake. In this bone scintigraphy.
case, CT imaging should be carefully reviewed to reduce the • This is also the case of most benign primary bone
false positive rate and thus avoid unnecessary biopsies. tumors.
• In malignant primary bone tumors, bone scintigra-
24.3.2.1 Giant Cell Tumor phy is usually not indicated to diagnose or charac-
The giant cell tumor (GCT) accounts for about 5% of all terize the primary lesion but rather to define the
primary osseous tumors, 80% of cases occurring between extent of osseous diffusion.
the second and the fifth decades [49]. GCT arises follow- • This is also true of [18F]FDG PET/CT, although in
ing fusion of the growth plate and extends from the some conditions such as sarcomas, uptake of [18F]
bone’s metaphysis to the epiphysis; the most common FDG has been shown to correlate with tumor
site is the knee, involved in more than half of the cases. aggressiveness.
GCTs are generally benign tumors but may transform
24.3.2.3 Myositis Ossificans 24.3.3 Metastatic Bone Tumors

This rare benign condition most commonly presents as a
localized, self-limiting lesion secondary to contusion trauma. Bone metastases occur frequently in patients with breast,
Scintigraphy can be useful, especially to confirm the unique- prostate, lung, and renal carcinomas; the frequency of osse-
ness of the lesion and for the differential diagnosis with both ous involvement in postmortem examinations of patients
nonmalignant (nodular fasciitis, juxtacortical osteoma, with these tumors ranges between 30% and 70%. Bone
osteochondroma, chondroma) and malignant lesions (osteo- metastases reduce survival significantly, between 6 and
genic sarcoma). Figure 24.14 shows a typical example of 48 months depending on tumor type.
bone scintigraphy in a patient with myositis ossificans. As for other diagnostic tests, some knowledge on the
pretest probabilities of a given condition is also relevant
when bone scintigraphy is used for staging malignant dis-
ease [53, 54]. The pretest probabilities of bone metastases
Key Learning Point depend on tumor stage and other risk factors. In breast
• Myositis ossificans is characterized by high uptake cancer stages I and II, it is between 0.8% and 2.6%, rising
of 99mTc-diphosphonates. to 16.8–40.5% in stages III and IV. Risk factors for skele-
tal metastasis in prostate cancer patients are a serum PSA
a b c
d e
Fig. 24.14 Planar whole-body scintigraphy obtained about 3 h after posterior views (b and c) and especially in the lateral views of the right
administration of 99mTc-HDP in a young adult patient who had suffered thigh (d and e) better demonstrate that increased uptake is not associ-
contusion trauma of the right thigh muscles several months earlier. ated with bone but rather involves the soft tissues adjacent to bone
Besides the still metabolically active growth plates at the knees and (reproduced with permission from: Volterrani D, Erba PA, Mariani G,
ankles, the whole-body images (a) show an area of focally increased Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
tracer uptake that appears to extend beyond the bone at the upper third Milan: Springer Italy; 2010)
of the right femur. The planar spot images acquired in the anterior and
level >10 ng/mL and a Gleason score ≥8; above these two prostate carcinoma), also lesions smaller than 1 cm may be
thresholds, the pretest probability of osseous involvement detected by bone scintigraphy; whereas, metastatic lesions
in prostate cancer rises to 16.2%. Tumor patients experi- with low uptake (as observed in, e.g., the predominantly
encing bone pain also have a higher risk of harboring bone lytic metastases of breast cancer) may not be detected, even
metastases than those who are asymptomatic. if larger than 1 cm.
Purely lytic metastases may escape detection by bone Osseous metastatic spread begins with tumor cells infil-
scintigraphy. Therefore, patients with renal cancer, myeloma, trating the bone marrow. Bone scintigraphy can detect metas-
or lymphomas are usually not examined by bone scintigra- tases only when the neoplastic cells invading the bone marrow
phy. This limitation does not apply to breast or prostate can- activate involvement of the mineralized component of bone,
cer because these tumors cause predominant osteoblastic thus initiating a reactive increase in bone metabolism. Since
bone lesions. However, the broader availability of positron- MRI and radionuclide imaging with tumor-seeking radio-
emitting radiopharmaceuticals such as 18F-fluoride, choline pharmaceuticals (as typically occurring with PET) enable
(labeled with either 11C or 18F), the synthetic amino acid direct visualization of the neoplastic foci, they usually become
anti-1-amino-3-18F-fluorocyclobutane-1-carboxylic acid positive for metastatic disease earlier than bone scintigraphy
(18F-FACBC, also known as 18F-fluciclovine), and 68Ga and are thus more sensitive than the conventional bone scan.
ligands that target the prostate specific membrane antigen It is also known that, in face of a relatively high sensitivity,
(PSMA) are reducing the need for bone scintigraphy in pros- the specificity of bone scintigraphy for bone metastases is quite
tate cancer patients, due to their higher diagnostic accuracy low, due in particular to the fact that numerous benign condi-
compared to conventional bone scintigraphy and to the pos- tions also lead to focal increases of bone metabolism. Among
sibility of detecting both skeletal and extra-skeletal meta- these, degenerative changes of the skeleton (including osteo-
static disease with a single imaging procedure. chondrosis of the spine due to degeneration of the interverte-
A similar scenario is developing also for patients with bral disk and facet’s osteoarthritis) raise in elderly patients the
bronchial carcinoma, in whom conventional bone scintigra- most frequent differential diagnosis. Nonetheless, these condi-
phy is being replaced with [18F]FDG PET/CT, which is useful tions have a typical appearance on CT; therefore, the specificity
not only for staging but also for assessing response to of bone scintigraphy in staging malignant disease is markedly
treatment. improved by hybrid SPECT/CT imaging (Figs. 24.17–24.19).
Since bone metastases develop by hematogenous spread, Further benign differential diagnoses of a hot spot such as
they are usually found in the parts of the skeleton that harbor vertebral fractures (Fig. 24.20) or benign bone tumors are
the well-perfused red bone marrow. Therefore, hot spots on also amenable to CT characterization. Since the advent of
the bone scan caused by metastases are usually not joint- SPECT/CT, a considerable amount of evidence has been pub-
related. Multifocal involvement is usually more frequent lished analyzing its utility for staging compared to planar or
than solitary bone metastases (Fig. 24.15). In disseminated stand-alone SPECT imaging. It has thus been established that
metastatic disease, diffusely increased bone metabolism may more than 90% of the foci of abnormal uptake considered
be visualized by bone scintigraphy, raising the need for dif- indeterminate on stand-alone radionuclide imaging can be
ferential diagnosis between the so-called superscan due to elucidated by SPECT/CT [55] (Figs. 24.21 and 24.22).
disseminated metastatic disease (Fig. 24.16) and metabolic Hybrid imaging has, thus, turned bone scintigraphy from a
bone disease (due to, e.g., severe hyperparathyroidism). The sensitive but nonspecific imaging procedure to a highly accu-
markedly increased avidity of the skeleton for 99mTc-diphos- rate diagnostic tool for staging the skeleton in malignant dis-
phonates or for 18F-fluoride occurring in patients with the ease [55].
superscan reduces radioactivity accumulation in soft tissues On the other hand, bone scintigraphy with 99mTc-
as well as excretion through the nephro-urinary tract. diphosphonates can also be used to monitor the efficacy of
Published evidence on the diagnostic accuracy of bone therapy in patients with skeletal metastatic disease.
scintigraphy for staging cancer patients is of low quality, in Nevertheless, caution should be taken in the interpretation
particular because of the lack of a reliable gold standard of the images so obtained (especially if no hybrid SPECT/
[53, 54]. Published sensitivities range between 85% and CT imaging is employed), since abnormally increased
96%, higher in prostate than in breast cancer. Sensitivity is tracer uptake may persist at the sites of metastasis up to
limited by suboptimal spatial resolution of the planar and/ several months after start of favourable response to treat-
or SPECT gamma cameras—currently between 8 and ment. This limitation is less stringent when using PET
10 mm. Very small bone lesions may, therefore, escape with 18F-fluoride, where changes in the pattern of tracer
detection by bone scintigraphy. However, detectability distribution reflect more readily the pathophysiologic
depends not only on size but also on the uptake ratio changes occurring during therapy – either in the case of
between lesion and background. In case of very high lesion progression despite treatment or in the case of favourable
uptake (as observed in, e.g., the osteoblastic metastases of response to treatment (see Fig. 24.23)

planar whole-body scans a b
obtained with 99mTc-
diphosphonates in different
patients with cancer. The
cases are presented here with
increasing severity of
metastatic bone disease,
including single lesion in a
vertebral body in the lower
thoracic spine (a); multiple
lesions in the spine, ribs, and
skull (b); and disseminated
lesions throughout the
skeleton (c and d)
c d
Fig. 24.16 Examples of “superscans” in patients with advanced met- nary tract (superscan 99mTc-MDP image modified from: Volterrani D,
astatic prostate cancer as visualized in two different patients by con- Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare -
ventional 99mTc-MDP bone scintigraphy (left) and by PET with Tecniche e Applicazioni. Milan: Springer Italy; 2010. Superscan
18
F-fluoride (right). Due to markedly increased avidity of bones for 18
F-fluoride image kindly provided by Guofan Xu, MD, PhD,
the bone seeking agents, in both cases there is very little radioactivity Department of Nuclear Medicine, M.D. Anderson Cancer Center,
remaining in the soft tissues and very little excretion through the uri- Houston, Texas, USA)
Fig. 24.17 Bone scintigraphy in a 55-year-old patient with breast car- Images of SPECT/CT fusion (center panel) and low-dose CT (right
cinoma. On the dorsal planar view (left panel), a discrete focus of panel) show that this corresponds to an area of osteolysis, possibly
uptake projects to the lateral aspects of the fifth lumbar vertebral body. caused by metastasis
Fig. 24.18 Bone scintigraphy in a 72-year-old patient with breast car- dose CT (right panel) disclose them as degenerative: osteochondroses
cinoma. Dorsal planar view (left panel) features two hot spots in the between L5 and S1 as well as between L4 and L5 showing disk space
lower spine. The images of SPECT/CT fusion (center panel) and low- narrowing, vacuum phenomena (“gas in disk”), and osteophytes
Fig. 24.19 Bone scintigraphy in a 71-year-old patient referred for cervical spine. Hybrid fused SPECT/CT image (center panel) and CT
staging of breast carcinoma. Dorsal planar view (left panel) shows image (right panel) disclose left-sided facet’s osteoarthritis and right-
focus of moderately increased uptake in the left-lateral aspects of the sided uncovertebral arthropathy, the latter featuring a geode
tion may be overcome by SPECT/CT hybrid imaging.

Key Learning Points Whereas the focal increase of uptake in a joint of the
• Bone scintigraphy is useful for staging breast and extremities is quite often accompanied by pain, this is not
prostate cancer, in particular in those patients with a necessarily true for the spine. Therefore, the specificity
high risk for osseous metastases. of bone scintigraphy to detect the correlate of skeletal
• Purely lytic metastases (such as those occurring in, pain is in the case of spinal degenerative disease also
e.g., renal cancer or myeloma) may escape detec- limited.
tion by bone scintigraphy. When reading bone scans of patients affected by osseous
• Bone scintigraphy can detect metastases only when pain, identifying major differential diagnoses is helpful to the
the neoplastic cells invading the bone marrow cause referring physician. These categories are malignant and benign
a reactive increase in bone metabolism. bone tumors, inflammation, trauma, degenerative disease, vas-
• The sensitivity of bone scintigraphy to detect osse- cular disorders, and some miscellaneous conditions eluding
ous metastases may thus be lower than that of MRI, this categorization. The pretest probabilities of these categories
which visualizes the integrity of bone marrow, or vary largely with clinical presentation. Therefore, before inject-
lower than that of imaging with tumor-seeking ing the radiopharmaceutical, a detailed medical history with
agents (e.g., PET/CT), which visualize the tumor regard to the occurrence and nature of the pain as well as to that
cells directly. of possible trauma is highly relevant. Furthermore, concomi-
• The specificity of bone scintigraphy to detect osse- tant laboratory findings such as the serum level of the C-reactive
ous metastases is greatly increased when SPECT/ protein (CRP) or leukocyte counts should be available as well
CT is used. as the results and—ideally—also the image datasets from pre-
vious radiological examinations.
24.4.1.1 Degeneration
24.4 Benign Bone Disease Degenerative conditions are one of the most frequent
causes of bone pain. Joint degeneration is caused by
24.4.1 Workup of Bone Pain degeneration of the joint cartilage, which is ultimately
destroyed in this process. In the case of spinal degenera-
Pain affecting the skeleton is very frequent. Its differen- tion, a progressive damage to the intervertebral disk
tial diagnosis is vast and can be roughly subdivided into occurs, which loses its elasticity and height. On CT and
extraosseous and osseous conditions. Bone scintigraphy radiographic imaging, cartilaginous degeneration presents
can be used to detect osseous pain generators with rather as thinning of the joint spaces; furthermore, calcifications
high sensitivity, although specificity is reduced by its as well as gas deposits can be seen. The latter is referred to
inability to morphologically characterize foci of increased as the so-called vacuum phenomenon. As a consequence
tracer uptake. Many studies have shown that this limita- of progressive destruction of these cartilaginous struc-
Fig. 24.20 Bone
scintigraphy obtained in a
patient with breast carcinoma.
Planar views (upper panel)
depict increased tracer uptake
in at least four vertebral
bodies. This is due to recent
vertebral compression
fractures, as demonstrated by
SPECT/CT fusion images
(middle panel) and low-dose
CT images (bottom panel)
tures, the mechanical stress on the joint-forming bones bility. Degeneration may lead to inflammatory processes,
increases, leading to subchondral bone edema and—subse- causing pain. This so-called active osteoarthritis is accom-
quently—to subchondral sclerosis. Further sequelae are panied by an increase in bone metabolism as well as an
the so-called osteophytes or—in the case of the spine— increase of uptake in the earlier phases of three-phase bone
spondylophytes, which are an indirect sign of joint insta- scintigraphy.
Fig. 24.21 Upper and lower rows represent transaxial SPECT (left sis, and an osteophyte. In the lower row, signs of osteoarthropathy are
panel), CT (center panel), and SPECT/CT fusion images (right panel) absent, and the hot spot is found on one side of the joint only.
in two different patients with breast carcinoma referred for staging by Furthermore, bone is hypodense compared to contralateral, indicating a
bone scintigraphy. On the SPECT scans, increased tracer uptake pro- small osteolytic metastasis (reproduced with permission from: Mariani
jecting to the facet’s joint of a vertebral body is seen in both patients. In G, Bruselli L, Kuwert T, Kim EE, Flotats A, Israel O, et al. A review on
the upper row, findings are due to facet’s arthropathy with increased the clinical uses of SPECT/CT. Eur J Nucl Med Mol Imaging.
tracer uptake projecting to both sides of the joint with associated signs 2010;37:1959–1985)
of osteoarthropathy such as joint space narrowing, subchondral sclero-
Fig. 24.22 Bone scintigraphy in a patient with bronchial carcinoma. corresponding CT image (rightmost two panels) demonstrate findings
Dorsal planar view (left panel) shows a cold spot in the 8th thoracic typical of a hemangioma, with so-called salt-and-pepper appearance of
vertebra. Hybrid fused SPECT/CT image (second panel from left) and the vertebra on the transaxial CT image—also containing fat
SPECT/CT helps in localizing foci of increased uptake floridity of the lesion, which in the case of the peripheral
and at the same time increases the diagnostic specificity joints correlates well with increases in bone metabolism. In
considerably, as the hallmarks of osteoarthritis are readily the case of the spine, the relationship between an increase
visualized by the CT component of the scan. The degree of in bone metabolism and pain generation is less direct.
morphological alterations correlates only weakly with the However, spinal regions with normal bone metabolism can
Fig. 24.23 Different patterns of response to androgen-deprivation of osseous metastatic disease, now involving multiple thoracic vertebral
therapy in two different patients with skeletal metastases from prostate bodies. (b) The baseline, pre-therapy PET scan obtained from this
cancer, as visualized by PET/CT with 18F-fluoride. (a) The baseline, 72-year-old man demonstrates diffuse skeletal metastatic disease
pre-therapy PET/CT scan obtained from this 65-year-old man (top row) involving the skull, spine, upper limbs, ribs, pelvis, and lower limbs
reveals metastatic lesions within the vertebral bodies of C3, T2, L1, and (left panel). Considerable improvement is observed in the scan obtained
L4 (fusion PET/CT image in left panel, CT image in center panel, PET 3 months after initiating androgen-deprivation therapy (center panel),
image in right panel). Six months after initiation of androgen-depriva- with further improvement at 6 months (right panel)
tion therapy (bottom row), the PET/CT scan demonstrates progression
b
be ruled out as generators of bone pain. Considerable evi- 24.5 Metabolic Bone Diseases
dence has been published showing that SPECT/CT can
guide minimal-invasive treatment of joint disease. The term metabolic bone disease includes a group of altera-
tions that involve the skeleton in full, characterized by an
increase in bone turnover. The corresponding scintigraphic
Key Learning Points pattern is characterized by a diffuse increased uptake of the
• When reading bone scans of patients with skeletal radiopharmaceutical. There are, in addition, features of
pain, listing of the possible differential diagnoses focally increased uptake, such as the one involving long
into major categories might be helpful. bones typical of hypertrophic pulmonary osteopathy
• These categories are malignant and benign bone (Fig. 24.24).
tumors, inflammation, trauma, degenerative dis-
ease, vascular disorders, and some miscellaneous
conditions. 24.5.1 Osteoporosis
• Joint degeneration is one of the most frequent
causes of bone pain and is caused by degeneration Osteoporosis is a metabolic skeletal disorder characterized by
of the joint cartilage. decreased bone density and quality, reducing bone strength
• Joint degeneration is accompanied by an increase in and predisposing the patient to fractures even for minor trau-
bone metabolism, when active. mas. There are two types of osteoporosis: postmenopausal
• SPECT/CT helps in localizing foci of increased osteoporosis and senile osteoporosis. Given the availability of
uptake and at the same time increases the diagnostic effective treatments, early diagnosis of osteoporosis is crucial
specificity considerably, as the hallmarks of osteo- to slow down its evolution and thus reduce the risk of compli-
arthritis are readily visualized by the CT component cations. Dual X-ray absorptiometry (DEXA) has completely
of the scan. replaced earlier techniques based on a sealed radionuclide
dystrophy. Nevertheless, scintigraphy does not allow one to

diagnose reduced bone mass.
In patients with multiple vertebral fractures, bone scintig-
raphy may also disclose reduced height of the spine and chest
cage closer to the pelvis. Detection of unknown vertebral or
rib fractures is also possible, raising the suspicion of osteopo-
rosis. Vertebral crush is visible at scintigraphy as an intense
linear uptake of the tracer, corresponding to the site of frac-
ture. Abnormal uptake at the site of fractures can be detected
in 80% of cases within 24 h and in 95% of cases within 72 h.
This high uptake decreases progressively over a period of
time ranging from 6 to 18 months but extending even up to
3 years depending on fracture site and patient age, with a time
trend that can help determine the age of the fracture.
In case of back pain, bone scintigraphy can help deter-
mine whether pain is attributable to vertebral fracture
(Fig. 24.26); a normal scan excludes a recent fracture and
directs the clinician toward other diagnostic options.
Bone scintigraphy can also reveal other causes of back
pain, such as tumor metastasis, Paget’s disease, and infection
(Fig. 24.27). SPECT and SPECT/TC have increased the diag-
nostic value of bone scintigraphy, especially for the lumbar
region, based on better scintigraphic contrast and the possibil-
ity to localize the area of hyperactivity more accurately.
SPECT images show that joint lesions are more common
in patients with multiple vertebral fractures. Since most often
the area of increased uptake occurs at the level of L4-L5 and
L5-S1, it can be assumed that hyperactivity of the articulated
joints in these areas is due to the simultaneous presence of
osteoarthritis and to the stress due to the collapse of a verte-
Fig. 24.24 Planar whole-body scan obtained about 3 h after adminis-
tration of Tc-HDP in a patient with hypertrophic pulmonary osteopa-
99m bral body above or below those levels. Patients with vertebral
thy. Increased tracer uptake with a typical “railway” pattern is clearly crush can present other fractures that bone scintigraphy can
recognized at the femurs and tibias (reproduced with permission from: detect at an early stage, when traditional radiological imag-
Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina ing is still negative or nonconclusive. Unknown fractures of
Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
the pelvis detected only by bone scintigraphy can mimic pain
caused by vertebral crush. Similarly, unknown fractures of
source emitting γ-rays either with a single energy peak (e.g., the femoral neck, scapula, radium and chest cage can be
125
I and/or 241Am) or with two different energy peaks (such as detected by bone scintigraphy.
153
Gd). Therefore, nuclear medicine does not have a role in The use of [18F]FDG PET/CT can help distinguish a
screening patients for osteoporosis. pathological vertebral fracture caused by tumor metastasis
Regarding instead the most common radionuclide bone from an osteoporotic fracture. The latter tends to have no
imaging modality, i.e., bone scintigraphy with 99mTc- pathologically increased [18F]FDG uptake, while high [18F]
diphosphonates, there are several nonspecific findings that FDG uptake is characteristic for malignant and inflamma-
might suggest the diagnosis of osteoporosis, such as reduced tory processes [56].
bone-to-soft-tissue uptake ratio, decreased resolution of ver- Nuclear medicine imaging might be of value also in the
tebral body endplates, and relatively increased skull uptake assessment of side effects from osteoporotic therapies. For
of the radiopharmaceutical (Fig. 24.25). However, the value example, bone scan can be used to assess possible fractures
of nuclear medicine imaging in patients with osteoporosis is related to bisphosphonate therapy [57]. Long-standing
predominantly in the assessment of complications, such as bisphosphonate therapy changes the bone architecture,
pathological fractures. This is particularly important with weakening the bone and resulting in an increased risk of low-
regard to vertebral fractures, since bone scintigraphy can energy atypical subtrochanteric and proximal diaphyseal
exclude the coexistence of other skeletal diseases that may femoral fractures. These fractures are often subtle and missed
cause fractures. Bone scintigraphy is also of great help for on initial radiography [58]; however, because of the risk of a
the diagnosis of regional osteoporosis syndromes and algo- delayed diagnosis of displaced complete fracture, early
a b
Fig. 24.25 Planar bone scintigraphy obtained about 3 h after adminis- ing to prior fractures cause by minor traumas—as typically observed in
tration of 99mTc-HDP in a patient with osteoporosis previously submit- patients with osteoporosis. Increased tracer uptake is observed also at
ted to bilateral hip joint prosthetic implants. The whole-body anterior both hips, reflecting ongoing bone post-implantation remodeling
and posterior views (a) show diffuse inhomogeneities in tracer distribu- (reproduced with permission from: Volterrani D, Erba PA, Mariani G,
tion throughout the skeleton, with multiple areas of focally increased Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
uptake in the ribcage, better shown in the lateral views (b), correspond- Milan: Springer Italy; 2010)
Fig. 24.26 Bone scintigraphy performed in an elderly patient with Honda sign; in the right os ischiaticum, a further hot spot can be
long-standing pelvic pain. Dorsal planar view (left panel) shows zones detected. On hybrid fused SPECT/CT imaging (center panel) and CT
with markedly increased tracer uptake in/near both iliosacral joints, imaging (right panel), increased uptake corresponds to a fracture cleft
connected by a thin line of increased tracer accumulation, the so-called in the os sacrum. The findings are due to a sacral insufficiency fracture
assessment for such fractures using a highly sensitive tech- the efficacy of therapeutic interventions for osteoporosis.
nique, such as scintigraphy, is recommended. While there does not yet seem to be a consensus on whether
The kinetics 18F-NaF uptake in bone and the plasma clear- to use static or dynamic PET, it is suggested that a single
ance of this tracer in patients with osteoporosis have been static PET acquisition may be able to closely approximate
investigated. As reviewed by Raynor et al., measurements of the 1-h dynamic PET values of regional blood clearance,
plasma clearance to assess response to new treatments for while completing a scan in only 5 min [8, 60, 61].
osteoporosis [59] have validated the use of 18F-fluoride PET A recent study compared the measurements obtained by
imaging as a noninvasive method of imaging and evaluating static PET to those obtained by dynamic PET, reporting a cor-
Fig. 24.27 Bone scintigraphy performed in a 67-year-old woman with panel) as well as low-dose CT (right panel) disclose this to be due to
considerable, long-standing back pain. Dorsal view of planar scan (left spondylodiscitis with contour defects of endplates and normal width of
panel) shows markedly enhanced tracer uptake in the lower lumbar disk space. Other signs of osteochondrosis are also missing
spine extending over at least two segments. Hybrid images (center
relation coefficient between estimated Ki values and calcu- scenario of skeletal deformities in rickets includes bowed legs,
lated Ki values 30–60 min after injection even >0.99. The skull bossing, pectus carinatum, hyphoscoliosis, and costo-
potential of static scans to replace dynamic scans would elimi- chondral swelling. Osteomalacia may present with a mild and
nate the need for arterial blood sampling, as well as greatly subtle clinical picture without the visible bone deformities
reduce the duration of the imaging session. Bone turnover in seen in the pediatric population. Bone scans in patients with
the femoral neck—the site of most traumatic osteoporotic osteomalacia can be normal, especially in early stages. The
fractures—has also been investigated as marker for osteoporo- typical scintigraphic pattern in later stages of this condition is
sis. The femoral neck region was defined using anatomical characterized by intense uptake of 99mTc-diphosphonates at the
landmarks from CT scans, and quantification of 18F-fluoride metaphysis and long bone ossification centers (so-called
uptake in the entire neck was found to reflect global bone turn- chicken bone). Figure 24.28 shows two typical examples.
over. The metabolically active volume, maximum SUV, mean Osteomalacia shares with hyperparathyroidism some
SUV, and total calcium metabolism were calculated. A corre- metabolic features that on bone scintigraphy appear as
lation coefficient of −0.43 was found between age and global increased bone-to-soft tissue uptake ratio, increased uptake
values, indicating a decline in global activity with aging [59]. in long bones and skull, beading of the costochondral junc-
These preliminary data demonstrate the potential role of tions, and the “tie sternum” appearance [62, 63]. “Metabolic”
global bone turnover in the femoral neck for early detection of fractures have also been described associated with osteoma-
osteoporosis but also the need for larger-scale studies to evalu- lacia and often misinterpreted as metastasis.
ate the clinical utility of this approach. Oncogenic osteomalacia is a rare form of osteomalacia
but deserves a special note for its clinical significance. This
condition is also known as tumor-induced osteomalacia and
24.5.2 Osteomalacia is usually associated with mesenchymal tumors. This para-
neoplastic syndrome is caused by the secretion of a sub-
Osteomalacia and rickets are disorders characterized by stance that interferes with the normal metabolism of
abnormal mineralization of the skeleton resulting from defects calcium and phosphate. If there is suspicion of tumor-
in the phosphate, calcium, and/or vitamin D metabolism. induced osteomalacia, a whole-body scan such as [18F]FDG
Rickets is a childhood disease (most frequent in developing PET/CT is recommended to locate the primary cancer. [18F]
countries), whereas osteomalacia occurs in adults. The clinical FDG [64, 65] and the somatostatin receptor imaging agents
a b
Fig. 24.28 Planar whole-body scans (anterior and posterior views) scintigraphic pattern for the patient shown in panel (b) is instead char-
obtained in two different patients with osteomalacia about 3 h after acterized by multiple foci of markedly increased tracer uptake involv-
administration of 99mTc-HDP. Both patients exhibit diffusely increased ing also pelvic bones and the lower limbs (reproduced with permission
tracer uptake at multiple sites especially in the ribs for patient shown from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
in panel (a), in whom the “chicken bone” pattern is visible particu- Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
larly in the posterior view of the femora and tibiae. The predominant
such as 68Ga-DOTA-TATE [66, 67], 68Ga-DOTA-NOC [68], roid stimulation. 99mTc-sestamibi scans play a key role in the
and 68Ga-DOTA-TOC have each been reported to detect the preoperative localization of parathyroid adenoma(s) (see
primary tumor in these cases. 18F-DOPA might be also an Chap. 28 of this book).
option. The typical bone scan pattern of hyperparathyroidism is
represented by generalized increased radiopharmaceutical
uptake in bones of the skull, jaw, sternum, and shoulder
24.5.3 Primary and Secondary (Fig. 24.29). With high-resolution scintigraphy (e.g., using
Hyperparathyroidism the pinhole collimator for acquiring spot views at selected
sites), the characteristic sign of the linearly increased uptake
Hyperparathyroidism is characterized by an excess of para- in the subperiosteal bone, for example, at the level of the
thyroid hormone in the blood that results in weakening of the phalanxes, might be detected. Other signs associated with
bones through loss of calcium. Depending on the etiology, hyperparathyroidism are the “rosary beads sign” (increased
hyperparathyroidism can be primary, secondary, or tertiary. uptake at the costochondral junctions) and the “tie sign”
Primary hyperparathyroidism is due to parathyroid adenoma (prominent uptake in the sternum).
in 80% of cases, parathyroid gland hyperplasia in 15–20%, Hyperparathyroidism has also been described associated
and rarely cancer (<0.5%). Secondary hyperparathyroidism is with brown tumors, soft tissue calcifications, fractures, and
usually related to vitamin D deficiency, chronic kidney dis- subchondral bone resorption. Brown tumors are also known
ease, or other causes of low blood calcium levels resulting in as osteitis fibrosa cystica or osteoclastomas and result from
hyperstimulation of the parathyroid glands. Tertiary hyper- replacement of the normal bone marrow with hemorrhage
parathyroidism is characterized by autonomous parathyroid and granulation tissue. These tumors can be visualized with
function following resolution of the initial cause of parathy- whole-body 99mTc-sestamibi scintigraphy, although [18F]FDG
Fig. 24.29 Composite planar

bone scintigraphy obtained
about 3 h after administration
of 99mTc-MDP in a patient with
hyperparathyroidism
secondary to renal failure.
Markedly increased tracer
uptake virtually in the whole
head, including the calvarium,
maxillary bones, and
mandible. This scintigraphic
pattern of the head is rather
typical and resembles the
appearance of a grenadier
soldier (reproduced with
permission from: Volterrani D,
Italy; 2010)
PET/CT has been reported as a more sensitive technique [69]. 24.5.5 Hypertrophic Osteoarthropathy
Soft tissue calcifications, also known as metastatic calcifica-
tions, have been described in the soft tissue of both upper and Hypertrophic osteoarthropathy can be idiopathic or second-
lower extremities [70], lungs [71], thyroid, and stomach [72]. ary to malignancy. This condition is characterized by perios-
Improvement of these features has been reported following teal reaction, more prominent in the long bones. In the bone
resection of the hyperfunctioning parathyroid tissue [72]. scan, linearly increased uptake in parallel lines along the cor-
tex of the diaphysis of the radii, tibias, and femurs can be
detected [73]. Secondary forms, for example, due to lung can-
24.5.4 Renal Osteodystrophy cer, often resolve following treatment of the underlying cause.
The most typical scintigraphic pattern of renal osteodystro-

phy (a condition characterized by complex metabolic Key Learning Points
derangements including, among others, secondary hyper- • The term metabolic bone disease includes a group
parathyroidism) is represented by homogeneously increased of alterations that involve the skeleton in full, char-
uptake of 99mTc-diphosphonates in the whole skeleton but acterized by an increase in bone turnover.
particularly in the bones of the skull and jaw (see Fig. 24.29). • The corresponding scintigraphic pattern is charac-
The images acquired 24 post-administration are character- terized by a diffusely increased uptake of the
ized by very high bone-to-soft-tissue tissue ratios; pulmo- radiopharmaceutical.
nary, gastric, or renal calcifications may also be visualized.
24.6 Trauma and Fractures tures (though varying according to the site involved and
patient age) is >95%, with a negative predictive value close
On bone scintigraphy, fractures are characterized by to 99%.
increased uptake of 99mTc-diphosphonates along the fracture The three-phase bone scintigraphy becomes abnormal
lines. In the acute stage, uptake may also be seen in the first shortly after the fracture; 95% of fractures in patients aged
and second phase of a three-phase bone scan. The time <65 years may be visualized at scintigraphy within 24 h and
elapsed between the trauma and positivity of the fracture in a 100% within 72 h, while in patients aged >65 years the bone
bone scan may vary, though. For instance, fractures in the scan becomes positive within 72 h. Bone scintigraphy can
extremities usually become positive within the first 24 h after also be used to estimate the age of the fracture in terms of
trauma, but fractures of the spine, pelvis, or femoral neck time elapsed since the acute event.
may become positive only after several days. This also The pattern of three-phase bone scintigraphy can be clas-
depends on patients’ age and on the concomitant occurrence sified into three types: acute, subacute, and healing. In the
of osteoporosis. In elderly patients, even hip fractures may acute phase (1–4 weeks), hyperemia and hematoma
not become sites of increased uptake for several days after formation around the fracture site cause increased blood flow
the traumatic event. (in the vascular phase of the three-phase scan) and increased
Patients with trauma are usually referred for planar X-rays activity in the blood pool image. In the subacute phase
and/or high-resolution CT. In rare cases, these modalities (6–12 weeks), with the formation of bone callus and a rela-
might not allow a diagnosis of a fracture, and bone scintigra- tive decrease of regional blood flow, increased tracer uptake
phy might be helpful. This applies in particular to fractures is noticed in particular around the fracture fissure (Fig. 24.30).
in osteoporotic bone, in which the fracture lines may be In the final healing phase, the bone callus is reabsorbed and
detected only with some difficulty, for instance, in case of replaced by normal bone tissue, thus gradually reducing the
sacral insufficiency fractures. Sacral insufficiency fractures intensity of tracer uptake. In uncomplicated fractures scintig-
are scintigraphically characterized by linear uptake lines raphy normalizes over 1 or 2 years; if increased uptake per-
next to the iliosacral joints connected by a horizontal line of sists, suspicion of infection or lack of union of the bone
increased uptake so that an “H” is formed, the so-called fracture should be considered.
Honda sign (see Fig. 24.26). They are usually accompanied
by fractures of the pubis and/or ischium. In osteoporotic ver-
tebral fractures, bone scintigraphy with SPECT/CT not only 24.6.2 Stress Fractures
provides the diagnosis but can also assess their age and flo-
ridity. This feature has some therapeutic relevance, since The proper definition and pathophysiology of the stress frac-
only patients with younger fractures benefit from vertebral ture currently encompass two processes with a s imilar end
osteoplasty. result. The fatigue fracture is the result of an abnormal load
In athletes presenting with pain of the lower extremities, upon normal bone, while the insufficiency fracture is the
bone scintigraphy can distinguish between insertion tendi- result of normal loading upon abnormal bone (osteopenic
nopathies (such as the so-called shin splints) and stress frac- bone). The fatigue fracture is the result of an abnormal repet-
tures. Again, SPECT/CT helps considerably in differentiating itive load upon normal bone and occurs during an abrupt
between increased uptake caused by fractures and that increase in frequency, duration, or intensity of activity when
caused by other conditions, in particular in anatomically bone resorption (osteoclasts) is greater than bone replace-
complex regions such as in the hands or feet. On CT imag- ment (osteoblasts).
ing, fracture lines can in many cases be seen or also contour This type of stress fracture commonly occurs in the young
disruptions of the bone involved. active individual, such as the athlete or military recruit.
Contributing risk factors, which are not mutually exclusive
of one another, can be divided into two broad categories:
24.6.1 Occult Fractures or Trauma extrinsic (training regimen, footwear, training surface, and
type of sport) and intrinsic factors (gender, age, race, and
While standard radiography is the primary imaging tech- overall fitness level, as well as skeletal, muscle, joint, and
nique in case of suspected fracture, bone scintigraphy is use- biomechanical factors). Females have a higher incidence of
ful to reveal occult fractures and to exclude a pathological fatigue fractures compared with males, in part as a conse-
cause when the radiological examination is still negative. quence of the “female athlete triad”: eating disorders, amen-
The areas most frequently affected by occult fractures are the orrhea, and osteoporosis. Through a complex interplay of
carpal bones, proximal femora, ankle, foot, and spine. nutritional deficiency, hypothalamic and estrogen abnormal-
Sensitivity of bone scintigraphy in disclosing occult frac- ities, as well as delayed menarche, reduced bone mineral
Fig. 24.30 Planar whole-

body scintigraphy obtained
about 3 h after administration
of 99mTc-HDP in a patient
with recent traumatic fracture
just below the left humeral
head. Markedly increased
tracer uptake at the fracture
site reflects active ongoing
mineralization of the healing
bone callus (modified from:
Volterrani D, Erba PA,
density places the female athlete at a significant risk for CT is useful for identification of longitudinal fracture
stress fractures. lines, for the differential diagnosis with osteoid osteoma, and
Insufficiency fractures are the result of normal loading for the evaluation of stress injury of the spine, whereas has
upon abnormally weakened bone. Several predisposing fac- limited value in case of chronic lesions and for the evaluation
tors have been identified as the cause of insufficiency frac- of transverse fractures [76].
tures, the common entity often being primary or secondary A repetitive stress on a given musculoskeletal tract causes
osteoporosis. Other risk factors include all the conditions a reabsorption and neoformation reaction of the bone tissue,
where bone elasticity and mineral content are compromised, mediated by the concerted action of osteoblasts and osteo-
such as rheumatoid arthritis, metabolic bone disease, neuro- clasts; this response is a functional adaptation of the bone
logical disorders, prior irradiation, total hip replacement, tissue, which is constantly stretched under load. However,
corticosteroid therapy, high-dose fluoride therapy, and this process implies that there is a phase where osteoclast
bisphosphonate therapy. Elderly and postmenopausal women activity is predominant; the occurrence of further stress in
are most at risk for developing insufficiency fractures. this period can cause microfractures (or even macroscopic
Stress fractures almost universally present with pain upon fractures).
activity and point tenderness. The pain is relieved with rest Bone scintigraphy plays a well-defined role in the diag-
and worsens when the activity is continued. Typically, the nosis of stress fractures, with 100% sensitivity even at an
presentation occurs when there has been a change in inten- early stage when 80% of such lesions are not still evident in
sity of the activity. Insufficiency fractures of the pelvis fre- the conventional radiological examination. In fact, bone
quently present as low back, buttock, and groin pain in the scintigraphy becomes positive about 2 weeks before the
elderly. Subchondral insufficiency fractures may present radiological appearance. Early diagnosis is very important,
with sudden onset of severe pain in the absence of or follow- for example, for athletes who practice sports at an advanced
ing only minor trauma and, therefore, may present in an level, both to prevent complications of an undiagnosed
emergency setting. fracture and to allow a faster recovery. The most common
X-ray is a first-line imaging procedure, despite its low fracture site is the tibia, followed by the metatarsal bones.
sensitivity in case of early stress fracture and difficulties of On bone scintigraphy, the stress fracture appears as a focal
interpretation in the presence of osteopenia. Stress frac- area of hyperactivity in the 3-h image, variably associated
tures may appear as subtle linear sclerosis (often perpen- with different degrees of changes in the flow blood and
dicular to major trabeculae), focal endosteal or periosteal blood pool phase, depending on the time elapsed from the
reaction, and/or fracture through one cortex with superim- fracture or if it has been subjected to new stresses. Based on
posed periosteal reaction. MR imaging is extremely sensi- the scintigraphic pattern, a fracture can be classified as
tive (100% sensitivity and 85% specificity), although follows:
typically a second-line modality, obtained when radio-
graphs are normal and pain persists or in athletes requiring • Grade I—slightly increased tracer uptake localized to the
a definitive diagnosis. A linear hypointense fracture line on cortex
T1-weighted and T2-weighted images with adjacent bone • Grade II—more intense and extended increased uptake
marrow and soft tissue hyperintensity on T2 fat saturated or • Grade III—increased uptake extended to more than half
short tau inversion recovery (STIR) sequences (edema) are the bone amplitude
typical findings. Periosteal new bone will exhibit low sig- • Grade IV—increased uptake extended to the entire bone
nal in all sequences, and adjacent soft tissue edema may thickness
also be present. Fredericson et al. provided a classification
into discrete grades based on fluid-sensitive and In general, stress fractures of the proximal tibia tend to
T1-weighted characteristics of the stress injury of the tibial heal earlier than those of the distal tract. Furthermore, frac-
bone in runners [74]. Further work by Kijowski et al. found tures of the anterior aspect of the tibia and the malleolus are
a correlation between the grade of the lesions and the esti- more frequently associated with complications (such as non-
mated time to return to sport activities [75]. MRI is also conjunction, poor positioning, and avascular necrosis) as
important for diagnosing subchondral insufficiency frac- compared to the corresponding posterior aspects. Femoral
tures, which are not apparent on radiographs unless there is neck fractures are mostly diagnosed in athletes and in older
linear subchondral lucency or collapse. The main limitation people with osteoporosis and may be of two types: with
of MR is the possibility of misdiagnosis, since peripheral compression (located at the lower margin of the femoral
edema in the T2-weighted sequences can be present in case neck) and with loading (located at the top edge).
of infection and tumor. Therefore, MR imaging is currently Pelvic occult fractures are mainly due to osteoporosis;
reserved for the evaluation or follow-up of athletes per- they are usually located along the sacrum or pubis and can be
forming sports activities at a competitive level or in combi- misinterpreted as metastatic disease at scintigraphy. In 85%
nation with scintigraphic data, if they are not conclusive. of patients with sacral osteoporosis, bone scintigraphy can
also detect other unknown fractures, located above all in the spot” on bone scintigraphy. Reparative processes, starting at
spine, ribs, and pubic bone. the periphery of the necrotic bone, begin several days after
Spondylolysis, characterized by a defect in the posterior the insult and persist for months or years thereafter. In this
arch of the vertebrae at the pars interarticularis, represents process, the core of the necrosis remains hypometabolic; this
the primary diagnosis in 47% of young athletes with negative creates a typical scintigraphic pattern of a “cold-in-hot” spot,
history for previous trauma but with lumbar back pain. Often often termed as halo or crescent sign.
it is bilateral and can, therefore, cause spondylolisthesis. For Transient osteoporosis/bone marrow edema, a usually
the diagnosis, in addition to the standard scintigraphic study self-limiting condition, also affects the epiphysary bone; is
(which shows an increased uptake at the lesion), SPECT and painful, in its early stages markedly hypermetabolic; and
SPECT/CT acquisitions provide best results, especially for thus raises the issue of differential diagnosis with femoral
accurate localization of the site involved and as prognostic head necrosis. Aseptic necrosis may also affect other bones,
index for potential healing. In fact, a negative scintigraphy in such as the distal femur, the os lunatum, or the os scaphoi-
presence of a positive radiography indicates either inactive deum. Bone infarctions are also due to vascular disturbances
disease or healed fracture and excludes spondylolysis as a but—at variance with bone necrosis—in most cases are
cause of pain. asymptomatic. They may be idiopathic or triggered by a
[18F]FDG PET/CT can also demonstrate increased tracer variety of risk factors such as alcohol, sickle cell anemia,
uptake at the site of stress fracture. Some patterns of uptake, radiation therapy, cytotoxic medication, or steroids. Their
such as the “Honda sign” for sacral insufficiency fractures, scintigraphic appearance depends on the stage when they are
can be diagnostic, with sensitivity and positive predictive imaged, similarly as in the case of bone necrosis. On CT
value of up to 96% and 92%, respectively [77, 78]. imaging, they tend to have a sclerotic rim with a central lytic
zone that may also feature calcifications.
In osteochondrosis dissecans, bone ischemia can lead to
Key Learning Points
bone fragmentation of the joint end of a bone. This condition
• On bone scintigraphy, fractures are characterized
is found above all at the knee or talus. The involved zones are
by increased uptake of 99mTc-diphosphonates along
generally visible on planar/SPECT imaging as areas with
the fracture lines.
increased 99mTc-diphosphonate uptake. SPECT/CT greatly
• Fractures in the extremities usually become positive
increases the diagnostic accuracy.
within the first 24 h after trauma, but fractures of the
spine, pelvis, or femoral neck may become positive
only after several days.
24.7.1 Aseptic Necrosis of the Femoral Head
• In uncomplicated fractures scintigraphy normalizes
over 1 or 2 years.
Osteonecrosis (often termed avascular necrosis with
• In osteoporotic vertebral fractures, bone scintigra-
involvement of the epiphyseal regions) is a relatively com-
phy with SPECT/CT not only provides the diagno-
mon disease in which ischemic death of the cellular compo-
sis but can also assess their age and floridity.
nents of bone and marrow occurs. The femoral heads are the
• The fatigue fracture is the result of an abnormal
most commonly affected sites, with estimates of symptom-
load upon normal bone, while the insufficiency
atic femoral head osteonecrosis of 2–4.5 per patient-year,
fracture is the result of normal loading upon abnor-
resulting in 10,000–20,000 new cases annually in the United
mal bone (osteopenic bone).
States [79]. Since most patients are asymptomatic, most
• Bone scintigraphy plays a well-defined role in the
likely this incidence significantly underestimates the true
diagnosis of stress fractures, with 100% sensitivity
prevalence of osteonecrosis. Osteonecrosis affects both
even at an early stage when 80% of such lesions are
children and adults, and the predisposing factors include
not yet obvious on conventional X-ray
dislocation of the hip, femoral neck fracture, corticosteroid
examination.
usage, alcoholism, collagen vascular disease, hemoglobin-
opathies, Gaucher disease, caisson disease, and some skel-
etal dysplasias [80]. The spontaneous form (or
Legg-Calvé-Perthes disease) affects children at an age of
24.7 Vascular Disorders 4–8 years. Secondary forms are more common in adulthood.
Nontraumatic osteonecrosis is bilateral in 70–80% of cases,
Aseptic bone necroses are caused by acute disturbances in which further increases the disability due to femoral head
blood supply, the exact cause remaining unclear in most collapse.
cases. In their very early phase, the necrotic bone is hypoper- No specific physical findings or laboratory tests can reli-
fused and hypometabolic, presenting therefore as a “cold ably establish the diagnosis of osteonecrosis. When sus-
pected on clinical ground, osteonecrosis must be confirmed In addition, the diagnostic value of scintigraphy is less
by diagnostic imaging, or through biopsy. Imaging methods affected than CT and MR by artifacts due to recent surgical
that can assist in establishing the diagnosis include planar interventions. In Legg-Calvé-Perthes disease, a photopenic
X-rays, CT, bone scintigraphy, and MRI, with or without area can be seen in the early stages, followed by hyperactiv-
contrast enhancement. These methods vary considerably in ity due to repair reaction in later stages.
their cost, diagnostic accuracy, and the information
provided.
Although the optimal treatment for femoral head osteone- 24.7.2 Sympathetic Reflex Dystrophy
crosis is debated, early diagnosis is important both for
excluding other causes of pain (infection, cancer, fracture, This is mainly a post-traumatic syndrome, characterized by
arthritis, femoro-acetabular impingement syndrome, etc.) pain, functional impotence, vasomotor instability, swelling,
and for assessing the efficacy of treatment. and skin dystrophy. Because of an altered neurovegetative
X-ray should be always performed as the first-line balance, opening of vascular shunt occurs at the affected site,
imaging. In children, the earliest radiographic findings of resulting in low blood flow resistances. Diagnostic sensitiv-
osteonecrosis include a smaller ossific nucleus, increased ity of the three-phase bone scintigraphy is very high (>90%),
radiodensity, subchondral fracture, and metaphyseal especially at an early stage when radiological imaging can
radiolucencies. Subsequently, fragmentation, resorption, still be completely negative. The characteristic scintigraphic
reossification, and remodeling of the affected femoral pattern is represented by increased blood flow and increased
head and neck are seen. The main role of CT is to deter- capillary permeability (compared to the contralateral limb),
mine the severity of articular collapse and its location usually only apparent in the first 2–3 months after the onset
and to detect early secondary degenerative joint disease. of the disease (Fig. 24.31). At later stages, increased uptake
This information is useful for surgical planning, e.g., for in the periarticular region at the affected limb, which can per-
rotational osteotomy, arthroplasty, resurfacing proce- sist for up to 1 year from the onset of symptoms, is generally
dures, or joint replacement [81]. In the pediatric popula- observed. A particular form, more common in childhood and
tion, CT is not commonly used for assessment of adolescence, is characterized by low activity accumulation in
osteonecrosis. the hip, both in the vascular phase and in the delayed, meta-
MRI is the most sensitive and specific imaging modality bolic phase of the scan [90].
to detect osteonecrosis [82, 83]. In the adult population, the
differential diagnosis should always be made versus tran-
sient osteoporosis and subchondral insufficiency fracture
[84, 85]. In the pediatric population, there is a definite role Key Learning Points
for contrast-enhanced MRI [86, 87]. Using a dynamic tech- • Aseptic bone necroses are caused by acute distur-
nique, the disease can be detected at an earlier stage than bances in blood supply, the exact cause remaining
with other MRI techniques; in particular, in Legg-Calvé- unclear in most cases.
Perthes disease, there is increased peak enhancement and • In the very early phase, the necrotic bone is hypo-
delayed time-to-peak enhancement [88]. Furthermore, this perfused and hypometabolic, presenting therefore
technique has been used to identify femoral heads at risk for as a “cold spot” on bone scintigraphy.
the development of osteonecrosis subsequent to femoral • Reparative processes start at the periphery of the
neck fracture [89]. necrotic bone and begin several days after the insult,
Although planar bone scintigraphy is less sensitive than persisting for months or years thereafter. In this pro-
MRI (90% versus 100%) [82], it remains a valid option in cess, the core of the necrosis remains hypometabolic;
patients with suspected osteonecrosis. SPECT or preferably this creates the typical scintigraphic pattern of a
SPECT/CT improves overall sensitivity up to 97%. Different “cold-in-hot” spot, often termed halo or crescent sign.
sensitivities reported by different studies are mainly due to the • Sympathetic reflex dystrophy is a mainly post-trau-
different etiopathogenetic factors that can cause aseptic necro- matic syndrome, characterized by pain, functional
sis and by the dynamic nature of the process. For example, impotence, vasomotor instability, swelling, and
osteonecrosis following fracture of the femoral neck is due to a skin dystrophy.
sudden and substantial reduction of blood flow (the scinti- • The characteristic scintigraphic pattern of sympa-
graphic images, thus, show a photopenic area) (see Figs. 24.3 thetic reflex dystrophy represented by increased
and 24.4). In the form secondary to corticosteroid therapy, the blood flow, increased capillary permeability, and
presence of small photopenic lesions with microfractures and increased bone metabolism in the affected limb as
repair might be visible in the bone scan as an area of increased compared to the contralateral.
uptake.
Fig. 24.31 Dynamic three-phase bone scintigraphy in a patient with ity/time curves from 0 to 300 s reported in lower left panel show mark-
recent sympathetic reflex dystrophy (also known as Sudeck’s syn- edly increased blood perfusion of the right foot versus the left foot. The
drome) of the right foot—evaluated in the subacute phase of recovery. right panel depicts the late scintigraphic image (acquired about 3 h
The left panel depicts some selected initial frames from the dynamic post-administration of the bone-seeking agent), showing markedly
acquisition, showing enhanced perfusion of right foot versus the left increased tracer uptake in right foot, reflecting ongoing remodeling of
foot (ROIs defined on the two feet are depicted in last frame). The activ- the involved bones
24.8 Inflammation treated properly, this process leads to articular deformities

with consequent pain and loss of function. The disease typi-
Inflammation usually leads to increased bone metabolism. cally affects the small joints of the hands and feet symmetri-
Depending on its floridity, increased uptake can be visible in cally, with erosive arthritis that sometimes progresses to
the first two phases of a three-phase bone scintigraphy. systemic involvement. The spine is affected late, with pre-
However, these findings are often nonspecific. dominant cervical involvement of the atlanto-axial joint and
Because of its ability to produce whole-body images, of the inter-apophyseal joints.
bone scintigraphy can be helpful in evaluating the activity of The disease, easily diagnosed by clinical and laboratory
rheumatic diseases. In patients with infection/inflammation, parameters, is characterized by quiescent and exacerbation
bone scintigraphy provides information on the floridity of phases. Therefore, it is very important to define the state of
individual lesions as well as important diagnostic informa- activity of the disease to modulate therapy and prevent
tion on the specific entity—based on the distribution pattern progression.
of the segments involved. The radiological pattern is well defined, with altera-
tions occurring isolated or in association (according to the
stage), representing the key diagnostic signs. In the initial
24.8.1 Rheumatoid Arthritis stages, swelling of the soft joints associated with effusion
is observed, as well as juxta-articular subchondral corti-
Rheumatoid arthritis is an autoimmune inflammatory polyar- cospinal osteoporosis, symmetric reduction of articular
thritis that affects about 1% of the Western population and is interlines (as a consequence of cartilage damage), and
characterized by lymphocytic infiltration of the synovia that bone erosions. CT imaging does not have specific indica-
erodes the articular cartilage and the underlying bone. If not tions, except in special cases to estimate bone damage of
Fig. 24.32 Planar whole-

body imaging (a) and spot a b
acquisition of the hands (b)
obtained about 3 h after
administration of 99mTc-MDP
in a patient with active
rheumatoid arthritis involving
both small and large joints
from: Versari A. Nuclear
medicine imaging in chronic
inflammatory diseases. In:
Lazzeri E, Signore A, Erba
PA, Prandini N, Versari A,
D’Errico G, Mariani G, Eds.
Radionuclide Imaging of
Infection and Inflammation –
A Pictorial Case-Based Atlas.
Milan: Springer;
2013:289–331)
large joints. Ultrasound is useful in the evaluation of the anatomic sites/joints that are difficult to explore with X-rays
initial forms (which still do not involve the bone), thanks (sternoclavicular, acromioclavicular, temporomandibular,
to its ability to detect early-stage arthrosynovitis and and atlanto-axial joints). Bone scintigraphy can be used to
tenosynovitis; this method constitutes an optimal guide identify early rheumatoid arthritis, the form of recent onset
for endoarticular and infiltration procedures and monitor- (<3 months), in which the erosive bone damage has not yet
ing the response to therapy over time. MR imaging can developed and where proper therapy has the maximum prob-
evaluate joint damage and extraarticular involvement, ability of healing. At this stage, traditional radiological imag-
identifying the initial damage of the bone and bone coming is negative, and MR imaging is suboptimal. Instead, bone
partment as capsular ligamentous. This technique also scintigraphy reveals a typical pattern of symmetrically
provides prognostic indications on disease severity and increased tracer uptake in the affected joints.
response to specific therapies. In addition, it permits to Another important application concerns the identification
evaluate lesions of the atlanto-axial joint and, in particu- of complications that may occur during the disease, such as
lar, erosions of the epistropheus tooth. fractures (e.g., during corticosteroid therapy). The sites most
The role of bone scintigraphy (which has been relevant in frequently affected are the vertebral bodies, femoral neck,
the past) is currently somewhat limited. However, the exam proximal tibial metaphysis, sacrum, and metatarsal bone. In
is still able to provide whole-body imaging and to detect these cases traditional radiology is negative at an early stage
early joint involvement. In particular, the degree of 99mTc- and only later can reveal radiotransparency or thickening
diphosphonate uptake in the juxta-articular area is propor- (linked to trabecular interpenetration) at the fracture
tional to the activity of the disease (Figs. 24.32 and 24.33) fissures.
and is higher in recent acute lesions. Juxta-articular osteopo-
rosis presents as a more or less intense bipole shaped dif-
fusely increased uptake. 24.8.2 Seronegative Spondyloarthropathies
Bone erosion areas appear as areas of intense tracer
uptake on a background of diffusely increased uptake. Seronegative spondyloarthropathies constitute a heteroge-
Identification of the involved joints is very early, before it neous group of diseases that include ankylosing spondyli-
can be detected by traditional radiology. In addition, the tis, psoriatic arthropathy, Reiter’s syndrome, and
total-body imaging mode permits localization of disease in inflammatory bowel disease. They are called seronegative
Fig. 24.33 Bone
scintigraphy obtained in a
22-year-old man with pain in
the right foot and pelvis.
Planar anterior whole-body
scan (left panel) shows
increased tracer uptake in the
calcanear region of the left
foot. 3D-volume-rendered
hybrid image (right panel)
demonstrates Achilles
enthesitis and calcaneo-talar
arthritis. Spot dorsal view of
the pelvis (not shown here)
discloses predominantly
left-sided sacroiliitis. The
findings are due to psoriatic
arthropathy
as they are not characterized by the production of anti-IgG rheumatoid arthritis, as the evolution of this group of
antibodies as it occurs in rheumatoid arthritis. They are fur- pathologies is slower and with a minor inflammatory com-
ther characterized by HLA-B27 positivity. In these dis- ponent. Moreover, axial joints are difficult to evaluate due
eases, the spine is often affected, appearing on X-rays with to the complex three-dimensional geometry and overlap-
the so-called syndesmophytes that resemble osteophytes ping with others skeletal structures. This diagnostic delay
but run in vertical direction. Sacroiliitis is also frequently can be significantly reduced thanks to bone scintigraphy
found in these disorders and readily diagnosed by enhanced that, with a total body scan, is able to show multiple juxta-
tracer uptake of these joints on the bone scan. Furthermore, articular area of increased uptake with the typical spatial
these conditions are characterized by the common feature distribution in axial and appendicular segments and at the
of enthesitis lesions, i.e., inflammation and degeneration of level of the enthesis.
the insertion site of tendons, bands, and ligaments of the Increased tracer uptake at the sternal, costotransverse, and
bone. These events create dystrophic ossification. The dis- sacroiliac bones and in the tendon insertion areas (such as
tribution of these lesions is quite typical, given the main posterior calcareous apophysis, large trochanter, ischial
involvement of the axial skeleton and the synchondrosis tuberosity) is highly characteristic and diagnostic of
sacroiliac joints, which are invariably affected, although at spondylarthritis.
different times. Less frequently the small joints of the Currently, MRI is the preferred modality for early assess-
wrists and of the hands are involved, in an asymmetrical ment of these diseases, providing morphological evidence of
way. The diagnostic workup is based on plain X-ray, degeneration. Bone scintigraphy allows the early diagnosis
although this imaging modality is less satisfactory than in of sacroiliitis, which presents with atypical clinical signs.
Bone scintigraphy analyzed by calculating the uptake enzymes (thereby causing osteolysis), pus can be formed,
indexes of synchondrosis (on either planar and/or SPECT and the small vessels can be blocked causing sequesters
imaging) has a 100% negative predictive value. In contrast, and necrosis; with progressing infection, OM can become
an abnormal uptake index makes the diagnosis of sacroiliitis chronic, causing sclerosis and deformity. Although any
very likely already at the initial stage. SPECT increases by bone may in principle be affected, the maxilla, mandible,
20–50% the number of lesions detected in the lumbosacral long bones, and spine are most often involved, due to their
region as compared to planar scintigraphy. Nonetheless, high blood supply [94, 95]. Acute OM most often occurs
diagnostic confirmation of the synchondrosis by CT is nec- in children and in adults with compromised host resis-
essary, due to the low specificity of scintigraphy since tance (immunosuppressive therapy, i.v. drug abuse, infec-
increased tracer uptake might be present also in case of tious tooth-root canal teeth, etc.) [94]. The incidence of
osteoarthritis, functional overload in patients with scoliosis,
OM in developed countries is less than 2% of the general
and dissection of the limbs, and can be physiologically population per year (13/100,000 among children) [96].
observed in subjects in prepuberal age with cartilages of OM requires prolonged antibiotic therapy (usually
florid growth. Intensity of uptake of the bone-seeking radio- 6–12 weeks) and sometimes may require surgical debride-
pharmaceutical declines over time, with the evolution of ment. Severe cases may even lead to amputation of a limb
lesions to the chronic phase, in which the tendency to synos- [97].
tosis is manifested. Imaging has the primary role of achieving a prompt diag-
PET provides an alternative means that can elucidate nosis of OM and of evaluating its extent. Standard X-rays
molecular and cellular changes of the affected joints. Since represent the first imaging modality used when OM is sus-
changes in bone metabolism are present long before clear pected; however, plain radiography may be normal up to
morphological signs develop, PET with either 18F-fluoride or 2–4 weeks after onset of OM. Although extremely variable
[18F]FDG can identify and characterize early-stage disease, values are reported, sensitivity and specificity of radiography
allowing for more timely treatment and closer monitoring of for acute OM are generally considered to be rather low (28–
disease progression [91, 92]. Furthermore, 18F-fluoride has 93% and 50–91%, respectively). At later stages, reliable
greater accuracy and sensitivity than 99mTc-MDP scintigraphy X-ray signs of OM are cortical erosions, bone resorption
and can reliably identify areas with associated inflammation-and/or destruction, and periosteal reaction. Even if rarely
induced changes and high mechanical stress [93]. (depending on the specific bacteria responsible for the infec-
tion), gas can be seen in soft tissues.
Key Learning Points Ultrasonography has no major role for evaluating bone
• Inflammation usually leads to increased bone involvement when OM is suspected, although it can be used
metabolism. (with power Doppler) to evaluate soft tissues involvement,
• Depending on its floridity, increased uptake can be presence of abscesses or sinus tracts, and periosteal reaction.
observed in the first two phases of a three-phase In addition, ultrasound guidance can aid to guide needle
bone scintigraphy. aspiration of fluid or to place a drainage tube.
• Bone scintigraphy has a high sensitivity in detect- The diagnostic performance of CT imaging is clearly
ing joint inflammation and may be used to visualize superior to plain X-rays, particularly by allowing an earlier
joint involvement in the rheumatic diseases. diagnosis. Its use is, however, limited by the high amount of
ionizing radiation and the impossibility of evaluating bone
marrow edema.
MRI is superior to other radiological imaging modali-
24.9 Infection ties in evaluating patients with suspected OM, because of
its ability to simultaneously assess soft tissues and bone/
Osteomyelitis (OM), the condition characterized by microor- bone marrow structures. Sensitivity and specificity have
ganisms infecting the bone and/or bone marrow, is not a been reported to reach up to 100% and 95%, respectively.
single condition but rather a spectrum of different diseases In particular, MRI can be useful for distinguishing viable
that vary based on the host’s immunological status, the pre- from necrotic tissues and for detecting the presence of air/
existing pathology in bone, and the type of the infectious gas bubbles within soft tissues. OM is usually character-
microorganism. All of these factors affect the site of infec- ized by low signal on T1-weighted images and bright sig-
tion onset and the result of diagnostic investigations. nal on T2-weighted images, better appreciated when fat
OM can be the result of direct infection after injury or saturation is used. Increased signal on T2-weighted images
can occur if microorganisms reach the bones via the without signal modification on T1-weighted images usu-
bloodstream—from other infective foci in the body. Once ally represents reactive edema of the surrounding soft tis-
the bone is infected, the activated leukocytes release sues due to infection rather than bone infection itself. The
use of i.v. gadolinium-based contrast agents in conjunction necessarily infection. In the latter case, scintigraphy with
with fat-saturated T1-weighted sequences increases the labeled leukocytes will be useful to exclude the presence of
diagnostic accuracy. MRI is helpful in detecting septic infection. The pattern of increased uptake at bone scintigra-
involvement of the joint space and nearby soft tissues. phy is especially important in patients with chronic osteomy-
MRI can also detect joint effusion and synovial thicken- elitis, in whom it is crucial, for proper management, to
ing; in this instance, the absence of bone marrow changes distinguish between increased uptake caused by underlying
on T1-weighed images and marrow changes confined to infection and increased uptake due to bone remodeling.
the subchondral bone indicate simple reactive edema SPECT/CT acquisition can help define bone involvement
rather than bone infection. (Fig. 24.34). Indeed, in the presence of orthopedic devices,
the sensitivity and specificity of planar bone scintigraphy are
relatively low (67% and 50%, respectively).
24.9.1 Radionuclide Imaging of In patients with high clinical probability of OM, scintigra-
Infection/Inflammation phy with labeled autologous leukocytes (usually 99mTc-
HMPAO-leukocytes or 99mTc-HMPAO-WBCs—although
Different radiopharmaceuticals can be used in the diagnostic 111
In-oxine-WBCs may be preferred in selected cases) repre-
flowchart of infection and/or inflammation of peripheral sents the nuclear medicine procedure of choice [99] (see exam-
bones, including 99mTc-diphosponates, labeled autologous ples in Figs. 24.35–24.38). Nonetheless, in patients with
leukocytes, anti-granulocytes antibodies, [18F]FDG, and neutropenia (caused by, e.g., chemotherapy or various immu-
67
Ga-citrate [98]. A high or low clinical probability of OM nodeficiencies), the use of [18F]FDG PET/CT should be con-
dictates the choice of the radiopharmaceutical to be used. sidered. The use of [18F]FDG PET/CT can also be considered
In case of low probability of peripheral OM, dynamic in patients with suspected OM in general, particularly if no
three-phase bone scintigraphy with 99mTc-diphosponates rep- fractures are present. Scintigraphy with labeled 99mTc-HMPAO-
resents the front-line nuclear medicine procedure of choice. WBCs represents the gold standard for diagnosing neutrophil-
As described above, dynamic bone scintigraphy permits mediated infectious diseases. The scintigraphic images should
evaluation of blood flow immediately following i.v. bolus be acquired 30 min (early images), 3–4 h (delayed images), and
injection of the radiopharmaceutical, as well as the presence 20–24 h (late images) post reinfusion of the labeled leuko-
of local edema (blood pool image at 5 min post-injection) cytes. Time over which to acquire the images should be
and the level of osteoblastic activity (standard late imaging adjusted so as to accumulate counts with sufficient statistics
3–4 h after tracer injection). A negative scan is sufficient to according to radionuclide decay in the intervened time.
exclude presence of acute inflammation and/or infection. A Scintigraphic images should be evaluated by adopting
positive scan, i.e., increased perfusion in early imaging and the correct interpretation criteria to distinguish infective
delayed increased tracer uptake in the region of suspected from inflammatory disease of bone. The diagnostic accu-
infection, suggests the presence of inflammation—but not racy of 99mTc-HMPAO-WBC scintigraphy is >95% [100].
Fig. 24.34 99mBone scintigraphy performed in a patient who had suf- increased uptake. CT scan (right panel) discloses sclerotic as well as
fered a bullet injury to the right foot about 1 year earlier. Planar scan lytic changes in that bone. The findings are due to osteomyelitis. Note
(left panel) discloses a hot spot in the right midfoot. On the fusion metallic remnants of bullet on the CT image
SPECT/CT image (center panel), an area of the navicular bone exhibits
Fig. 24.35 99mTc-HMPAO-WBC scintigraphy performed in a patient cent soft tissue (modified from: Lazzeri E. Nuclear medicine imaging of
with osteomyelitis of the left tibia to evaluate the extent of infection. bone and joint infection. In: Lazzeri E, Signore A, Erba PA, Prandini N,
Top panel: the 24-h planar image (anterior and posterior, left and right) Versari A, D’Errico G, Mariani G, Eds. Radionuclide Imaging of
shows clear accumulation of labeled leukocytes in the diaphysis of the Infection and Inflammation – A Pictorial Case-Based Atlas. Milan:
left tibia. Bottom panel: fused SPECT/CT images allow the precise Springer; 2013:39–80)
assessment of the extent of the infection in the bone, extending to adja-
Baseline SPECT/CT Post-therapy SPECT/CT
Fig. 24.36 99mTc-HMPAO-WBC scintigraphy performed at two differ- strating disappearance of the abnormal focus of labeled leukocyte accu-
ent times in a patient with osteomyelitis of right tibia. Left panel: fused mulation at the site of the prior infection (modified from: Lazzeri
SPECT/CT acquired to confirm the clinical suspicion of osteomyelitis, E. Nuclear medicine imaging of bone and joint infection. In: Lazzeri E,
showing clear accumulation of labeled leukocytes at the site of infec- Signore A, Erba PA, Prandini N, Versari A, D’Errico G, Mariani G,
tion, confined to the bone and bone marrow. Right panel: fused SPECT/ Eds. Radionuclide Imaging of Infection and Inflammation – A Pictorial
CT of the scan acquired after completion of antibiotic therapy, demon- Case-Based Atlas. Milan: Springer; 2013:39–80)
Fig. 24.37 99mTc-HMPAO-WBC scintigraphy acquired 4 h post- depicting exact localization of the infection site (modified from: Lazzeri
injection in a patient with suspected infection of the right foot. Left E. Nuclear medicine imaging of bone and joint infection. In: Lazzeri E,
panel: fused SPECT/CT images (transaxial, left; sagittal, middle; coro- Signore A, Erba PA, Prandini N, Versari A, D’Errico G, Mariani G,
nal, right) of the feet showing focal accumulation of labeled leukocytes Eds. Radionuclide Imaging of Infection and Inflammation – A Pictorial
in right talonavicular joint, consistent with septic arthritis. Right panel: Case-Based Atlas. Milan: Springer; 2013:39–80)
3D reconstruction of fused SPECT/CT images of lower limbs clearly
Fig. 24.38 99mTc-HMPAO-WBC scintigraphy performed in a patient intense accumulation of labeled leukocytes in frontal sinus (modified
with suspected frontal osteomyelitis. Left panel: planar anterior (upper) from: Lazzeri E. Nuclear medicine imaging of bone and joint infection.
and left lateral (lower) images acquired 4 h post-injection, showing In: Lazzeri E, Signore A, Erba PA, Prandini N, Versari A, D’Errico G,
focal accumulation of labeled leukocytes in the frontal region. Right Mariani G, Eds. Radionuclide Imaging of Infection and Inflammation –
panel: fused SPECT/CT images (sagittal, left; axial, right), showing A Pictorial Case-Based Atlas. Milan: Springer; 2013:39–80)
SPECT/CT acquisition is crucial especially to evaluate the in general bone marrow accumulation is higher when using
exact extension of infection, i.e., whether the infection is radiolabeled anti-granulocyte antibodies than with the in vitro
confined to the bone and bone marrow or it also involves labeled leukocytes. Sensitivity of radiolabeled anti-granulocyte
the adjacent soft tissues. This evaluation is the optimal antibody scintigraphy for the diagnosis of peripheral bone OM is
basis for proper clinical management, particularly in some about 85%, with 83% specificity [106]. The combined scan with
conditions, e.g., the diabetic foot (see Figs. 24.39 and 99m
Tc-colloids can also be performed in case of equivocal find-
24.40). ings at anti-granulocyte antibody scintigraphy [102].
Scintigraphy with combined labeled leukocytes and a Scintigraphy with 67Ga-citrate (images being acquired at
bone marrow agent (generally a 99mTc-colloid) can be used 6, 24, and 48 h after tracer injection) is an alternative option
for diagnosing bone infection in case of equivocal findings at for radionuclide imaging where scintigraphy with labeled
leukocyte scintigraphy, particularly in those cases where an leukocytes or anti-granulocyte antibodies is not available.
expanded bone marrow activity is present. Both labeled leu- Nevertheless, this imaging technique has become obsolete in
kocytes and radiocolloids accumulate in the healthy bone most Nuclear Medicine centres around the world.
marrow, while in OM labeled leukocytes accumulate, PET/CT with [18F]FDG and/or 18F-fluoride are being
whereas radiocolloids do not [101] (see Fig. 24.41). increasingly used to investigate patients with suspected OM,
Radiolabeled anti-granulocyte monoclonal antibody prepa- although neither of these two tracers can be considered
rations can also be used, particularly for imaging chronic OM infection-specific [107, 108]. [18F]FDG has also been sug-
[102–105]. In this case, images should be acquired at 3–4 h gested to label autologous leukocytes in vitro [109], although
and 20–24 h post i.v. injection of the radiopharmaceutical. the short physical half-life of 18F (110 min) does not allow
Biodistribution of the leukocytes labeled in vivo with the anti- satisfactory evaluation of the kinetics of labeled leukocyte
body agents is similar to that of 99mTc-HMPAO-WBCs, although accumulation at the site under investigation.
Fig. 24.39 SPECT/CT acquisition during 99mTc-HMPAO-WBC scin- bone structure of the toe (reproduced with permission from: Prandini N,
tigraphy performed in a diabetic patient with suspected osteomyelitis of Beretta F. Nuclear medicine imaging of diabetic foot. In: Lazzeri E,
the right toe (MIP image in right lower panel). Sagittal section (CT, Signore A, Erba PA, Prandini N, Versari A, D’Errico G, Mariani G,
upper left; SPECT, upper right; fused SPECT/CT, lower left) demon- Eds. Radionuclide Imaging of Infection and Inflammation – A Pictorial
strating accumulation of labeled leukocytes selectively involving the Case-Based Atlas. Milan: Springer; 2013:253–269)
The European Association of Nuclear Medicine their conclusions concerning radionuclide imaging, the
(EANM), the European Society of Radiology (ESR), the pre-test probability of infection should be considered for
European Bone and Joint Infection Society (EBJIS), and choosing between three-phase bone scintigraphy and scin-
the European Society of Microbiology and Infectious tigraphy with labeled autologous WBCs. High probability
Diseases (ESCMID) have recently released a Consensus of infection based on findings such as fracture, recent sur-
Document for the diagnosis of peripheral bone infection in gery, osteosynthesis, and highly positive serological tests
adults (see below under “Further reading”). According to favours the choice of scintigraphy with labeled autologous
Fig. 24.40 SPECT/CT acquisition during 99mTc-HMPAO-WBC scin- soft tissues, thus ruling out osteomyelitis (reproduced with permission
tigraphy performed in a diabetic patient with right plantar ulcer and sus- from: Prandini N, Beretta F. Nuclear medicine imaging of diabetic foot.
pected osteomyelitis of the right foot (SPECT, upper row; CT, middle In: Lazzeri E, Signore A, Erba PA, Prandini N, Versari A, D’Errico G,
row; fused SPECT/CT, lower row). Sections in different planes demon- Mariani G, Eds. Radionuclide Imaging of Infection and Inflammation –
strate that the focus of labeled leukocyte accumulation involves only the A Pictorial Case-Based Atlas. Milan: Springer; 2013:253–269)
WBCs. Because of its high sensitivity and negative predic- CT acquisitions). When autologous leukocyte scintigraphy
tive value, three-phase bone scintigraphy is recommended (or anti-granulocyte antibody scintigraphy) turns out to be
as the first imaging modality of choice in patients with low negative, [18F]FDG PET/CT is generally the conclusive test
pre-test probability of infection; in case of a positive three- for infection versus non infection. On the other hand, PET/
phase bone scan, the next radionuclide imaging modality is CT with [18F]FDG is recommended as the first imaging
either scintigraphy with autologous leukocytes or with modality of choice especially in patients with high suspi-
anti-granulocyte antibodies (with both planar and SPECT/ cion of hematogenous spread of the infection.
negative bacteria may also cause postsurgical SD, the most

frequently isolated Gram-negatives including Escherichia
coli, Pseudomonas aeruginosa, Klebsiella pneumoniae,
Enterobacter cloacae, Bacteroides, and Proteus species.
The diagnosis of SD may be challenging due to the non-
specific clinical presentation and laboratory findings and to
the need of performing different diagnostic tests including
serologic, radiographic, and microbiological examinations.
Early diagnosis of SD, with the identification of the etiology
and description of location, extent, and severity of the infec-
99mTc-HMPAO-WBC 99mTc-Colloid tive process, is crucial to choose appropriate management.
Radiologic imaging is important to define the correct diag-
Fig. 24.41 Evaluation of a patient with suspected osteomyelitis of left nosis as early as possible, to assist in the percutaneous biopsy
tibia, to rule out the presence of acute infection. Left panel: 99mTc-
HMPAO-WBC scintigraphy (planar image), showing mild accumulation
(if needed), for follow-up of the disease, and to evaluate pos-
of labeled leukocytes in proximal and middle portion of diaphysis of left sible complications. Plain X-ray imaging has low sensitivity
tibia, suggesting the presence of inflammation without infection. Right and specificity, even if it is usually the first imaging modality
panel: scintigraphy with 99mTc-nanocolloidal albumin, showing uptake requested when SD is suspected. The detection of signs and
at the same site of 99mTc-HMPAO-WBC accumulation. The overall scin-
tigraphic pattern confirms the presence of normally functioning bone
abnormalities depends on their entity and can be seen only
marrow, thus ruling out the presence of acute infection (modified from: when bone destruction exceeds 30–50% of cortical bone.
Lazzeri E. Nuclear medicine imaging of bone and joint infection. In: Erosion of the anterior corner of the vertebral endplate, pro-
Lazzeri E, Signore A, Erba PA, Prandini N, Versari A, D’Errico G, gressive loss of disk height, narrowing of inter-vertebral space,
Mariani G, Eds. Radionuclide Imaging of Infection and Inflammation –
A Pictorial Case-Based Atlas. Milan: Springer; 2013:39–80)
and osteolysis with further destruction of the subchondral
plate can be seen in case of advanced d isease—associated
with soft tissue swelling and pre- or paravertebral soft tissue
24.9.2 Spondylodiscitis densities, occasionally with fluid level. Erosions can involve
the entire vertebral endplate and signs of new bone formation
Spinal infections include vertebral osteomyelitis (infection (including peripheral sclerosis, osteophytosis, and a buildup of
of the vertebral body), discitis (infection of the interverte- osteolytic lesions) can coexist. Reactive sclerosis may some-
bral disk), and spondylodiscitis (SD) (infection of two times produce an “ivory” sign as the only evident change.
adjacent vertebral bodies and their intervertebral disk) MRI has high diagnostic sensitivity, specificity, and accu-
[110]. Incidence of SD in developed countries ranges from racy (reported as 96%, 92%, and 94%, respectively) for SD
4 to 24 per million per year [111], men being affected more due to its high contrast resolution, direct multiplanar i maging
frequently than women and the most frequently affected capability, and high sensitivity for soft tissue and bone mar-
age being between 50 and 70 years. The most frequent site row abnormalities. MRI is generally positive in the early
of vertebral infection is the lumbar spine (45%) followed stages (first 2 weeks) after the onset of infection in >50% of
by the dorsal (35%) and the cervical tract (20%) [111]. cases, when other radiologic imaging modalities are still nor-
SD can be primary, i.e., when pathogens infect the spine by mal; these features currently make MR the gold standard for
hematogenous spread. Staphylococcus aureus is the most fre- imaging of spinal infections.
quently isolated pathogen (55–80% of cases), while 7–33% of The MRI examination should include the following
pyogenic SDs are caused by Enterobacteriaceae, mostly sequences:
Escherichia coli followed by Proteus spp. and Klebsiella spp.
Coagulase-negative staphylococci (CoNS) account for 5–16% • STIR or fat saturation T2-weigthed, which detect inflam-
of cases, while brucellosis, a common zoonosis in endemic matory edema with high sensitivity
areas (Mediterranean basin, Latin America, the Middle East, • T1-weighted SE fat saturation pre- and post-administration
parts of Africa, and Western Asia), can account for 21–48% of of i.v. contrast media, for evaluation of morphology and
spinal infections. In developed countries, between 9 and 46% vascularization of lesions.
of SDs are due to tuberculosis (TB) infection.
Secondary SD is due to direct contamination of microor- Usually vertebral infection and inflammation cause
ganisms during surgical or interventional procedures. The marrow edema, with ensuing signal reduction in
incidence of secondary SD varies according to the type of sur- T1-weighted sequences and increase in T2-weighted
gical procedure, ranging from 1% to 7% of patients undergo- sequences, associated with contrast enhancement in
ing surgery. Secondary SD is mainly caused by Staphylococcus T1-weighted sequences after i.v. contrast. When erosion of
aureus, isolated in almost 50% of cases and CoNS. Gram- the vertebral endplates occurs, the hypointense band of
cortical bone appears thinned or even absent. The MRI interstitial pressure and thus prevent efficient accumulation
pattern referred to as “Modic change” corresponds to of labeled leukocytes at the site of infection/inflammation
abnormal bone signals under the vertebral endplate sug- within the useful imaging time window. Based on the above
gesting lesions of vertebral endplate as well as of the adja- considerations, it follows that the available radiopharmaceu-
cent bone marrow in the vertebral body. Loss of the ticals to diagnose spine infection are: 99mTc-diphosphonates
internuclear cleft in T2-weighted sequences is an early (99mTc-MDP/HDP), 67Ga-citrate, and [18F]FDG. Each of
sign of discitis. If the infection involves all the vertebral them has pros and cons.
body and also the intervertebral disk and the nearby soft
tissues, T2-weighted sequences show a hyperintense sig- Scintigraphy with 99mTc-MDP/HDP and/or 67Ga-citrate
nal and a moderate and adjacent tissue enhancement in the Bone scintigraphy with 99mTc-MDP/HDP has a sensitivity of
T1-weighted sequences that can be seen post i.v. contrast 81.4% and 40.7% specificity for diagnosing spine infection
injection. Abscesses, seen as fluid collection, show hetero- (see example in Fig. 24.27). This imaging procedure has the
geneous signal in T1-weighted sequences with ring- advantages of low cost, low radiation burden, and single-day
enhancing thickened walls. Granulation tissue causes imaging procedure, whereas low specificity is its main limi-
diffuse enhancement throughout the mass. tation [112].
The main limitations of MRI are related to overestimation 67
Ga-citrate scintigraphy has a sensitivity of 86.3% and
of the extent of infection (as some of the signal changes are 35.8% specificity. 99mTc-diphosphonate bone scintigraphy
reactive) and to false-positive results (e.g., in presence of combined with 67Ga-citrate can be used for increased diag-
severe degenerative disk disease and postsurgery—due to the nostic accuracy, particularly in case of postsurgical infec-
increased T2-weighted signal and contrast enhancement at tions or to complement equivocal MRI findings in primary
the surgery site). Since some MR changes may persist or infections [113–115].
even worsen during treatment (even with clinical improve- Although 67Ga-citrate and 99mTc-MDP are widely avail-
ment and negative lab tests), they may lead to unnecessary able, the 67Ga-citrate scan is time-consuming (with image
surgery. acquisitions up to 48–72 h post-injection) and results in a
CT is readily available, easy to perform, and faster than relatively high radiation burden to patients.
MRI and represents the best modality to detect bony abnor- If the uptake of 67Ga-citrate is higher than 99mTc-MDP/
malities (such as minimal erosion of the endplates—before HDP uptake, the vertebral lesion is most likely due to infec-
they become visible on plain X-ray) and small vertebral foci tion, whereas degenerative changes are most likely when
of infection as well as sequestration or pathological calcifi- 99m
Tc-diphosphonate uptake is higher [113]. The overall
cation suggesting TB. In spine infection CT shows loss of accuracy of combined 99mTc-MDP/67Ga-citrate imaging
bone architecture with areas of transparency and soft tissue ranges between 65% and 80% [113, 114]. Nevertheless, it
replacement, endplate erosion, and collapse of the disk space should be considered that, except circumstances where the
with abscess formation. use of other radionuclide imaging procedures is problematic,
Furthermore, CT is currently used for guiding percutane- the use of 67Ga-citrate has markedly declined both because of
ous needle biopsy and percutaneous drainage of abscesses. a relatively high radiation dose and because of poor quality
Image quality may suffer from artifacts related to implants. images, linked to the suboptimal energies of the gamma
Image-guided biopsy of the infected tissue is indeed impor- emission generated by 67Ga.
tant for diagnosing SD. Histopathological and microbiologi-
cal examinations of biopsy specimens can establish definite PET/CT with [18F]FDG
diagnosis of SD and identify the causative pathogen. A CT or The role of PET (currently used as hybrid PET/CT) with
fluoroscopy-guided biopsy is the first invasive diagnostic [18F]FDG for diagnosing spinal infection has been exten-
step since it identifies the bacterial agent in up to 91% of sively investigated. The degree of [18F]FDG uptake is
patients. Biopsy is the principal investigation in patients with related to the metabolic rate, therefore to the expression of
postoperative spondylodiscitis since in these patients, blood glucose transporters on the cell membrane [116]. Similarly
cultures are frequently negative. as in tumor cells, in activated inflammatory cells (such as
neutrophils, lymphocytes, monocytes, and macrophages),
24.9.2.1 Radionuclide Imaging both the number of glucose transporters and affinity of
of Spondylodiscitis these transporters for glucose (or its analog—deoxyglu-
Scintigraphy with 99mTc-HMPAO-leukocytes has virtually cose) are enhanced [117–119]. These features explain the
no role in the diagnosis of spondylodiscitis. In fact, both the very high sensitivity of PET (especially PET/CT) with
vertebral bodies and the intervertebral disks are spaces [18F]FDG in diagnosing spine infections—which is how-
enclosed in a virtually non-expandable structure; therefore, ever associated with relatively low specificity (see
all pathological events inducing local edema increase the Fig. 24.42).
Fig. 24.42 PET/CT with [18F]FDG in a patient with suspected lumbo- clinical suspicion of spondylodiscitis (modified from: Lazzeri
sacral spondylodiscitis. The sagittal sections (CT, left; PET, middle; E. Nuclear medicine imaging of bone and joint infection. In: Lazzeri E,
fused PET/CT, right) of the lumbosacral skeleton show increased Signore A, Erba PA, Prandini N, Versari A, D’Errico G, Mariani G,
uptake of [18F]FDG (SUVmax 3.6) at the vertebral body of L5 with Eds. Radionuclide Imaging of Infection and Inflammation – A Pictorial
involvement of the disk interposed between L5 and S1, confirming the Case-Based Atlas. Milan: Springer; 2013:39–80)
Ohtori et al. showed that a definitive diagnosis of spine [121–123]. After reporting their experience demonstrating
infection was achieved more often when [18F]FDG was 86% sensitivity and 95% specificity for assessing disease
utilized specifically in patients with spinal infections pre- activity, Hungenbach et al. have proposed some interpreta-
senting as Modic type 1 signal on MRI—a condition tion criteria based on the [18F]FDG uptake patterns, classi-
where distinguishing between common Modic changes fied into different scores (from 0 to 4) related to different
and spine infection is challenging [120]. [18F]FDG PET is states of spine infection [124]:
useful also in the evaluation of patients with metallic
implants, since this imaging modality is not affected by • Score 0: normal findings and physiological [18F]FDG dis-
artifacts. tribution (consistent with no infection)
[18F]FDG PET/CT imaging has an important role espe- • Score 1: slightly enhanced uptake in the inter- or paraver-
cially for the evaluation of response to treatment in patients tebral region (consistent with no infection)
with spine infections; it is therefore recommended to acquire • Score 2: clearly enhanced uptake with a linear or disci-
a scan in baseline conditions as the reference for a repeat form pattern in the intervertebral space (consistent with
scan following (medical) therapy to assess disease activity discitis)
• Score 3: clearly enhanced uptake with a linear or disci- be also fixed by applying a hydroxyapatite-based compound
form pattern in the intervertebral space and involvement to the prosthetic surface, on which new bone tissue is formed.
of ground or cover plate or both plates of the adjacent Acetabular components can be placed under pressure within
vertebrae (consistent with spondylodiscitis) the acetabulum and, if necessary, further assured with ortho-
• Score 4: clearly enhanced uptake with a linear or disci- pedic nails.
form pattern in the intervertebral space and involvement Clinical outcomes of these procedures in terms of pain
of ground or cover plate or both plates of the adjacent relief and restoration of function are usually very satisfac-
vertebrae + surrounding soft tissue abscess (consistent tory. However, complications may arise due to detachments,
with spondylodiscitis) dislocations, fractures, infections, and/or heterotopic ossifi-
cation. These conditions must be diagnosed and adequately
Alternative interpretation criteria proposed by Riccio et al. treated. About 10 years post-implant, 50% of the prostheses
state that [18F]FDG uptake in bone or soft tissue is consistent show clear radiographic signs of mobilization, and about
with active infection, while uptake confined to the edges of a 30% must be reimplanted/replaced. Aseptic mobilization is
destroyed disk, after antibiotic therapy of pyogenic spine infec- often due to an immune/inflammatory reaction that causes
tion, must not be considered indicative of persistent infection the formation of a pseudomembrane similar to synovium
and likely represents mechanically induced inflammation. On (constituted mainly by giant cells), which is located between
the other hand, semiquantitative analysis could not reliably dis- the prosthesis and bone; lymphocytes and plasma cells are
tinguish patients with active infection from those without active present in 25% of cases and neutrophils in <10%. Material
infection and those who had had a successful response to ther- (debris) produced from fragmentation of the prosthetic com-
apy [125]. Whereas, according to some authors, SUVmax reflects ponents attracts and activates tissue phagocytes which, how-
the activity of infectious spondylitis, although this conclusion ever, cannot destroy with their lithic enzymes this material;
is based on limited observations [126]. this process results in the secretion of inflammatory cyto-
The findings at [18F]FDG PET/CT have been reported to kines and proteolytic enzymes (which damage the surround-
complement the MRI findings for distinguishing pyogenic ing bone tissue and cartilage), in turn resulting in the
from tuberculous spondylitis [126]. It should also be noted activation of immune cells, osteolysis, and mobilization of
that the combination of MRI with [18F]FDG PET/CT results in the prosthesis.
detection of spinal infection in 100% of patients [127–129]. The incidence of prosthetic joint infection (PJI) of pri-
On the other hand, it has consistently been shown that mary implants is approximately 2%, increasing to 20% after
[18F]FDG PET/CT is superior to 67Ga-citrate scintigraphy revision surgery [131–134]. Once microorganisms reach the
for detecting spinal infection and paraspinal soft tissue prosthetic region, they can attach to the biomaterials and
infection and could be used in selected cases as an alterna- within 2 days form a biofilm on the implants, which makes
tive to MRI. them hard to detect and to eradicate. Furthermore, osteolysis
In summary, the advantages of [18F]FDG PET/CT are a induced by the inflammatory cells results in loosening of the
very high negative predictive value, high-quality imaging, prosthesis, which can become painful. Due to the often low-
and short time for completing the acquisition. Limitations of grade nature of the infection, it is very difficult to establish
[18F]FDG PET/CT are its relatively low specificity (ranging whether there is a septic or aseptic loosening [133, 134].
from 35.8% to 87.9%) and its relative inability to distinguish The onset of symptoms after implantation can be classi-
infection from neoplastic lesions or pronounced degenera- fied as early, delayed, or late infection. Early infection
tive vertebral disease [130]. (<3 months) is predominantly acquired during implant sur-
gery (or over the following 2–4 days) and is caused by highly
virulent organisms (e.g., Staphylococcus aureus or Gram-
24.9.3 Infection/Inflammation of Prosthetic negative bacilli). Delayed or low-grade infection is predomi-
Joint Implants nantly acquired during implant surgery and caused by less
virulent organisms (3–24 months) (e.g., coagulase-negative
Arthroplasty has revolutionized treatment of patients with staphylococci or Propionibacterium acnes). Late infection
advanced degenerative joint disease, especially of the hip (>24 months) is predominantly caused by hematogenous
and knee. The introduction of modular prostheses made of seeding from remote infections [135]. The inflammatory
metal (cobalt, chrome, or titanium), plastic, or ceramic has reaction induced by infection is similar to aseptic mobiliza-
made it possible to adapt the prosthetic implant to the indi- tion, except for a fundamental feature: neutrophils are invari-
vidual patient’s needs. These components can be assembled ably present. The main inflammation indices (ESR and PCR)
on bone tissue with different techniques: cemented prosthe- can be altered both during aseptic mobilization and during
ses are fixed with polymethylmethacrylate, while non- infection.
cemented prostheses use a porous surface between the Management of patients with aseptic loosening differs
prosthesis itself and adjacent bone tissue. The prosthesis can from that of patients with prosthetic infection. In fact, while in
case of aseptic loosening direct replacement of the prosthesis scintigraphy with 99mTc-diphosphonates. The specificity of
with a new one might be directly performed, in case of infec- bone scintigraphy in hip prosthesis infection ranges from
tion dual-time surgery is required, consisting of removal of the 50% to 70%, and it is related to the time elapsed after the
prosthesis followed by antibiotic therapy and/or antibiotic- implant. The 99mTc-diphosphonate late phase findings alone
loaded spacer and subsequent implant of new prosthesis. cannot distinguish between aseptic loosening and infection.
In fact, periprosthetic increased tracer uptake reflects
increased bone turnover, which can occur in both condi-
24.9.4 Non-radionuclide Imaging tions—infection and inflammation. When bone scintigra-
for Infection/Inflammation phy fails to visualize enhanced periprosthetic bone
of Prosthetic Joint Implants metabolism, it is very unlikely that the patient’s symptoms
are caused by the implant (i.e., the scan has a very high
X-ray constitutes the first-line diagnostic approach in case of negative predictive value). In these cases, referred pain
suspected PJI, even if it is well known that plain radiography from other regions of the body might be the cause of the
may be negative up to 4 weeks after onset of infection and problem, which in some cases can be diagnosed at bone
does not allow early diagnosis. At later stages, X-ray may scintigraphy (e.g., degenerative disease of the spine leading
show loosening signs, such as the presence of a radiolucent to pain projecting to the hip in patients).
area around the prosthesis. Other signs such as erosions, When increased uptake of 99mTc-diphosphonates is
periosteal reaction, and soft tissue swelling can be seen in observed near or around the prosthetic implant, differential
case of PJI. One advantage of plain X-ray is that image qual- diagnoses must be considered. In particular, physiologic
ity is not affected by the presence of metallic implants. bone remodeling may occur up to at least 1 year after implan-
Ultrasonography can be effectively used to detect the tation; this implies that bone scintigraphy performed within
presence of periprosthetic fluid collections and to track the such time window has a high yield of false-positive results,
presence of sinus tracts within soft tissues. The main advan- therefore resulting in quite difficult interpretation. Loosening
tages of ultrasonography are its wide availability, low cost, of the metallic implants is scintigraphically characterized by
and the possibility of being performed at bedside. It can also increased uptake at the bone/metal interface. Inflammation
be repeated whenever needed. Furthermore, it can be used to (not necessarily caused by infection) is defined as very likely
guide interventional procedures, such as needle aspiration. when increased tracer accumulation is observed in all three
CT imaging can be used to evaluate bone abnormalities and phases of a three-phase bone scintigraphy.
loosening of the prosthesis, although the presence of implant- In addition, the presence and pattern of uptake around a
related striking artifacts constitutes an important limitation. prosthetic implant vary depending on the time elapsed since
Nonetheless, recent technical developments, such as specific implantation surgery. In the case of cemented implants, most
algorithms for artifact reduction, partly overcome these limita- asymptomatic patients have a normal bone scan pattern
tions. Other limitations of CT imaging are represented by sub- 1 year after the implant. However, up to 10% of patients
optimal evaluation of periprosthetic soft tissues and by the shows persistently increased uptake even in the absence of
relatively high ionizing radiation burden to patients. complications. For non-cemented prostheses, nonspecific
Prosthetic implants do not constitute an absolute contra- increased uptake persists for even longer periods. In patients
indication to MRI. In fact, specific MR sequences (metal with knee prosthesis, increased tracer uptake may be
artifact reduction sequences, MARS) have been developed observed for more than 1 year after the implant (in 63% of
to overcome these problems. Among the various sequences, cases for the femoral component, in 89% for the tibial
turbo spin-echo and short tau inversion recovery sequences component).
appear to be affected by minor artifacts than gradient-echo- Most of the prosthetic infections occur within the first year
based sequences. MRI is excellent for evaluating involve- after the implant (i.e., when bone scintigraphy is mostly
ment of the skin, subcutaneous or deep soft tissues adjacent affected by the limitations described above), only the pres-
to the prosthesis, as well as the presence of a fistula and soft ence of a normal pattern can provide useful information (i.e.,
tissue edema. MRI is also helpful in characterizing the to exclude the presence of infection and/or inflammation). It
nature of a collection as serous, purulent, or hematic. follows that the overall accuracy of bone scintigraphy for
diagnosing prosthetic joint infection is suboptimal—at
50–70%. In particular, while a negative three-phase bone
24.9.5 Radionuclide Imaging for Infection/ scintigraphy can reasonably rule out prosthetic infection, a
Inflammation of Prosthetic Joint positive bone scintigraphy is generally not sufficiently spe-
Implants cific to diagnose prosthetic infection. In this case scintigraphy
with radiolabeled autologous leukocytes is required to distin-
In patients with low probability of PJI, the radionuclide guish between aseptic prosthetic loosening and PJI (see
imaging procedure of choice is a dynamic three-phase bone Figs. 24.43 and 24.44).
Fig. 24.43 99mTc-HMPAO-

WBC scintigraphy performed
in a patient with bilateral total
hip prosthesis and persisting
pain associated with
functional impairment of right
hip. The planar images
acquired at 3 and 20 h
post-injection (upper panel)
show accumulation of labeled
leukocytes in soft tissues
around the right hip. The
fused SPECT/CT images
shown in lower panel (axial,
left; sagittal, middle; coronal,
right) demonstrate that the
infection of soft tissues of the
hip is in continuity with the
prosthesis, consistent with an
abscess (modified from:
Prandini N, Caruso
G. Nuclear medicine imaging
of joint prosthesis infection.
In: Lazzeri E, Signore A, Erba
PA, Prandini N, Versari A,
D’Errico G, Mariani G, Eds.
Radionuclide Imaging of
Infection and Inflammation –
A Pictorial Case-Based Atlas.
Milan: Springer;
2013:81–106)
Fig. 24.44 Bone
scintigraphy in a patient
previously submitted to right
hip prosthesis. Broken screw
and enhanced tracer uptake in
the upper acetabulum indicate
loosening of a cup of the
endoprosthesis as visualized
on fused SPECT/CT image
(left panel) and on low-dose
CT image (right panel)
In hip prosthesis, extraarticular calcifications may also or above the instrumented region can be the cause of ongo-
be found. Their removal only makes sense when they are no ing pain and is recognized on scintigraphic imaging, prefer-
longer metabolically active. In loosened knee prosthesis, ably performed with SPECT/CT, especially because in
large resorptive cysts that are not hypometabolic and are patients treated with metallic implants planar imaging and
easily identifiable on CT may produce false-positive results. even stand-alone SPECT imaging cannot reliably localize
When lumbar implants do not stabilize the joint, spondylo- the anatomic sites of abnormalities observed on the scan.
sis and facet’s arthritis may progress in the instrumented On the other hand, the changing pattern of tracer uptake
region. Furthermore, spinal degeneration occurring below in the case of knee prosthesis differs somewhat from hip
Fig. 24.45 Scintigraphic
evaluation of a patient submitted
1 year earlier to total left knee
prosthesis, presenting with pain of
left knee, edema, and local signs of
inflammation. Upper panel: “blood
pool” phase (left) and delayed 3-h
(right) acquisitions from three-
phase bone scintigraphy with
99m
Tc-MDP, indicating the presence
of edema and active bone
remodeling. These findings are
consistent with ongoing
inflammation, but not conclusive as
to the presence of infection. Middle
panel: 3- and 20-h acquisitions (left
and right, respectively) during
99m
Tc-HMPAO-WBC scintigraphy,
showing accumulation of labeled
leukocytes at the left knee. Bottom
panel: fused transaxial (left),
sagittal (center), and coronal (right)
SPECT/CT images acquired during
showing involvement 99mTc-
HMPAO-WBC scintigraphy,
showing bone involvement of the
lateral compartment of left knee, of
the lateral portion of the tibial
plateau, of the lateral femoral
condyle, and also in the
infrapatellar bursae (modified from:
Prandini N, Caruso G. Nuclear
medicine imaging of joint
prosthesis infection. In: Lazzeri E,
Signore A, Erba PA, Prandini N,
Versari A, D’Errico G, Mariani G,
Eds. Radionuclide Imaging of
Infection and Inflammation – A
Pictorial Case-Based Atlas. Milan:
Springer; 2013:81–106)
prosthesis, as increased uptake of 99mTc-diphosphonates choice [100]. SPECT/CT acquisition and the use of correct
reflecting bone remodeling can persist much longer than in interpretation criteria allow the differential diagnosis
patients with hip prosthesis—up to several years. Therefore, between infective and inflammatory periprosthetic pro-
dynamic three-phase bone scintigraphy is less useful in these cesses. In case of doubtful/equivocal findings at labeled leu-
patients than in patients with hip prosthesis [136]. In case of kocyte scintigraphy, bone marrow scintigraphy can usefully
clinical high probability of periprosthetic knee infection be added to achieve the correct diagnosis [100], based on the
(Fig. 24.45) or in the evaluation of response to treatment rationale that, while both labeled leukocytes and radiocol-
(antibiotic therapy), scintigraphy with labeled autologous loids accumulate in the healthy bone marrow, only labeled
leukocytes constitutes the radionuclide imaging procedure of leukocytes accumulate in infection.
a b
Fig. 24.46 Patient with right hip arthroplasty, presenting with pain and inguinal lymph nodes, consistent with infection diagnosis of right hip
fistula around the right hip. (a) X-rays show radiolucent lines around prosthesis (modified from: Prandini N, Caruso G. Nuclear medicine
the shaft of the prosthesis consistent with hip loosening. (b) [18F]FDG imaging of joint prosthesis infection. In: Lazzeri E, Signore A, Erba PA,
PET (coronal sections at two different levels) shows irregularly Prandini N, Versari A, D’Errico G, Mariani G, Eds. Radionuclide
increased tracer uptake around both components (acetabulum and Imaging of Infection and Inflammation – A Pictorial Case-Based Atlas.
femur) of the total hip arthroplasty, in continuity with the fistula in lat- Milan: Springer; 2013:81–106)
eral side of the hip. Enhanced tracer uptake can also be seen in the right
Radiolabeled anti-granulocyte monoclonal antibody prepa-

rations can also be used if periprosthetic OM is suspected, with • Scintigraphy with labeled leukocytes or radiola-
the same imaging acquisition modes as described above for beled anti-granulocyte antibodies is useful to
peripheral OM. This imaging approach has 85% sensitivity and exclude or prove the presence of infection.
83% specificity for the diagnosis of PJI [102–104]. The combi- • Spinal infections include vertebral osteomyelitis
nation with bone marrow scintigraphy appears to increase spec- (infection of the vertebral body), discitis (infection
ificity of scintigraphy with labeled anti-granulocyte antibody of the intervertebral disk) and spondylodiscitis (SD)
agents [137]. (infection of two adjacent vertebral bodies and their
Controversial results have been reported on the diagnostic intervertebral disk).
accuracy of [18F]FDG PET/CT for diagnosing PJI • Unlike osteomyelitis at other sites, diagnostic imag-
(Fig. 24.46), and its utility is still debated [107, 138]. In par- ing of spine infection relies mostly on MR and PET/
ticular, while a negative PET/CT scan can certainly rule out CT with [18F]FDG.
the presence of PJI (because of its very high negative predic- • Radionuclide imaging for suspected infection/
tive value), a positive scan does not permit to distinguish inflammation of prostetic joint implants is based on
with certainty between infection and inflammation [139]. specific protocols to discriminate aseptic loosening
from infection.
Key Learning Points

• Different radiopharmaceuticals can be used in the
diagnostic flowchart of infection and/or inflamma- The European Association of Nuclear Medicine (EANM),
tion of peripheral bones, including 99mTc- the European Society of Radiology (ESR), the European
diphosponates, labeled autologous leukocytes, Bone and Joint Infection Society (EBJIS), and the European
anti-granulocytes antibodies, [18F]FDG, and Society of Microbiology and Infectious Diseases (ESCMID)
67
Ga-citrate. have recently released a Consensus Document for the diag-
• A negative bone scan is sufficient to exclude pres- nosis of prosthetic joint infections (see below under “Further
ence of inflammation associated with infection. reading”). According to their conclusions concerning radio-
nuclide imaging, the time elapsed since prosthesis implant
and the beginning of symptoms suggesting prosthetic joint sclerotic phase. In this phase the margins of resorption might
infection is the first discriminant for the choice of the initialfurther expand; coarse trabeculae and thickened cortical
imaging modality. In particular, due to continuing bone bones with a cotton wool appearance are hallmarks of the
remodeling for the majority of the implant sites, more than latter stage.
two years (but up to five years for knee prosthetic implants) Paget’s disease is generally asymptomatic, so often the
must elapse for either three-phase bone scintigraphy or PET/ disease is detected incidentally during radiologic imaging
CT with [18F]FDG to have a high negative predictive value performed for unrelated reasons or because of blood chemis-
for infection and/or inflammation – thus to rule out the diag- try tests revealing increased alkaline phosphatase values. In
nosis of infection. In case either of these two imaging modal- the clinically manifest cases, pain or onset of bone deformity
ity turns out to be positive, the clinical presentation must be are the most frequent manifestations. Pain is due to the
carefully considered to stratify patients with regard to the increased bone remodeling that characterizes the disease or,
suspicion of either acute or chronic infection. When acute more frequently, by indirect complications of the disease
infection is suspected, the first imaging modality of choice is such as degenerative osteoarthritis, nerve compression, or
labeled autologous WBC scintigraphy, the use of labeled osteosarcoma.
anti-granulocyte scintigraphy remaining the second choice. Other causes of pain include increased vascularization,
When instead chronic infection is suspected, anti-granulo- periosteum distortion caused by disorganized bone structure
cyte antibody scintigraphy is the first imaging modality of during remodeling, and the consequences of microtraumatic
choice, autologous WBC scintigraphy remaining the second stress to which the bone is most exposed. Bone hypertrophy
choice. If instead the clinical suspicion of infection arises in the subchondral region can cause osteoarthritis, and this
within two years after prosthetic joint implant, the first imag- phenomenon can make the differential diagnosis between
ing modality of choice is scintigraphy with autologous articular pain caused by Paget’s disease and pain caused by a
WBCs, possibly combined with bone marrow scintigraphy; flogistic degenerative osteoarticular process very difficult.
nonetheless, the use of [18F]FDG PET/CT can also be con- Diagnosis of Paget’s disease is often radiological, and the
sidered – keeping in mind however its low specificity associ- pathognomonic aspect is bone deformation; other radiologi-
ated with possible false-positive findings due to inflammation cal patterns include thickening of cortical bone and the pres-
in the case of aseptic loosening and/or recent surgery. ence of areas of osteolysis with alternating areas of
osteosclerosis. Lytic lesions are clearly evident in the initial
phase and occur as focal lesions (restricted osteoporosis) or
24.10 Paget’s Disease of the Bone appear as a “flame” on plain X-ray. In the later stages, areas
of sclerosis appear and lithic and sclerotic lesions coexist.
Paget’s disease of the bone is a focal, progressive alteration Although plain X-rays may allow a correct diagnosis, it may
of the remodeling processes of the bone, described in 1876 sometimes be difficult to distinguish between pagetic lesions
by James Paget. It is characterized initially by excessive and other skeletal lesions, such as metastases—either lytic or
bone resorption, followed by a secondary increase of osteo- sclerotic.
blast activity with bone reconstitution. The fundamental his- Bone scintigraphy is the most sensitive imaging proce-
tological lesion is an excessive amount of fibrotic connective dure for identifying Paget’s disease, and it is mainly indi-
tissue inside the medullary space, with hyper-vasculariza- cated in the initial diagnostic workup of patients affected by
tion. The prevalence of Paget’s disease is 1.5–3% in subjects this condition. In fact, bone scintigraphy is crucial to define
over the age of 60 years, more common in men than in overall distribution of the disease, to identify the affected
women (with a 1.8:1 ratio). Although the etiology of Paget’s bone segments and the possible development of bone com-
disease is unknown, it is believed that there is some interac- plications. The most affected bone segments belong to the
tion between genetic factors and environmental factors (viral axial skeleton: pelvis (65%), spine (42%), femur (37%), tibia
infections) at the basis of the disease. A family history is (37%), sacrum (30%), and skull (29%). The humerus (13%),
present in about 15% of patients, with a higher risk than in clavicle (5%), and forearm (5%) are less frequently affected.
the general population to develop the disease in the first- The disease may be monostotic (when only one bone seg-
degree relatives. Typically, three phases of disease are ment is involved) or polyostotic (when more bone segments
described: lytic, mixed, and sclerotic phases. The initial lytic are affected), in a continuum spectrum, with 65% probability
phase is characterized by geographic map osteolysis with of identifying new disease localizations during follow-up. To
advancing sharp edge bone resorption. In the long bones, the assist in the diagnosis of Paget’s disease, relatively specific
disease starts in the metaphyseal region and extends to the signs have been described on both X-rays and bone scintig-
diaphyseal region with the appearance described as resem- raphy [140]. On bone scintigraphy, the “picture frame sign”
bling a “blade of grass” or “flame-shaped.” In the skull, the represents cortical thickening associated with radiolucent
initial lytic phase is often referred to as “osteoporosis cir- center of an enlarged, flattened vertebra body. The “Ivory
cumscripta.” The intermediate phase is a mixed lytic- sign” is typical of the sclerotic phase and is characterized by
enlarged and homogenously dense vertebra body. The in patients with recurrent and/or metastatic prostate cancer.
“Mickey Mouse” sign describes the specific increased uptake Several reports describe nonspecific uptake of these two trac-
in the vertebral body, posterior elements, and its spinous pro- ers in Paget’s disease. 18F-fluoride PET/CT has been used to
cess. This same sign has been reported under different names quantify bone turnover in Paget’s disease, particularly for
such as the “T-sign” or “champagne glass” sign [141]. assessing response to treatment with bisphosphonates [150];
Bone scintigraphy can be also used for assessing possible however, this technique is not used routinely in clinical
complications of Paget’s disease, for instance, in the case of practice.
fractures, which are frequently complicated by nonunion.
Insufficiency fractures are usually seen in the cortex of
abnormally bowed long bones and have been defined as
Key Learning Points
“banana fracture” [142]. The altered biomechanics of the
• Bone scintigraphy is the most sensitive imaging
affected joints can lead to cartilaginous and bone abnormali-
procedure for identifying Paget’s disease, and it is
ties resulting in osteoarthritis.
mainly indicated in the initial diagnostic workup of
Malignant transformation of bone affected by Paget’s
patients affected by this condition.
disease can occur, most commonly in patients with wide-
• Bone scintigraphy is useful also to detect possible
spread osseous involvement, however with an incidence
transformation of pagetic lesions into a malignant
<1% [143]; sarcomatous Paget transformation may occur
condition.
simultaneously at different sites. This condition has a very
poor prognosis, yet it is difficult to diagnose, because sev-
eral reports noted unexpectedly low [18F]FDG uptake in
these tumors [144]. Other rarer tumors possibly complicat-
ing Paget’s disease are fibrosarcoma, chondrosarcoma, 24.11 T
reatment of Skeletal Metastases
malignant fibrous histiocytoma, and even giant cell tumor with Bone-Seeking
[145]. Radiopharmaceuticals
Figure 24.47 shows some examples of 99mTc-
diphosphonate bone scans observed in case of mono- and 24.11.1 Generalities on Metastatic
polyostotic Paget’s disease. Disease to the Skeleton
The most common conditions that are considered in the
differential diagnosis with Paget’s disease are metastasis Skeletal metastases constitute one of the most feared events
and fibrous dysplasia. Paget’s disease and fibrous dysplasia in patients with cancer, not only because they generally
are frequently difficult to distinguish on the basis of the indicate advanced disease but also because they translate
bone scan, because both lesions may have intense tracer into markedly reduced quality of life [151]. Some of the
uptake. However, clinical data and other radiologic tech- cancers that more frequently give rise to skeletal metastases
niques can contribute to the correct diagnosis; in particular, are also those with the highest incidence (e.g., prostate,
young age, bowing deformities and ground-glass appear- breast, and lung cancers), whereas other less frequent can-
ance are more typical of fibrous dysplasia. When the head is cers do exhibit a certain propensity to metastasize to the
involved, fibrous dysplasia tends to spare the skull, while skeleton (e.g., thyroid and bladder cancers or melanoma and
Paget’s disease often involves skull bones. Furthermore, myeloma) [152].
MRI and [18F]FDG PET/CT can help distinguish metastatic Except rare instances where tumors directly infiltrate
disease from Paget’s disease. In the lytic phase of Paget’s adjacent bony structures (e.g., in case of an intrapelvic cancer
lesion, the fatty marrow signal is preserved on MR imaging infiltrating the sacrum), most of the skeletal metastases arise
because the destruction of bone is due to resorption and not because tumor cells circulating in the blood home into the
to infiltration as in metastatic disease [146]. However, in the bone marrow. Therefore, skeletal segments that are more fre-
sclerotic phase, this MRI feature cannot help the differential quently involved by metastasis are those with a more impor-
diagnosis, because the signal would be low in all sequences. tant component of active bone marrow—that receive a higher
[18F]FDG PET/CT has some additional value in this differ- fraction of blood flow, such as the axial skeleton (from the
ential diagnosis, since pagetic lesions are often non-[18F] skull to pelvis) and the proximal portions of the humeri and
FDG-avid (or mildly avid when the alkaline phosphatase is femora [153].
markedly elevated), while metastases are usually hypermet- The most frequent complications of skeletal metastases
abolic [147]. are bone pain (initially of moderate intensity—but then pro-
Paget’s disease represents a pitfall also when using new ceeding to excruciating pain), pathologic bone fractures, and
PET tracers, such as 68Ga-PSMA ligand [148] and labeled spinal cord compression. The combined effects of these
choline [149], radiopharmaceuticals that are commonly used events result in loss of mobility, while the most important
a b c
d e f
Fig. 24.47 Different patterns observed at 99mTc-HDP scintigraphy in tibia. (d) Monostotic disease involving a lumbar vertebral body. (e)
patients with Paget’s disease of the bone; planar whole-body scans in Polyostotic disease involving the entire right pelvis and the lumbosacral
the anterior and posterior views are depicted for each patient. (a) passage. (f) Polyostotic disease involving a thoracic vertebral body, the
Predominant involvement of the skull. (b) Polyostotic disease involving left pelvis, and the proximal left tibia (reproduced with permission
the left humerus and two nonadjacent vertebral bodies with the typical from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
“Mickey Mouse” appearance. (c) Monostotic disease affecting the right Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
biochemical derangement associated with skeletal metasta- ing expansion and increased intramedullary pressure, perios-
sis is hypercalcemia and bone marrow aplasia. teal stress, or compression of nerves. Nonetheless, the most
Bone pain is due to the combined effects of several fac- important determinants of bone pain appear to be chemical
tors, some of which are related to anatomic changes caused mediators, such as prostaglandins, interleukins, and other
by the growth of a tumor mass within the bone marrow caus- cytokines inducing inflammation (with recruitment of
inflammatory cells) and stimulation of sensory nerve end- 24.11.2 athophysiologic Bases for
P
ings [154]. the Use of Bone-Seeking
Treatment of skeletal metastasis per se is based on a Radiopharmaceuticals
multifaceted approach that includes effective antitumor
regimens (chemotherapy, hormonal therapy when appropri- Bone-seeking radiopharmaceuticals used for treatment of
ate, other newer forms of therapy), local treatments (exter- bone pain localize at the site(s) of skeletal metastasis based
nal beam radiation therapy and/or surgery in case of highly on mechanisms that are similar to those taking place when
localized disease or of impending fracture in weight-bear- diagnostic radiopharmaceuticals are used for bone scintigra-
ing segments), and different systemic therapies including phy, i.e., 99mTc-diphosphonates for conventional gamma
pain- killing medications, bisphosphonates, and radiola- camera imaging or 18F-fluoride for PET imaging. In particu-
beled agents. The latter category includes both specific lar, the intensity of uptake of these agents is directly corre-
tumor-targeting agents (e.g., 131I-iodide in case of metasta- lated with the degree of osteoblastic reaction to osteolysis
ses from differentiated iodine-avid thyroid cancers, induced by the growth of cancer cells in bone. Therefore, a
131
I-MIBG, or 90Y/177Lu-DOTA-somatostatin analogs in more favorable outcome regarding bone pain palliation can
case of metastases from neuroendocrine tumors) and, more be predicted for those painful skeletal sites that exhibit more
commonly, bone-seeking radiopharmaceuticals that local- intense uptake of the diagnostic agent when performing bone
ize at sites of enhanced turnover of the mineral component scintigraphy as a preliminary to treatment with the therapeu-
of the bone. tic radiopharmaceutical (see Fig. 24.48).
Fig. 24.48 Conventional gamma camera bone scintigraphy obtained MDP. Case in left panel is a 67-year-old patient already treated with surgery
about 3 h post-administration of 99mTc-MDP in two patients with painful and adjuvant chemotherapy because of non-small cell lung cancer. Case in
bone metastases who were subsequently treated with 153Sm-EDTMP (which right panel is a 72-year-old man already treated for prostatic adenocarci-
ensued palliation of bone pain). The areas highlighted with pink color are noma, currently resistant to hormonal therapy and with progressing meta-
the metastatic sites that are also the site of bone pain. Both these patients are static skeletal disease during chemotherapy (reproduced with permission
good candidates for therapy with bone-seeking radiopharmaceuticals, since from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
the painful areas correspond to metastatic sites with avid uptake of 99mTc- Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
The clinical benefits of external beam radiation therapy in 1000
Z coordinate (µm)
patients with skeletal metastases have long been known in
terms of both bone pain palliation and accelerated healing of 500
pathologic fractures [155, 156]. In this form of therapy, ion-
izing radiation is delivered to the metastatic site as a whole 0
volume, including both the mineral component of the bone

-500
and the bone marrow space, where metastatic tumor cells
colonize and grow. Therefore, the efficacy of treatment is
-1000
linked, among other factors, to an important antitumor effect 1000
1000
of radiation which leads to reduced local production of the 0 0
500
inflammatory/algogenic compounds released in the presence -500
-1000 -1000
of tumor cells, such as prostaglandins, interleukins, and Y coordinate (µm) X coordinate (µm)
other cytokines.
The mechanism of bone pain palliation induced by bone- Fig. 24.49 Simulation of the tracks in water of β− particles emitted by
seeking radiopharmaceuticals is somewhat similar to that fol- a point source of 131I (full-energy spectrum); simulation has been per-
formed for only 100 particles, a number too limited to show the whole
lowing external beam radiation therapy regarding the local
spectrum of possibilities (including the maximum range, which is
production of inflammatory/algogenic compounds [157]; reached only by an extremely small fraction of the particles emitted).
instead, the mode of energy release by ionizing radiation fol- Due to interactions with surrounding matter, the β− particles deviate
lowing administration of therapeutic bone-seeking radiophar- repeatedly from their initial direction. Most of the energy (about 90%)
associated with the emissions is deposited within about 20% of the
maceuticals is somewhat different from external beam
maximum path range reported for each radionuclide. The maximum
radiation therapy regarding both the distribution of energy in energy of the β− particles emitted by 131I (0.606 MeV) is somewhat
the target lesion and time-related release of energy. Bone- intermediate between those of the β− particles emitted by 89Sr
seeking radiopharmaceuticals localize at the surface of the (0.585 MeV) and by 32P (0.696 MeV), respectively, used for treatment
of painful skeletal metastases (image provided by courtesy of Prof.
trabecular bone (the osteoid layer surface) undergoing active
Alberto Del Guerra, Department of Physics “E. Fermi,” University of
remodeling, and the ionizing particles emitted during their Pisa and National Institute of Nuclear Physics (INFN), Pisa Section,
decay travel a certain distance depending on their physical Pisa (Italy))
properties; in particular, β− particles travel in water/tissue lon-
ger distances than α++ particles (2–8 mm depending on maxi-
mum energy versus 0.1 mm, respectively); therefore their with dose rates that vary according to physical decay of the
linear energy transfer (LET) is lower than that of α++ particles. radionuclide and are in general lower than those involved in
Since electrically charged particles are emitted and travel in a external beam radiation therapy.
3D space (see Fig. 24.49), part of their energy is absorbed by
the trabecular structure of the bone, while a certain fraction of
the energy is absorbed by cells in the intertrabecular space, 24.11.3 Bone-Seeking
i.e., the bone marrow—which includes the normal hemato- Radiopharmaceuticals for Therapy
poietic cells, the tumor cells, and the inflammatory cells.
These events translate into some radiobiological effects not According to the main localization mechanism, bone-seeking
only on the mineral component but also on the cellular com- radiopharmaceuticals used for therapy can be classified into
ponent of the affected bone, thus exerting some antitumor (1) those that localize at sites of enhanced bone mineral turn-
activity and at the same time some myelotoxicity—which is over because of their affinity/chemisorption onto the newly
indeed the main dose-limiting factor for therapy with bone- formed hydroxyapatite crystals and (2) those that localize at
seeking radiopharmaceuticals emitting β− particles. Instead, sites of enhanced bone mineral turnover because they behave
the much shorter tissue penetration of the α++ particles emit- in vivo as chemical analogs of calcium. As described in
ted by 223Ra (the only α-emitting radionuclide currently details in Chap. 4, the first group includes emitters of β− par-
approved for treatment of skeletal metastases) limits bone ticles such as bisphosphonates labeled with different radio-
marrow toxicity following treatment with this agent. nuclides, most notably 153Sm-EDTMP and 186Re-HEDP or
Concerning time-related issues of therapy with bone- 188
Re-HEDP (and more recently 177Lu-EDTMP). The second
seeking radiopharmaceuticals versus external beam radiation group includes 89Sr-chloride (a β− emitter) and
therapy, it must be considered that the latter is administered 223
Ra-dichloride (an α++ emitter). On the other hand,
typically as treatments with high-dose rate (either with the 32
P-phosphate (a β− emitter still used clinically in some low-
conventional multiple fraction protocols or with the more income countries because of low cost—despite a non-negli-
recent hypofractionated protocols), whereas therapy with gible bone marrow toxicity) distributes widely throughout
bone-seeking radiopharmaceuticals is characterized by the bone mineral component because of the natural distribu-
release of ionizing energy over an extended period, therefore tion of phosphate.
Most of the clinical experience with the use of this cate- 24.11.4.2 Contraindications
gory of bone-seeking radiopharmaceuticals has been The only two absolute contraindications for therapy with β−-
acquired with 89Sr-chloride and with 153Sm-EDTMP. Therefore, emitting bone-seeking radiopharmaceuticals are pregnancy
most of the comprehensive analyses of the published litera- and breastfeeding.
ture concerning β− emitters (see further below) refer to these There are several relative contraindications that require
two therapeutic agents. tailoring of therapy after careful consideration of the whole
clinical scenario of each individual patient. They include
conditions related to compromised bone marrow function
24.11.4 linical Use of β− Particle-Emitting
C (due to either diffuse infiltration of the bone marrow by met-
Bone-Seeking astatic cancer cells or prior chemo- or radiotherapy regi-
Radiopharmaceuticals mens) and life expectancy.
Concerning bone marrow function, a certain reduction of
There has been a constant update of guidelines developed by blood cell counts (usually affecting platelets and/or leuko-
professional societies throughout the years for the use of cytes) is expected following therapy with β− emitting bone-
bone-seeking radiopharmaceuticals emitting β− particles in seeking radiopharmaceuticals. Such reductions may reach
the management of patients with skeletal metastases [158– the levels of hematological toxicity in patients with evidence
160]. In this chapter we will refer to the most recently of compromised bone marrow function prior to treatment.
updated guidelines on the use of β−-emitting radiopharma- Although the scenario has changed after the introduction of
ceuticals, developed by the European Association of Nuclear colony-stimulating factors to support/stimulate the produc-
Medicine [160]. Although these guidelines formally regard tion of granulocytes in particular, the following blood cell
the three approved β−-emitting radiopharmaceuticals count values have been defined as the threshold above which
(32P-sodium phosphate, 89Sr-chloride, 153Sm-EDTMP), simi- administration of β−-emitting bone-seeking radiopharmaceu-
lar considerations apply to the use of other bone-seeking ticals for therapy of skeletal metastases is expected to be
radiopharmaceuticals that have been clinically validated and safe:
are available in certain institutions under approval by the
local Institutional Review Board (e.g., 186Re-HEDP, • Hemoglobin >90 g/L
188
Re-HEDP, 177Lu-EDTMP). • Total leukocyte count >3.5 × 109/L
• Platelet count >100 × 109/L
24.11.4.1 Indications
As simply stated by Handkiewicz-Junak et al. [160], the Stability of blood cell counts should be ascertained prior
chief indication for therapy of skeletal metastases with this to treatment, in order to rule out the occurrence of deteriorat-
class of radiopharmaceuticals is for “Painful metastatic bone ing bone marrow function before administering therapy with
lesions with osteoblastic response, as indicated by areas of β−-emitting bone-seeking radio-pharmaceuticals.
intense uptake on radionuclide bone scans.” Furthermore, Furthermore, since the main route of clearance for these
clinical experience has cumulated on the beneficial effect of radiopharmaceuticals is through renal excretion, poor renal
this therapy also in patients with primary painful bone malig- function results in prolonged blood retention and therefore
nancies with similarly enhanced uptake on the radionuclide possibly increasing bone marrow toxicity. As a consequence,
bone scan, although not at sufficient level for the proper treatment with β−-emitting bone-seeking radiopharmaceuti-
authorization bodies to approve this indication for routine cals should not be performed in patients with serum creati-
clinical use. nine >180 μmol/L (or >2.05 mg/dL) and/or estimated
Moreover, although not mentioned in the published glomerular filtration rate <30 mL/min.
guidelines the role of high-dose 153Sm-EDTMP in patients Moreover, the coagulation profile should be investigated
with multiple myeloma has been investigated. The results prior to treatment in order to rule out subclinical dissemi-
obtained indicate that there are no additional hematologi- nated intravascular coagulation, a condition that constitutes a
cal or non-hematological toxicities following this treat- risk factor for severe thrombocytopenia following therapy
ment in combination with other therapies—even when with these radiolabeled agents.
administered with multiple cycles [161, 162]. Furthermore, Concerning life expectancy, it must be considered that the
clinical benefit has been demonstrated following adminis- palliative effect of this form of therapy typically takes place
tration of 153Sm-EDTMP combined with bortezomib (a up to 4 weeks after administration. Therefore, therapy with
proteasome inhibitor, used in this regimen as a radiosensi- bone-seeking radiopharmaceuticals is expected to be more
tizing agent) in patients with relapsed or refractory multi- beneficial in patients with longer life expectancy as well as in
ple myeloma [163]. patients with earlier stages of skeletal metastatic disease.
A final consideration concerning selection of patients 100

with bone pain from skeletal metastases for palliative ther- 90
apy with bone-seeking radiopharmaceuticals has to do with Strontium-89
80
Percent fraction of energy

special features of the metastatic lesions, as it may occur Samarium-153
70
when such lesions cause spinal cord compression or imply
the risk of impending fracture. In these cases the critical 60
lesions must be treated with modalities that ensue faster con- 50
trol of the lesions, such as surgery or external beam radiation 40
therapy. This approach does not preclude concomitant treat- 30
ment of other noncritical lesions with administration of
20
bone-seeking radiopharmaceuticals.
10
24.11.4.3 Efficacy 0
Several meta-analyses and systematic reviews have evaluated 0 10 20 30
the results of a vast number of clinical studies that have been Time elapsed, days
published on this matter [164–173]. Such results consistently
Fig. 24.50 Comparison of the time-related pattern of energy release
demonstrate the beneficial effects of therapy with bone-seek- over a certain period between 153Sm and 89Sr. The plot has been drawn
ing agents in patients with osteoblastic or mixed-pattern skel- in such a way that the areas (or integrals) under the two curves (i.e., the
etal metastases. Since predominantly osteoblastic metastases total energy released from time zero to total decay) are identical for the
are observed primarily in patients with prostate cancer, most two radionuclides. Thus, the two curves basically correspond to the
dose rates for the two radionuclides. The plot shows that, with the same
of the studies have focused on these patients; nonetheless, this total radiation dose released, the dose rate of 153Sm (physical half-life
therapy is equally effective in patients with skeletal metasta- 1.9 days) is much higher than that of 89Sr (physical half-life 50.5 days),
ses originating from other cancers (e.g., breast cancer, lung but it declines to virtually zero over approximately 20 days. Whereas,
cancer, etc.), provided that patients are adequately selected on
89
Sr continues to release radiation energy (though lower than that of
153
Sm) over a much longer period (reproduced with permission from:
the basis of bone scintigraphy demonstrating increased tracer Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
uptake at the sites of painful bone metastases. Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
Overall response rates for palliation of bone pain average
70% of the patients treated, with reports up to 90%, are
reported; these figures are remarkably consistent regardless Studies on the quality of life following treatment of pain-
of the radiopharmaceutical used—153Sm-EDTMP or ful skeletal metastases with bone-seeking radiopharmaceuti-
89
Sr-chloride. In about one out of the three patients respond- cals indicate that these treatments can be recommended to
ing to treatment, there is complete relief of bone pain, while improve the quality of life of patients with skeletal metasta-
in the remaining patients there is partial response, i.e., ses characterized by high tracer uptake on conventional bone
marked reduction of bone pain which becomes therefore scintigraphy.
amenable to treatment with minor analgesic drugs rather Besides symptomatic relief of bone pain, treatment with β−-
than opioid medications. Similar results have been reported emitting bone-seeking radiopharmaceuticals has been shown to
regarding bone pain palliation following administration of induce also objective tumor responses, as demonstrated by, e.g.,
188
Re-HEDP, with an average 32% of complete responses reduction in the serum levels of tumor- associated antigens,
and 44% of partial responses. improved imaging patterns (see Fig. 24.51), and/or reduction in
The onset of bone pain palliation varies between about 1 new skeletal-related serious events. Furthermore, several reports
and 4 weeks post-administration, the shorter interval occur- have shown improved survival after treatment with these agents,
ring when radionuclides with shorter physical half-life are either alone or especially in combination with chemotherapy in
used (i.e., 153Sm and 186Re, 1.9 and 3.8 days, respectively) patients with cancers of either the prostate or breast.
versus 89Sr (50.5 days). This difference is due to the higher
dose rate for shorter-lived radionuclides compared to longer- 24.11.4.4 Administration Protocols
lived radionuclides (see Fig. 24.50). Duration of pain pallia- and Possible Side Effects
tion is generally reported in the 6–15-month range. In case of The standard administration route for 32P-phosphate,
recurrence of bone pain, treatment can be repeated, and effi- 89
Sr-chloride, 153Sm-EDTMP, and 186Re- or 188Re-HEDP is
cacy of such therapy is maintained throughout several cycles, by intravenous injection; slow infusion is recommended,
provided that all indications for therapy with β−-emitting especially when administering 89Sr-chloride. Nonetheless,
bone-seeking radiopharmaceuticals are met—with special 32
P-phosphate can also be administered orally—but in this
attention to the pattern of bone scintigraphy and to the crite- case the activity to be administered is about twofold the
ria of bone marrow function [174]. activity administered intravenously. Particular attention
Fig. 24.51 Efficacy of treatment with 153Sm-EDTMP administered obtained about 6 months posttreatment, without any intervening che-
for bone pain palliation purposes to a patient with progressing motherapy; at this time serum PSA was 45 ng/mL, and the bone scan
castration-resistant prostate cancer. Left panel shows the baseline shows marked improvement of the scintigraphic pattern (reproduced
bone scintigraphy (prior to therapy with the bone-seeking radiophar- with permission from: Volterrani D, Erba PA, Mariani G, Eds.
maceutical) obtained about 3 h after injection of 99mTc-HDP: at this Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
time serum PSA was 540 ng/mL. Right panel shows the bone scan Springer Italy; 2010)
must be paid to avoid extravasation of the β−-emitting radio- • 32

P-phosphate: 185–370 MBq (or 5–10 mCi) when admin-
pharmaceutical, because the high local concentration of the istered intravenously (or about twofold this activity when
therapeutic agent in interstitial space may cause tissue administered orally)
necrosis. • 89
Sr-chloride: 1.5–2.2 MBq/kg body weight (or 40–60 μCi/
No special preparation of the patient is required before kg)
administration of the radiopharmaceutical, including recent, • 153
Sm-EDTMP: 37 MBq/kg (or 1 mCi/kg)
prior, or subsequent therapy with latest generation bisphos- • 186
Re-HEDP: 1110–3515 MBq (or 30–95 mCi)
phonates (such as zoledronic acid), considering that no com- • 188
Re-HEDP: 2500–5500 MBq (or 67–148 mCi)
petition was found for bone localization between such
bisphosphonates and 153Sm-EDTMP or 89Sr-chloride. Although distribution of these therapeutic radiopharma-
Recommended activities of the β−-emitting bone-seeking ceuticals is in general expected to mirror the pattern observed
radiopharmaceuticals administered for palliation of bone on conventional bone scintigraphy with any 99mTc-labeled
pain from osteoblastic or mixed-pattern skeletal metastases bone-seeking phosphonate agent, it may be useful to acquire
are listed here below, as derived from either the EANM scintigraphic images after administration of the β−-emitting
guidelines or from other sources: radiopharmaceutical, to depict in the individual patient actual
Fig. 24.52 Whole-body imaging: 99mTc-MDP bone scan and Pandit-Taskar N, Divgi CR. Targeted radionuclide therapy for bone
153
Sm-EDTMP scan. Bone scan (left) shows multiple areas of abnor- metastases. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
mally increased uptake in the bones consistent with metastatic disease. Nuclear Oncology: From Pathophysiology to Clinical Applications.
Targeting of all sites by 153Sm-EDTMP is seen on the whole-body scan New York: Springer; 2017:1307–1336)
obtained 3 h post-therapy (right) (reproduced with permission from:
distribution of the therapeutic agent throughout the skeleton. effect (occurring within approximately 72 h post-adminis-
When 89Sr-chloride is administered, gamma camera imaging tration and affecting about 10% of the patients treated) is
can be recorded by taking advantage of the photon emission the so-called “flare” phenomenon, i.e., a transitory self-
associated with the bremsstrahlung effect caused by interac- limiting and readily managed increase in bone pain which
tion with tissues of the high-energy β− particles emitted by is in the majority of cases associated with good clinical
89
Sr. However, the images so obtained are in general of rather response.
poor quality. Instead, good-quality gamma camera images The second side effect, which is consistently observed in
are obtained after administration of radiopharmaceuticals the majority of patients treated with β−-emitting bone-
labeled with either 153Sm, 188Re, 186Re, or 177Lu, since the seeking radiopharmaceuticals, is some degree of myelotox-
modes of decays of these radionuclides include the emission icity, which is in general mild and recovers spontaneously
of γ-rays with energies suitable for gamma camera imaging. without requiring special additional therapies. In particular,
Imaging can be acquired either few hours post-administration as a result of irradiation of the hemopoietic cells in the bone
(Fig. 24.52) or, for better scintigraphic contrast, 24 h marrow by the β− particles emitted by the radiopharmaceuti-
post-administration. cal adsorbed on the surfaces of trabecular bone, there is drop
Therapy with β−-emitting bone-seeking radiopharma- in leukocyte and especially in platelet counts, with a nadir
ceuticals implies some side effects that are in general mild that is observed at 3–5 weeks post-administration when
and exhibit a rather predictable pattern. The earliest side using the agents labeled with the shorter-lived radionuclides
(153Sm, 186Re, or 188Re); recovery is usually complete within metastases” [178]. Thus, it should be noted that in the clini-
3 months. When using instead the longer-lived 89Sr or 32P, the cal routine skeletal metastases originating from tumors other
nadir in leukocyte/platelet counts can occur at 12–16 weeks than prostate cancer are not amenable to treatment with
post-administration. Myelotoxicity is usually more 223
Ra-dichloride.
pronounced when patients have received prior myelosup- Although not specifically mentioned in the approval
pressive chemotherapy regimens. document, it is common practice to verify the presence of
predominantly osteoblastic metastases with either a conven-
tional gamma camera bone scintigraphy or with 18F-fluoride
24.11.5 Clinical Use of the α++ Emitter PET/CT.
223
Ra-dichloride Furthermore, the term “visceral metastases” does not
apply to the presence of lymph node metastases as a criterion
This radiopharmaceutical constitutes the most recent addition to for exclusion from therapy with 223Ra-dichloride.
the list of approved bone-seeking agents for therapy of patients
with skeletal metastases. It should be noted that, while other 24.11.5.2 Contraindications
radiopharmaceuticals have been approved with mention of the Considering that prostate cancer in children and adolescents
palliative effects of such treatment, FDA and EMA approvals of under the age of 18 years is an exceptionally rare occurrence,
223
Ra-dichloride read as follows: “For the treatment of patients no safety and efficacy of treatment with 223Ra-dichloride has
with castration-resistant prostate cancer, symptomatic bone ever been investigated in this age group. There are no abso-
metastases and no known visceral metastatic disease.” This defi- lute contraindications to the use of 223Ra-dichloride, pro-
nition, which is based on the results of a large-scale randomized vided that the indications are correctly adopted, chiefly the
seminal study by Parker et al. [175] subsequently supported by absence of visceral metastases.
additional reports [176, 177], implies not only a mere palliation Although hematological toxicity following the use of
effect on bone pain but also an additional clinical benefit in 223
Ra-dichloride is lower than with the β−-emitting agents,
terms of new serious skeletal-related events and overall survival the bone marrow reserve must be carefully investigated
following therapy with 223Ra-dichloride. before administration. Before the first administration/cycle,
In this chapter we will refer to the recently published the minimum threshold values for treatment are listed as
guidelines on the use of 223Ra-dichloride, developed by the follows:
European Association of Nuclear Medicine [178].
Similarly as other bone-seeking radiopharmaceuticals • Hemoglobin >10.0 g/dL
utilized for either diagnostic or therapeutic purposes, 223Ra • Absolute neutrophil count >1.5 × 109/L
(a member of the second main group of elements, which • Platelet count >100 × 109/L
possess chemical and physiological similarities with cal-
cium) selectively accumulates in the bone, particularly in For all subsequent administrations/cycles, the following
areas of high bone turnover—as typically are the border minimum threshold values for treatment are required:
zones of bone and bone metastases. 223Ra decays with
physical half-life of 11.43 days, emitting α++ particles that • Absolute neutrophil count >1.0 × 109/L
have a very limited penetration in tissue (less than 100 μm, • Platelet count >50 × 109/L
which corresponds to about 10 cell diameters). This results
in a very high linear energy transfer and therefore in high Regarding the overall clinical condition of patients who
probability of inducing lethal DNA damage (double-strand are candidates for treatment with 223Ra-dichloride in an out-
breaks) in the adjacent cells that are mostly represented by patient setting, their Karnofsky score should be >50% or
osteoblasts and osteoclasts in addition to metastatic tumor their ECOG performance status <2. If patients do not meet
cells. On the other hand, the limited tissue penetration of these criteria, hospitalization should be considered.
the α++ particles minimizes irradiation of the normal hemo- Furthermore, fecal incontinence requires hospitalization,
poietic components of the bone marrow, thus resulting in since 223Ra undergoes fecal excretion (13% of the injected
hematological toxicity of therapy with 223Ra-dichloride activity within 48 h post-administration versus 2% for renal
considerably lower than that observed following therapy excretion) and incontinence would imply an increased risk of
with the β−-emitting bone-seeking radiopharmaceuticals. contamination for the attending caregivers/household and for
the environment.
24.11.5.1 Indications
The approved indication for 223Ra-dichloride is “… the treat- 24.11.5.3 Efficacy
ment of adult patients with castration-resistant prostate can- Treatment with 223Ra-dichloride is generally followed by
cer, symptomatic bone metastases and no known visceral rapid and marked reduction of bone pain, although pain
relief alone is not a reliable indicator of response to treat- The risk of hematological adverse events (e.g., neutrope-
ment in terms of survival benefit [175, 179, 180]. Objective nia and thrombocytopenia) is increased in patients whose
markers of response to treatment are represented by the bone marrow reserve is reduced because of prior cytotoxic
serum levels of PSA and alkaline phosphatase, although fluc- chemotherapies and/or radiation therapy or in patients with
tuations in serum PSA levels may be observed throughout advanced diffuse metastatic infiltration of the bone—appear-
the six therapy cycles and make this marker a somewhat ing as “superscan” on conventional bone scintigraphy.
unreliable predictor of long-term response. The most frequent adverse side effects observed after
223
Ra-dichloride is the first bone-seeking radiopharma- treatment with 223Ra-dichloride (occurring in ≥10% of the
ceutical that has shown a definite survival benefit in the patients treated) are diarrhea, nausea, vomiting, and throm-
patients treated with this radiopharmaceutical as a single bocytopenia, the most serious adverse reactions being repre-
therapeutic agent, with median overall survival of sented by thrombocytopenia and neutropenia.
14.9 months versus 11.3 months for the group treated with Since 223Ra is excreted mainly through the gastrointestinal
placebo. Investigations of the possible synergistic effect of route, the intestine may be exposed to radiation exposure
223
Ra-dichloride used concomitantly with chemotherapy are sufficient to cause detrimental effects. Therefore, the possible
currently under way. risks caused by administration of 223Ra-dichloride in patients
with inflammatory bowel disease should carefully be consid-
24.11.5.4 A dministration Protocol and Possible ered in terms of benefit/risk ratio.
Side Effects A few patients treated concomitantly with 223Ra-dichloride
According to the EANM guidelines (based on key consensus and bisphosphonates combined with chemotherapy have
recommendations developed among 11 nuclear medicine developed osteonecrosis of the jaw. However, it is currently
centers across six European countries [181]), fasting is not unclear whether this occurrence was attributable to prior
required prior to administration of 223Ra-dichloride, and treatment with bisphosphonates and chemotherapy or
patients should be well hydrated. Although in general whether therapy with 223Ra-dichloride further increased the
patients should take their medications as usual, the only rec- risk of osteonecrosis of the jaw.
ommended precaution is to temporarily discontinue supple-
mentation with calcium, phosphates, or vitamin D for at least
4 days before and after each administration of the radiophar-
maceutical. A careful benefit/risk evaluation should be per-
formed in case patients are receiving concomitant Key Learning Points
chemotherapy. • The skeleton is the most frequent site of metastases
Patients with urinary or fecal incontinence should be han- arising from cancers originating at other sites.
dled in such a way so as to minimize the risk of contamina- • Metastatic tumor cells home in the bone marrow
tion, through bladder catheterization or with other padding mostly by the hematogenous route.
devices, respectively. Although therapy with 223Ra-dichloride • The complications of skeletal metastases include
is generally performed on an outpatient basis, hospitalization bone pain, pathologic bone fractures, spinal cord
should be considered for seriously ill patients of for patients compression, hypercalcemia, and bone marrow
with fecal incontinence. aplasia.
223
Ra-dichloride is administered by slow intravenous • Bone pain is due mainly to local production of
injection (over approximately 1 min), paying special atten- chemical mediators produced in the bone marrow
tion to avoid extravasation. The standard protocol for this environment both by the tumor cells and by inflam-
treatment is based on administration of 223Ra-dichloride in mavtory cells.
the amount of 55 kBq/kg of body weight, corresponding to • Treatment of skeletal metastases includes systemic
3.85 MBq (or 104 μCi) for a 70 kg patient. At variance antitumor regimens, local treatment (radiation therapy
with the β−-emitting bone-seeking radiopharmaceuticals and/or surgery), pain-killing medications, bisphospho-
described in the prior section (for which the rule is a single nates, and bone-seeking radiopharmaceuticals.
administration—considering repeat administration after sev- • The mechanism of bone pain palliation induced by
eral weeks or months in case of recurrence of bone pain), the bone-seeking radiopharmaceuticals is mostly due to
standard protocol for therapy with 223Ra-dichloride involves reduced local production of inflammatory/algo-
six administrations at 4-week intervals. Before each admin- genic compounds.
istration, the hematological profile of patients must be care- • The radiobiological effects caused by bone-seeking
fully evaluated in order to verify that the requisites for radiopharmaceuticals affect not only the mineral
treatment are maintained throughout.
7. Segall G, Delbeke D, Stabin MG, Even-Sapir E, Fair J, Sajdak

R, et al. SNM practice guideline for sodium 18F-fluoride PET/CT
component but also the cellular component of the
bone scans 1.0. J Nucl Med. 2010;51:1813–20.
affected bone, thus exerting some antitumor activity 8. Blake GM, Siddique M, Puri T, Frost ML, Moore AE, Cook
and at the same time some myelotoxicity. GJ, et al. A semipopulation input function for quantifying static
• The basic indication for therapy with bone-seeking and dynamic 18F-fluoride PET scans. Nucl Med Commun.
2012;33(8):881–8.
radiopharmaceuticals is that the lesions to be treated 9. Blake GM, Siddique M, Frost ML, Moore AE, Fogelman I. Imaging
show a predominant osteoblastic pattern on conven- of site specific bone turnover in osteoporosis using positron emis-
tional bone scintigraphy with 99mTc-phosphonates. sion tomography. Curr Osteoporos Rep. 2014;12(4):475–85.
• Bone-seeking radiopharmaceuticals utilized for 10. Hawkins RA, Choi Y, Huang SC, Hoh CK, Dahlbom M, Schiepers
C, et al. Evaluation of the skeletal kinetics of fluorine-18-fluoride
therapy include agents labeled with or carrying ion with PET. J Nucl Med. 1992;33(5):633–42.
radionuclides emitting β− particles (32P, 89Sr, 153Sm, 11. Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation
18
6Re, 188Re, 177Lu) or α++ particles (223Ra). of blood-to-brain transfer constants from multiple-time uptake
• While indications for the approved β−-emitting data. J Cereb Blood Flow Metab. 1983;3(1):1–7.
12. Pfannkuchen N, Meckel M, Bergmann R, Bachmann M, Bal
radiopharmaceuticals (32P-phosphate, 89Sr-chloride, C, Sathekge M, et al. Novel radiolabeled bisphosphonates
153
Sm-EDTMP) state their clinical use for palliation for PET diagnosis and endoradiotherapy of bone metasta-
of bone pain, the indications for the use of ses. Pharmaceuticals (Basel). 2017;10(2). pii: E45. https://doi.
223
Ra-dichloride state its use for therapy of skeletal org/10.3390/ph10020045.
13. Holub J, Meckel M, Kubicek V, Rosch F, Hermann P. Gallium(III)
metastatic disease from prostate cancer with the complexes of NOTA-bis (phosphonate) conjugates as PET
purpose of prolonging overall survival. radiotracers for bone imaging. Contrast Media Mol Imaging.
• All bone-seeking radiopharmaceuticals are effec- 2015;10(2):122–34.
tive not only in terms of bone pain palliation but 14. Bastiaannet E, Groen H, Jager PL, Cobben DC, van der Graaf
WT, Vaalburg W, et al. The value of FDG-PET in the detection,
also in terms of objective tumor responses to grading and response to therapy of soft tissue and bone sarco-
treatment. mas; a systematic review and meta-analysis. Cancer Treat Rev.
• There are specific indications, contraindications, 2004;30(1):83–101.
possible adverse side effects, and expected clinical 15. Biermann S, Chow W, Reed DR, Lucas D, Adkins DR, Agulnik
M, et al. NCCN Guidelines® insights – bone cancer, Version
outcomes for each category of bone-seeking radio-
2.2017. Featured updates to the NCCN guidelines. J Natl Compr
pharmaceuticals used for therapy of metastatic skel- Cancer Netw. 2017;15(2):155–67.
etal disease. 16. Charest M, Hickeson M, Lisbona R, Novales-Diaz JA, Derbekyan
V, Turcotte RE. FDG PET/CT imaging in primary osseous and
soft tissue sarcomas: a retrospective review of 212 cases. Eur J
Nucl Med Mol Imaging. 2009;36(12):1944–51.
17. Fuglo HM, Jorgensen SM, Loft A, Hovgaard D, Petersen
References MM. The diagnostic and prognostic value of 18F-FDG PET/CT
in the initial assessment of high-grade bone and soft tissue sar-
1. Van den Wyngaert T, Strobel K, Kampen WU, Kuwert T, van coma. A retrospective study of 89 patients. Eur J Nucl Med Mol
der Bruggen W, Mohan HK, et al. The EANM practice guide- Imaging. 2012;39(9):1416–24.
lines for bone scintigraphy. Eur J Nucl Med Mol Imaging. 18. Mirabello L, Troisi RJ, Savage SA. Osteosarcoma inci-
2016;43(9):1723–38. dence and survival rates from 1973 to 2004: data from the
2. Treves ST, Lassmann M, EANM/SNMMI Pediatric Dosage Surveillance, Epidemiology, and End Results Program. Cancer.
Harmonization Working Group. International guidelines for pedi- 2009;115(7):1531–43.
atric radiopharmaceutical administered activities. J Nucl Med. 19. Peller PJ. Role of positron emission tomography/computed
2014;55(6):869–70. tomography in bone malignancies. Radiol Clin North Am.
3. Ayres KL, Spottswood SE, Delbeke D, Price R, Hodges PK, Wang 2013;51(5):845–64.
L, et al. Dose optimization of the administered activity in pediatric 20. Quartuccio N, Treglia G, Salsano M, Mattoli MV, Muoio B,
bone scintigraphy: validation of the North American consensus Piccardo A, et al. The role of Fluorine-18-Fluorodeoxyglucose
guidelines. J Nucl Med. 2015;56(9):1391–4. positron emission tomography in staging and restaging of patients
4. Cachovan M, Vija AH, Hornegger J, Kuwert T. Quantification with osteosarcoma. Radiol Oncol. 2013;47(2):97–102.
of 99mTc-DPD concentration in the lumbar spine with 21. Schulte M, Brecht-Krauss D, Heymer B, Guhlmann A, Hartwig
SPECT/CT. EJNMMI Res. 2013;3(1):45. https://doi. E, Sarkar MR, et al. Grading of tumors and tumor-like lesions
org/10.1186/2191-219X-3-45. of bone: evaluation by FDG PET. J Nucl Med. 2000;41(10):
5. Beck M, Sanders JC, Ritt P, Reinfelder J, Kuwert T. Longitudinal 1695–701.
analysis of bone metabolism using SPECT/CT and 99mTc- 22. Tateishi U, Yamaguchi U, Seki K, Terauchi T, Arai Y, Kim
diphosphono-propanedicarboxylic acid: comparison of visual and EE. Bone and soft-tissue sarcoma: preoperative staging with fluo-
quantitative analysis. EJNMMI Res. 2016;6(1):60. https://doi. rine 18 fluorodeoxyglucose PET/CT and conventional imaging.
org/10.1186/s13550-016-0217-4. Radiology. 2007;245(3):839–47.
6. Beheshti M, Mottaghy FM, Payche F, Behrendt FFF, Van den 23. London K, Stege C, Cross S, Onikul E, Graf N, Kaspers G,
Wyngaert T, Fogelman I, et al. 18F-NaF PET/CT: EANM proce- et al. 18F-FDG PET/CT compared to conventional imaging
dure guidelines for bone imaging. Eur J Nucl Med Mol Imaging. modalities in pediatric primary bone tumors. Pediatr Radiol.
2015;42:1767–77. 2012;42(4):418–30.
24. Hongtao L, Hui Z, Bingshun W, Xiaojin W, Zhiyu W, Shuier Z, 42. Feldman F, Van Heertum R, Saxena C, Parisien M. 18FDG-
et al. 18F-FDG positron emission tomography for the assessment PET applications for cartilage neoplasms. Skelet Radiol.
of histological response to neoadjuvant chemotherapy in osteosar- 2005;34(7):367–74.
comas: a meta-analysis. Surg Oncol. 2012;21(4):e165–70. 43. Kransdorf MJ, Moser RP Jr, Gilkey FW. Fibrous dysplasia.
25. Gaston LL, Di Bella C, Slavin J, Hicks RJ, Choong PF. 18F-FDG Radiographics. 1990;10(3):519–37.
PET response to neoadjuvant chemotherapy for Ewing sarcoma 44. Fitzpatrick KA, Taljanovic MS, Speer DP, Graham AR, Jacobson
and osteosarcoma are different. Skelet Radiol. 2011;40(8): JA, Barnes GR, et al. Imaging findings of fibrous dysplasia
1007–15. with histopathologic and intraoperative correlation. AJR Am J
26. Cheon GJ, Kim MS, Lee JA, Lee SY, Cho WH, Song WS, et al. Roentgenol. 2004;182(6):1389–98.
Prediction model of chemotherapy response in osteosarcoma by 45. D’Souza MM, Jaimini A, Khurana A, Tripathi M, Sharma R,
18
F-FDG PET and MRI. J Nucl Med. 2009;50(9):1435–40. Mondal A, et al. Polyostotic fibrous dysplasia on F-18 FDG PET/
27. Healey JH, Ghelman B. Osteoid osteoma and osteoblastoma. CT imaging. Clin Nucl Med. 2009;34(6):359–61.
Current concepts and recent advances. Clin Orthop Relat Res. 46. Case DB, Chapman CN Jr, Freeman JK, Polga JP. Best cases
1986;(204):76–85. from the AFIP: atypical presentation of polyostotic fibrous dys-
28. Kransdorf MJ, Stull MA, Gilkey FW, Moser RP Jr. Osteoid oste- plasia with myxoma (Mazabraud syndrome). Radiographics.
oma. Radiographics. 1991;11(4):671–96. 2010;30(3):827–32.
29. Iyer RS, Chapman T, Chew FS. Pediatric bone imaging: diag- 47. Berrebi O, Steiner C, Keller A, Rougemont AL, Ratib O. F-18
nostic imaging of osteoid osteoma. AJR Am J Roentgenol. fluorodeoxyglucose (FDG) PET in the diagnosis of malignant
2012;198(5):1039–52. transformation of fibrous dysplasia in the pelvic bones. Clin Nucl
30. Saccomanni B. Osteoid osteoma and osteoblastoma of the Med. 2008;33(7):469–71.
spine: a review of the literature. Curr Rev Musculoskelet Med. 48. Qu N, Yao W, Cui X, Zhang H. Malignant transformation in
2009;2(1):65–7. monostotic fibrous dysplasia: clinical features, imaging features,
31. Chotel F, Franck F, Solla F, Dijoud F, Kohler R, Berard J, et al. outcomes in 10 patients, and review. Medicine (Baltimore).
Osteoid osteoma transformation into osteoblastoma: fact or fic- 2015;94(3):e369.
tion? Orthop Traumatol Surg Res. 2012;98(6 Suppl):S98–S104. 49. Murphey MD, Nomikos GC, Flemming DJ, Gannon FH, Temple
32. Kole AC, Nieweg OE, Hoekstra HJ, van Horn JR, Koops HS, HT, Kransdorf MJ. From the archives of AFIP. Imaging of giant
Vaalburg W. Fluorine-18-fluorodeoxyglucose assessment of cell tumor and giant cell reparative granuloma of bone: radiologic-
glucose metabolism in bone tumors. J Nucl Med. 1998;39(5): pathologic correlation. Radiographics. 2001;21(5):1283–309.
810–5. 50. Aoki J, Watanabe H, Shinozaki T, Takagishi K, Ishijima H,
33. Al-Muqbel KM, Al-Omari MH, Audat ZA, Alqudah Oya N, et al. FDG PET of primary benign and malignant bone
MA. Osteoblastoma is a metabolically active benign bone tumor tumors: standardized uptake value in 52 lesions. Radiology.
on 18F-FDG PET imaging. J Nucl Med Technol. 2013;41(4): 2001;219(3):774–7.
308–10. 51. Strauss LG, Dimitrakopoulou-Strauss A, Koczan D, Bernd
34. Brenner W, Bohuslavizki KH, Eary JF. PET imaging of osteosar- L, Haberkorn U, Ewerbeck V, et al. 18F-FDG kinet-
coma. J Nucl Med. 2003;44(6):930–42. ics and gene expression in giant cell tumors. J Nucl Med.
35. Volker T, Denecke T, Steffen I, Misch D, Schonberger S, Plotkin 2004;45(9):1528–35.
M, et al. Positron emission tomography for staging of pediatric 52. O’Connor W, Quintana M, Smith S, Willis M, Renner J. The
sarcoma patients: results of a prospective multicenter trial. J Clin hypermetabolic giant: 18F-FDG avid giant cell tumor identified on
Oncol. 2007;25(34):5435–41. PET-CT. J Radiol Case Rep. 2014;8(6):27–38.
36. Bredella MA, Caputo GR, Steinbach LS. Value of FDG positron 53. Davila D, Antoniou A, Chaudhry MA. Evaluation of osseous
emission tomography in conjunction with MR imaging for evalu- metastasis in bone scintigraphy. Semin Nucl Med. 2015;45:3–15..
ating therapy response in patients with musculoskeletal sarcomas. Erratum in: Semin Nucl Med. 2015;45:266
AJR Am J Roentgenol. 2002;179(5):1145–50. 54. Even-Sapir E. Imaging of malignant bone involvement by mor-
37. Hyun OJ, Luber BS, Leal JP, Wang H, Bolejack V, Schuetze phologic, scintigraphic, and hybrid modalities. J Nucl Med.
SM, et al. Response to early treatment evaluated with 18F-FDG 2005;46:1356–67.
PET and PERCIST 1.0 predicts survival in patients with Ewing 55. Mariani G, Bruselli L, Kuwert T, Kim EE, Flotats A, Israel O,
sarcoma family of tumors treated with a monoclonal anti- et al. A review on the clinical uses of SPECT/CT. Eur J Nucl Med
body to the insulin-like growth factor 1 receptor. J Nucl Med. Mol Imaging. 2010;37(10):1959–85.
2016;57(5):735–40. 56. Schmitz A, Risse JH, Textor J, Zander D, Biersack HJ, Schmitt
38. Eary JF, Conrad EU, Bruckner JD, Folpe A, Hunt KJ, Mankoff O, et al. FDG-PET findings of vertebral compression frac-
DA, et al. Quantitative [F-18]fluorodeoxyglucose positron emis- tures in osteoporosis: preliminary results. Osteoporos Int.
sion tomography in pretreatment and grading of sarcoma. Clin 2002;13(9):755–61.
Cancer Res. 1998;4(5):1215–20. 57. Bush LA, Chew FS. Subtrochanteric femoral insufficiency frac-
39. Hwang JP, Lim I, Kong CB, Jeon DG, Byun BH, Kim BI, et al. ture in woman on bisphosphonate therapy for glucocorticoid-
Prognostic value of SUVmax measured by pretreatment fluorine-18 flu- induced osteoporosis. Radiol Case Rep. 2009;4(1):261.
orodeoxyglucose positron emission tomography/computed tomogra- 58. Patel RN, Ashraf A, Sundaram M. Atypical fractures follow-
phy in patients with Ewing sarcoma. PLoS One. 2016;11(4):e0153281. ing bisphosphonate therapy. Semin Musculoskelet Radiol.
https://doi.org/10.1371/journal.pone.0153281. 2016;20(4):376–81.
40. Hawkins DS, Schuetze SM, Butrynski JE, Rajendran JG, Vernon 59. Raynor W, Houshmand S, Gholami S, Emamzadehfard S,
CB, Conrad EU III, et al. [18F]Fluorodeoxyglucose positron emis- Rajapakse CS, Blomberg BA, et al. Evolving role of molecular
sion tomography predicts outcome for Ewing sarcoma family of imaging with 18F-sodium fluoride PET as a biomarker for calcium
tumors. J Clin Oncol. 2005;23(8834):8828–34. metabolism. Curr Osteoporos Rep. 2016;14(4):115–25.
41. Lee FY, Yu J, Chang SS, Fawwaz R, Parisien MV. Diagnostic 60. Siddique M, Blake GM, Frost ML, Moore AE, Puri T, Marsden
value and limitations of fluorine-18 fluorodeoxyglucose positron PK, et al. Estimation of regional bone metabolism from whole-
emission tomography for cartilaginous tumors of bone. J Bone body 18F-fluoride PET static images. Eur J Nucl Med Mol
Joint Surg Am. 2004;86-A(12):2677–85. Imaging. 2012;39(2):337–43.
61. Siddique M, Frost ML, Moore AE, Fogelman I, Blake 79. Murphey MD, Foreman KL, Klassen-Fischer MK, Fox MG,
GM. Correcting 18F-fluoride PET static scan measurements of Chung EM, Kransdorf MJ. From the radiologic pathology
skeletal plasma clearance for tracer efflux from bone. Nucl Med archives imaging of osteonecrosis: radiologic-pathologic correla-
Commun. 2014;35(3):303–10. tion. Radiographics. 2014;34(4):1003–28.
62. Fogelman I, McKillop JH, Bessent RG, Boyle IT, Turner JG, 80. Lavernia CJ, Sierra RJ, Grieco FR. Osteonecrosis of the femoral
Greig WR. The role of bone scanning in osteomalacia. J Nucl head. J Am Acad Orthop Surg. 1999;7(4):250–61.
Med. 1978;19(3):245–8. 81. Beckmann J, Matsuura M, Grässel S, Kock F, Grifka J, Tingart
63. Chakraborty PP, Bhattacharjee R, Mukhopadhyay S, Chowdhury M. A μCT analysis of the femoral bone stock in osteonecrosis of
S. ‘Rachitic rosary sign’ and ‘tie sign’ of the sternum in tumour- the femoral head compared to osteoarthrosis. Arch Orthop Trauma
induced osteomalacia. BMJ Case Rep. 2016. pii: bcr2016214766. Surg. 2009;129(4):501–5.
https://doi.org/10.1136/bcr-2016-214766. 82. Theodorou DJ, Malizos KN, Beris AE, Theodorou SJ, Soucacos
64. Kaneuchi Y, Hakozaki M, Yamada H, Hasegawa O, Tajino T, PN. Multimodal imaging quantitation of the lesion size in
Konno S. Missed causative tumors in diagnosing tumor-induced osteonecrosis of the femoral head. Clin Orthop Relat Res.
osteomalacia with 18F-FDG PET/CT: a potential pitfall of 2001;386:54–63.
standard-field imaging. Hell J Nucl Med. 2016;19(1):46–8. 83. Yeh LR, Chen CK, Huang YL, Pan HB, Yang CF. Diagnostic
65. Jain AS, Shelley S, Muthukrishnan I, Kalal S, Amalachandran performance of MR imaging in the assessment of subchondral
J, Chandran S. Diagnostic importance of contrast enhanced fractures in avascular necrosis of the femoral head. Skelet Radiol.
18
F-fluorodeoxyglucose positron emission computed tomogra- 2009;38(6):559–64.
phy in patients with tumor induced osteomalacia: our experience. 84. Balakrishnan A, Schemitsch EH, Pearce D, McKee
Indian J Nucl Med. 2016;31(1):14–9. MD. Distinguishing transient osteoporosis of the hip from avas-
66. Breer S, Brunkhorst T, Beil FT, Peldschus K, Heiland M, cular necrosis. Can J Surg. 2003;46(3):187–92.
Klutmann S, et al. 68Ga DOTA-TATE PET/CT allows tumor local- 85. Ikemura S, Yamamoto T, Motomura G, Nakashima Y, Mawatari
ization in patients with tumor-induced osteomalacia but negative T, Iwamoto Y. MRI evaluation of collapsed femoral heads in
111
In-octreotide SPECT/CT. Bone. 2014;64:222–7. patients 60 years old or older: differentiation of subchondral insuf-
67. Agrawal K, Bhadada S, Mittal BR, Shukla J, Sood A, Bhattacharya ficiency fracture from osteonecrosis of the femoral head. AJR Am
A, et al. Comparison of 18F-FDG and 68Ga DOTATATE PET/CT J Roentgenol. 2010;195(1):W63–8.
in localization of tumor causing oncogenic osteomalacia. Clin 86. Kim HK, Kaste S, Dempsey M, Wilkes D. A comparison of non-
Nucl Med. 2015;40(1):e6–e10. contrast and contrast-enhanced MRI in the initial stage of Legg-
68. Bhavani N, Reena Asirvatham A, Kallur K, Menon AS, Pavithran Calve-Perthes disease. Pediatr Radiol. 2013;43(9):1166–73.
PV, Nair V, et al. Utility of Gallium-68 DOTANOC PET/CT in the 87. Du J, Lu A, Dempsey M, Herring JA, Kim HK. MR perfusion
localization of Tumour-induced osteomalacia. Clin Endocrinol. index as a quantitative method of evaluating epiphyseal perfusion
2016;84(1):134–40. in Legg-Calve-Perthes disease and correlation with short-term
69. Gahier Penhoat M, Drui D, Ansquer C, Mirallie E, Maugars Y, radiographic outcome: a preliminary study. J Pediatr Orthop.
Guillot P. Contribution of 18-FDG PET/CT to brown tumor detec- 2013;33(7):707–13.
tion in a patient with primary hyperparathyroidism. Joint Bone 88. Chan WP, Liu YJ, Huang GS, Lin MF, Huang S, Chang YC,
Spine. 2017;84(2):209–12. et al. Relationship of idiopathic osteonecrosis of the fem-
70. Niemann KE, Kropil F, Hoffmann MF, Coulibaly MO, Schildhauer oral head to perfusion changes in the proximal femur by
TA. A 23-year-old patient with secondary tumoral calcinosis: dynamic contrast- e nhanced MRI. AJR Am J Roentgenol.
regression after subtotal parathyroidectomy: a case report. Int J 2011;196(3):637–43.
Surg Case Rep. 2016;23:56–60. 89. Kaushik A, Sankaran B, Varghese M. Prognostic value of dynamic
71. Ando T, Mochizuki Y, Iwata T, Nishikido M, Shimazaki MRI in assessing post-traumatic femoral head vascularity. Skelet
T, Furumoto A, et al. Aggressive pulmonary calcification Radiol. 2009;38(6):565–9.
developed after living donor kidney transplantation in a 90. Goldsmith DP, Vivino FB, Eichenfield AH, Athreya BH, Heyman
patient with primary hyperparathyroidism. Transplant Proc. S. Nuclear imaging and clinical features of childhood reflex neu-
2013;45(7):2825–30. rovascular dystrophy: comparison with adults. Arthritis Rheum.
72. Hwang GJ, Lee JD, Park CY, Lim SK. Reversible extraskeletal 1989;32:480–5.
uptake of bone scanning in primary hyperparathyroidism. J Nucl 91. Kobayashi N, Inaba Y, Tateishi U, Yukizawa Y, Ike H, Inoue T,
Med. 1996;37(3):469–71. et al. New application of 18F-fluoride PET for the detection of
73. Terry DW Jr, Isitman AT, Holmes RA. Radionuclide bone images bone remodeling in early-stage osteoarthritis of the hip. Clin Nucl
in hypertrophic pulmonary osteoarthropathy. Am J Roentgenol Med. 2013;38(10):e379–83.
Radium Ther Nucl Med. 1975;124(4):571–6. 92. Kobayashi N, Inaba Y, Tateishi U, Ike H, Kubota S, Inoue T, et al.
74. Fredericson M, Bergman AG, Hoffman KL, Dillingham Comparison of 18F-fluoride positron emission tomography and
MS. Tibial stress reaction in runners. Correlation of clinical symp- magnetic resonance imaging in evaluating early-stage osteoarthri-
toms and scintigraphy with a new magnetic resonance imaging tis of the hip. Nucl Med Commun. 2015;36(1):84–9.
grading system. Am J Sports Med. 1995;23(4):472–81. 93. Hirata Y, Inaba Y, Kobayashi N, Ike H, Yukizawa Y, Fujimaki H,
75. Kijowski R, Choi J, Shinki K, Del Rio AM, De Smet A. Validation et al. Correlation between mechanical stress by finite element
of MRI classification system for tibial stress injuries. AJR Am J analysis and 18F-fluoride PET uptake in hip osteoarthritis patients.
Roentgenol. 2012;198(4):878–84. J Orthop Res. 2015;33(1):78–83.
76. Sofka CM. Imaging of stress fractures. Clin Sports Med. 94. Lew DP, Waldvogel FA. Osteomyelitis. N Engl J Med.
2006;25(1):53–62. 1997;336(14):999–1007.
77. Fujii M, Abe K, Hayashi K, Kosuda S, Yano F, Watanabe S, et al. 95. Lew DP, Waldvogel FA. Osteomyelitis. Lancet.
Honda sign and variants in patients suspected of having a sacral 2004;364(9431):369–79.
insufficiency fracture. Clin Nucl Med. 2005;30(3):165–9. 96. Riise OR, Kirkhus E, Handeland KS, Flato B, Reiseter T,
78. Joshi P, Lele V, Gandhi R, Parab A. Honda sign on 18-FDG PET/ Cvancarova M, et al. Childhood osteomyelitis-incidence and dif-
CT in a case of lymphoma leading to incidental detection of sacral ferentiation from other acute onset musculoskeletal features in a
insufficiency fracture. J Clin Imaging Sci. 2012;2:29. population-based study. BMC Pediatr. 2008;8:45.
97. Cierny G III, Mader JT, Penninck JJ. A clinical staging system for 117. Kubota R, Yamada S, Kubota K, Ishiwata K, Tamahashi N, Ido
adult osteomyelitis. Clin Orthop Relat Res. 2003;414:7–24. T. Intratumoral distribution of fluorine-18- fluorodeoxyglucose
98. Chianelli M, Boerman OC, Malviya G, Galli F, Oyen WJ, Signore in vivo: high accumulation in macrophages and granula-
A. Receptor binding ligands to image infection. Curr Pharm Des. tion tissues studied by microautoradiography. J Nucl Med.
2008;14(31):3316–25. 1992;33(11):1972–80.
99. Weon YC, Yang SO, Choi YY, Shin JW, Ryu JS, Shin MJ, et al. 118. Yamada S, Kubota K, Kubota R, Ido T, Tamahashi N. High accu-
Use of Tc-99m HMPAO leukocyte scans to evaluate bone infec- mulation of fluorine-18-fluorodeoxyglucose in turpentine-induced
tion: incremental value of additional SPECT images. Clin Nucl inflammatory tissue. J Nucl Med. 1995;36(7):1301–6.
Med. 2000;25(7):519–26. 119. Kaim AH, Weber B, Kurrer MO, Gottschalk J, Von Schulthess
100. Palestro CJ, Swyer AJ, Kim CK, Goldsmith SJ. Infected knee GK, Buck A. Autoradiographic quantification of 18F-FDG
prosthesis: diagnosis with In-111 leukocyte, Tc-99m sulfur col- uptake in experimental soft-tissue abscesses in rats. Radiology.
loid, and Tc-99m MDP imaging. Radiology. 1991;179(3):645–8. 2002;223(2):446–51.
101. Palestro CJ, Love C, Tronco GG, Tomas MB, Rini JN. Combined 120. Ohtori S, Suzuki M, Koshi T, Yamashita M, Yamauchi K, Inoue G,
labeled leukocyte and technetium 99m sulfur colloid bone marrow et al. 18F-fluorodeoxyglucose-PET for patients with suspected spon-
imaging for diagnosing musculoskeletal infection. Radiographics. dylitis showing Modic change. Spine. 2010;35(26):E1599–603.
2006;26(3):859–70. 121. Kim SJ, Kim IJ, Suh KT, Kim YK, Lee JS. Prediction of resid-
102. Pakos EE, Trikalinos TA, Fotopoulos AD, Ioannidis JP. Prosthesis ual disease of spine infection using F-18 FDG PET/CT. Spine.
infection: diagnosis after total joint arthroplasty with antigranu- 2009;34(22):2424–30.
locyte scintigraphy with 99mTc-labeled monoclonal antibodies – a 122. Ito K, Kubota K, Morooka M, Hasuo K, Kuroki H, Mimori
meta-analysis. Radiology. 2007;242(1):101–8. A. Clinical impact of 18F-FDG PET/CT on the management
103. Meller J, Liersch T, Oezerden MM, Sahlmann CO, Meller and diagnosis of infectious spondylitis. Nucl Med Commun.
B. Targeting NCA-95 and other granulocyte antigens and recep- 2010;31(8):691–8.
tors with radiolabeled monoclonal antibodies (Mabs). Q J Nucl 123. Nanni C, Boriani L, Salvadori C, Zamparini E, Rorato G,
Med Mol Imaging. 2010;54(6):582–98. Ambrosini V, et al. FDG PET/CT is useful for the interim
104. Ruf J, Oeser C, Amthauer H. Clinical role of anti-granulocyte evaluation of response to therapy in patients affected by hae-
MoAb versus radiolabeled white blood cells. Q J Nucl Med Mol matogenous spondylodiscitis. Eur J Nucl Med Mol Imaging.
Imaging. 2010;54(6):599–616. 2012;39(10):1538–44.
105. Richter WS, Ivancevic V, Meller J, Lang O, Le Guludec D, Szilvazi 124. Hungenbach S, Delank KS, Dietlein M, Eysel P, Drzezga
I, et al. 99mTc-besilesomab (Scintimun) in peripheral osteomyeli- A, Schmidt MC. 18F-fluorodeoxyglucose uptake pattern in
tis: comparison with 99mTc-labelled white blood cells. Eur J Nucl patients with suspected spondylodiscitis. Nucl Med Commun.
Med Mol Imaging. 2011;38(5):899–910. 2013;34(11):1068–74.
106. Prandini N, Feggi L, Massari L. Diagnosis of bone infections 125. Riccio SA, Chu AK, Rabin HR, Kloiber R. Fluorodeoxyglucose
using 99mTc-HMPAO labelled leukocytes. Nucl Med Commun. positron emission tomography/computed tomography interpre-
2002;23(12):1245–8. tation criteria for assessment of antibiotic treatment response in
107. Chen SH, Ho KC, Hsieh PH, Lee MS, Yen TC. Potential
pyogenic spine infection. Can Assoc Radiol J. 2015;66(2):145–52.
clinical role of 18F FDG-PET/CT in detecting hip prosthesis 126. Lee IS, Lee JS, Kim SJ, Jun S, Suh KT. Fluorine-18-
infection: a study in patients undergoing two-stage revision fluorodeoxyglucose positron emission tomography/computed
arthroplasty with an interim spacer. Q J Nucl Med Mol Imaging. tomography imaging in pyogenic and tuberculous spondylitis:
2010;54(4):429–35. preliminary study. J Comput Assist Tomogr. 2009;33(4):587–92.
108. Gemmel F, Van den Wyngaert H, Love C, Welling MM, Gemmel 127. Mazzie JP, Brooks MK, Gnerre J. Imaging and management of
P, Palestro CJ. Prosthetic joint infections: radionuclide state- postoperative spine infection. Neuroimaging Clin North Am.
of-the-art imaging. Eur J Nucl Med Mol Imaging. 2012;39(5): 2014;24(2):365–74.
892–909. 128. Fuster D, Tomas X, Granados U, Soriano A. Prospective com-
109. Rini JN, Palestro CJ. Imaging of infection and inflammation parison of whole-body 18F-FDG PET/CT and MRI of the spine
with 18F-FDG-labeled leukocytes. Q J Nucl Med Mol Imaging. in the diagnosis of haematogenous spondylodiscitis: response
2006;50(2):143–6. to comments by Soussan. Eur J Nucl Med Mol Imaging.
110. Calderone RR, Larsen JM. Overview and classification of spinal 2015;42(2):356–7.
infections. Orthop Clin North Am. 1996;27(1):1–8. 129. Skanjeti A, Penna D, Douroukas A, Cistaro A, Arena V, Leo G,
111. Gouliouris T, Aliyu SH, Brown NM. Spondylodiscitis: update et al. PET in the clinical work-up of patients with spondylodis-
on diagnosis and management. J Antimicrob Chemother. citis: a new tool for the clinician? Q J Nucl Med Mol Imaging.
2010;65(Suppl 3):iii11–24. 2012;56(6):569–76.
112. Prandini N, Lazzeri E, Rossi B, Erba P, Parisella MG, Signore 130. Rosen RS, Fayad L, Wahl RL. Increased 18F-FDG uptake in
A. Nuclear medicine imaging of bone infections. Nucl Med degenerative disease of the spine: characterization with 18F-FDG
Commun. 2006;27(8):633–44. PET/CT. J Nucl Med. 2006;47(8):1274–80.
113. Gratz S, Dorner J, Oestmann JW, Opitz M, Behr T, Meller J, et al. 131. Blom AW, Brown J, Taylor AH, Pattison G, Whitehouse S,
67
Ga-citrate and 99Tcm-MDP for estimating the severity of verte- Bannister GC. Infection after total knee arthroplasty. J Bone Joint
bral osteomyelitis. Nucl Med Commun. 2000;21(1):111–20. Surg Br. 2004;86(5):688–91.
114. Fuster D, Sola O, Soriano A, Monegal A, Setoain X, Tomas X, 132. Kurtz S, Mowat F, Ong K, Chan N, Lau E, Halpern M. Prevalence
et al. A prospective study comparing whole-body FDG PET/CT to of primary and revision total hip and knee arthroplasty in the
combined planar bone scan with 67Ga SPECT/CT in the diagnosis United States from 1990 through 2002. J Bone Joint Surg Am.
of spondylodiskitis. Clin Nucl Med. 2012;37(9):827–32. 2005;87(7):1487–97.
115. Love C, Palestro CJ. Nuclear medicine imaging of bone infec- 133. Parvizi J, Della Valle CJ. AAOS clinical practice guideline: diag-
tions. Clin Radiol. 2016;71(7):632–46. nosis and treatment of periprosthetic joint infections of the hip and
116. Pauwels EK, Ribeiro MJ, Stoot JH, McCready VR, Bourguignon knee. J Am Acad Orthop Surg. 2010;18(12):771–2.
M, Maziere B. FDG accumulation and tumor biology. Nucl Med 134. Parvizi J, Zmistowski B, Adeli B. Periprosthetic joint infection:
Biol. 1998;25(4):317–22. treatment options. Orthopedics. 2010;33(9):659.
135. Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. 156. Shiloh R, Krishnan M. Radiation for treatment of painful
N Engl J Med. 2004;351(16):1645–54. bone metastases. Hematol Oncol Clin North Am. 2018;32:
136. Duus BR, Boeckstyns M, Kjaer L, Stadeager C. Radionuclide 459–68.
scanning after total knee replacement: correlation with pain 157. Krishnamurthy GT, Krishnamurthy S. Radionuclides for meta-
and radiolucent lines. A prospective study. Investig Radiol. static bone pain palliation: a need for rational re-evaluation in the
1987;22(11):891–4. new millennium. J Nucl Med. 2000;41:688–91.
137. Sousa R, Massada M, Pereira A, Fontes F, Amorim I, Oliveira 158. Silberstein EB, Buscombe JR, McEwan A, Taylor AT Jr. Society
A. Diagnostic accuracy of combined 99mTc-sulesomab and 99mTc- of Nuclear Medicine Procedure Guideline for Palliative Treatment
nanocolloid bone marrow imaging in detecting prosthetic joint of Painful Bone Metastases. Version 3.0. snmmi.files.cms-plus.
infection. Nucl Med Commun. 2011;32(9):834–9. com/docs/pg_ch25_0403.pdf. Accessed 25 Jan 2003.
138. van der Bruggen W, Bleeker-Rovers CP, Boerman OC, Gotthardt 159. Bodei L, Lam M, Chiesa C, Flux G, Brans B, Chiti A, et al.
M, Oyen WJ. PET and SPECT in osteomyelitis and prosthetic EANM procedure guideline for treatment of refractory meta-
bone and joint infections: a systematic review. Semin Nucl Med. static bone pain. Eur J Nucl Med Mol Imaging. 2008;35:
2010;40(1):3–15. 1934–40.
139. Delank KS, Schmidt M, Michael JW, Dietlein M, Schicha H, Eysel 160. Handkiewicz-Junak D, Poeppel TD, Bodei L, Aktolun C, Ezziddin
P. The implications of 18F-FDG PET for the diagnosis of endo- S, Giammarile F, et al. EANM guidelines for radionuclide therapy
prosthetic loosening and infection in hip and knee arthroplasty: of bone metastases with beta-emitting radionuclides. Eur J Nucl
results from a prospective, blinded study. BMC Musculoskelet Med Mol Imaging. 2018;45:846–59.
Disord. 2006;7:20. 161. Abruzzese E, Iuliano F, Trawinska MM, Di Maio M. 153Sm: its use
140. Estrada WN, Kim CK. Paget’s disease in a patient with breast can- in multiple myeloma and report of a clinical experience. Expert
cer. J Nucl Med. 1993;34(7):1214–6. Opin Investig Drugs. 2008;17:1379–87.
141. van Heerden BB, Prins MJ. The value of pinhole collimator imag- 162. Dispenzieri A, Wiseman GA, Lacy MQ, Hayman SR, Kumar
ing in the scintigraphic analysis of vertebral diseases. S Afr Med SK, Buadi F, et al. A phase II study of 153Sm-EDTMP and high-
J. 1989;75(6):280–3. dose melphalan as a peripheral blood stem cell conditioning
142. Mirra JM, Brien EW, Tehranzadeh J. Paget’s disease of bone: regimen in patients with multiple myeloma. Am J Hematol.
review with emphasis on radiologic features, Part I. Skelet Radiol. 2010;85:409–13.
1995;24(3):163–71. 163. Berenson JR, Yellin O, Patel R, Duvivier H, Nassir Y, Mapes R,
143. Mangham DC, Davie MW, Grimer RJ. Sarcoma arising in Paget’s et al. A phase I study of samarium lexidronam/bortezomib combi-
disease of bone: declining incidence and increasing age at presen- nation therapy for the treatment of relapsed or refractory multiple
tation. Bone. 2009;44(3):431–6. myeloma. Clin Cancer Res. 2009;15:1069–75.
144. Bush LA, Toresdahl B, Hoch B, Chew FS. Falsely negative F-18 164. McQuay HJ, Collins SL, Carroll D, Moore RA. Radiotherapy for
FDG PET of osteosarcoma arising in Paget disease. Radiol Case the palliation of painful bone metastases. Cochrane Database Syst
Rep. 2009;4(3):295. Rev. 2000;(2):CD001793.
145. Hadjipavlou A, Lander P, Srolovitz H, Enker IP. Malignant trans- 165. Bauman G, Charette M, Reid R, Sathya J. Radiopharmaceuticals
formation in Paget disease of bone. Cancer. 1992;70(12):2802–8. for the palliation of painful bone metastasis-a systemic review.
146. Sundaram M, Khanna G, El-Khoury GY. T1-weighted MR imag- Radiother Oncol. 2005;75:258–70.
ing for distinguishing large osteolysis of Paget’s disease from sar- 166. Finlay IG, Mason MD, Shelley M. Radioisotopes for the pallia-
comatous degeneration. Skelet Radiol. 2001;30(7):378–83. tion of metastatic bone cancer: a systematic review. Lancet Oncol.
147. Cook GJ, Maisey MN, Fogelman I. Fluorine-18-FDG PET in 2005;6:392–400.
Paget’s disease of bone. J Nucl Med. 1997;38(9):1495–7. 167. Reisfield GM, Silberstein EB, Wilson GR. Radiopharmaceuticals
148. Bourgeois S, Gykiere P, Goethals L, Everaert H, De Geeter for the palliation of painful bone metastases. Am J Hosp Palliat
FW. Aspecific uptake of 68GA-PSMA in Paget disease of the bone. Care. 2005;22:41–6.
Clin Nucl Med. 2016;41(11):877–8. 168. Christensen MH, Petersen LJ. Radionuclide treatment of pain-
149. Giovacchini G, Samanes Gajate AM, Messa C, Fazio F. Increased ful bone metastases in patients with breast cancer: a systematic
C-11 choline uptake in pagetic bone in a patient with coexist- review. Cancer Treat Rev. 2012;38:164–71.
ing skeletal metastases from prostate cancer. Clin Nucl Med. 169. D’Angelo G, Sciuto R, Salvatori M, Sperduti I, Mantini G, Maini
2008;33(11):797–8. CL, et al. Targeted “bone-seeking” radiopharmaceuticals for palli-
150. Cook GJ, Blake GM, Marsden PK, Cronin B, Fogelman ative treatment of bone metastases: a systematic review and meta-
I. Quantification of skeletal kinetic indices in Paget’s disease analysis. Q J Nucl Med Mol Imaging. 2012;56:538–43.
using dynamic 18F-fluoride positron emission tomography. J Bone 170. Tomblyn M. The role of bone-seeking radionuclides in the pallia-
Miner Res. 2002;17(5):854–9. tive treatment of patients with painful osteoblastic skeletal metas-
151. Coleman RE. Metastatic bone disease: clinical features, tases. Cancer Control. 2012;19:137–44.
pathophysiology and treatment strategies. Cancer Treat Rev. 171. Tunio M, Al Asiri M, Al Hadab A, Bayoumi Y. Comparative
2001;27:165–75. efficacy, tolerability, and survival outcomes of various radio-
152. Hage WD, Aboulafia AJ, Aboulafia DM. Incidence, location, and pharmaceuticals in castration-resistant prostate cancer with bone
diagnostic evaluation of metastatic bone disease. Orthop Clin metastasis: a meta-analysis of randomized controlled trials. Drug
North Am. 2000;31:515–28. Des Devel Ther. 2015;9:5291–9.
153. Macedo F, Ladeira K, Pinho F, Saraiva N, Bonito N, Pinto 172. Jong JM, Oprea-Lager DE, Hooft L, de Klerk JM, Bloemendal
L, Goncalves F. Bone metastases: an overview. Oncol Rev. HJ, Verheul HM, et al. Radiopharmaceuticals for pallia-
2017;11(1):321. https://doi.org/10.4081/oncol.2017.321. tion of bone pain in patients with castration-resistant pros-
154. Mantyh PW, Clohisy DR, Koltzenburg M, Hunt SP. Molecular tate cancer metastatic to bone: a systematic review. Eur Urol.
mechanisms of cancer pain. Nat Rev Cancer. 2002;2:201–9. 2016;70:416–26.
155. Kougioumtzopoulou A, Zygogianni A, Liakouli Z, Kypraiou 173. Zacho HD, Karthigaseu NN, Fonager RF, Petersen LJ. Treatment
E, Kouloulias V. The role of radiotherapy in bone metastases: a with bone-seeking radionuclides for painful bone metastases in
critical review of current literature. Eur J Cancer Care (Engl). patients with lung cancer: a systematic review. BMJ Support
2017;26(6) https://doi.org/10.1111/ecc.12724.. Epub 2017 Jun 20 Palliat Care. 2017;7:230–7.
174. Sartor O, Reid RH, Bushnell DL, Quick DP, Ell PJ. Safety 180. Liepe K, Shinto A. From palliative therapy to prolongation of sur-
and efficacy of repeat administration of samarium Sm-153 vival: 223RaCl2 in the treatment of bone metastases. Ther Adv Med
lexidronam to patients with metastatic bone pain. Cancer. Oncol. 2016;8:294–304.
2007;109:637–43. 181. Du Y, Carrio I, DeVincentis G, Fanti S, Ilhan H, Mommsen C,
175. Parker C, Nilsson S, Heinrich D, Helle SI, O’Sullivan JM, Fosså et al. Practical recommendations for radium-223 treatment of met-
SD, et al. Alpha emitter radium-223 and survival in metastatic astatic castration-resistant prostate cancer. Eur J Nucl Med Mol
prostate cancer. N Engl J Med. 2013;369:213–23. Imaging. 2017;44:1671–8.
176. Sartor O, Coleman R, Nilsson S, Heinrich D, Helle SI, O’Sullivan
JM, et al. Effect of radium-223 dichloride on symptomatic skel-
etal events in patients with castration-resistant prostate cancer and
bone metastases: results from a phase 3, double-blind, randomised Further Reading
trial. Lancet Oncol. 2014;15:738–46.
177. Parker CC, Coleman RE, Sartor O, Vogelzang NJ, Bottomley D, Glaudemans AWJM, Jutte PC, Cataldo MA, Cassar-Pullicino V,
Heinrich D, et al. Three-year safety of radium-223 dichloride in Gheyesens O, Bores O, et al. Consensus document for the diagnosis
patients with castration-resistant prostate cancer and symptom- of peripheral bone infection in adults: a joint paper by the EANM,
atic bone metastases from Phase 3 randomized Alpharadin in EBJS, and ESR (with ESCMID endorsement). Eur J Nucl Med Mol
symptomatic prostate cancer trial. Eur Urol. 2017; https://doi. Imaging. 2019;46:957–70.
org/10.1016/j.eururo.2017.06.021. Signore A, Scofienza LM, Borens O, Glaudemans AWJM, Cassar-
178. Poeppel TD, Handkiewicz-Junak D, Andreeff M, Becherer A, Pullicino V, Trampuz A, et al. Consensus document for the diag-
Bockisch A, Fricke E, et al. EANM guideline for radionuclide nosis of prosthetic joint infection: a joint paper by the EANM,
therapy with radium-223 of metastatic castration-resistant pros- EBJS, and ESR (with ESCMID endorsement). Eur J Nucl Med Mol
tate cancer. Eur J Nucl Med Mol Imaging. 2018;45:824–45. Imaging. 2019;46:971–88.
179. Oyen WJ, Sundram F, Haug AR, Kalevi K, Lewington V, Mäenpää Signore A, Scofienza LM, Borens O, Glaudemans AWJM, Cassar-
H, et al. Radium-223 dichloride (Ra-223) for the treatment of Pullicino V, Trampuz A, et al. Correction to: Consensus document
metastatic castration-resistant prostate cancer: optimizing clinical for the diagnosis of prosthetic joint infection: a joint paper by the
practice in nuclear medicine centers. J Oncol Pathol. 2015;3:1–25. EANM, EBJS, and ESR (with ESCMID endorsement). Eur J Nucl
https://doi.org/10.13032/tjop.2052-5931.100121. Med Mol Imaging. 2019;46:1203.
Hybrid Imaging of Melanoma and Other
Cutaneous Malignancies 25
Montserrat Estorch
Contents
25.2 Melanoma Staging, Classification, and Prognostic Factors 646
25.2.1 [18F]FDG PET/CT in Initial Melanoma Staging 647
25.2.2 [18F]FDG PET/CT in Follow-Up of Melanoma 649
25.2.3 [18F]FDG PET/CT in Preoperative Evaluation of Metastatic Melanoma 649
25.2.4 [18F]FDG PET/CT in Therapy Monitoring of Melanoma 650
25.3 Noncutaneous Melanoma 651
25.4 Cutaneous Non-melanoma Malignancies 652
References 652
Learning Objectives
• Summarize the main features of histology, predisposing 25.1 Introduction
risk factors and clinical presentation for cutaneous mela-
noma, noncutaneous melanoma, and cutaneous non- Imaging routinely used for melanoma and other cutaneous
melanoma malignancies. malignancies includes ultrasound; contrast-enhanced com-
• Summarize the main principles of treatment according to puted tomography (CT) of the chest, abdomen, and pelvis;
stage of the disease. magnetic resonance (MR) of the brain; and hybrid modal-
• Understand the rationale underlying the common clinical ity positron emission tomography/computed tomography
uses of [18F]FDG PET/CT in patients with cutaneous mel- (PET/CT).
anoma, noncutaneous melanoma, and cutaneous non- Melanoma is the first cause of all skin cancer deaths. Its
melanoma malignancies. incidence has increased over the past decades in white popu-
• Describe the established indications for [18F]FDG PET/ lations at an overall rate of 33% for men and 23% for women
CT in patients with cutaneous melanoma, noncutaneous [1]. This increase is an underestimate, because superficial
melanoma, and cutaneous non-melanoma malignancies, and in situ melanomas can be treated on an outpatient basis
including initial staging, assessment of response to treat- and may be not reported. The lifetime risk of developing
ments, and follow-up surveillance according to stage of cutaneous melanoma is 1 of 34 for women and 1 of 53 for
the disease. men, and the median age at diagnosis is 59 years [2].
• Summarize other potential uses of [18F]FDG PET/CT in Risk factors for melanoma include skin type, personal
patients with cutaneous melanoma, noncutaneous mela- history of prior melanoma, multiple clinical atypical moles
noma, and cutaneous non-melanoma malignancies. or dysplastic nevi, a positive family history of melanoma,
and, rarely, inherited genetic mutations. Increased ultraviolet
light exposure of a genetically predisposed population seems
to be responsible for the increased incidence [3].
M. Estorch (*) Diagnosis of melanoma should be based on a full-
Nuclear Medicine Department, Sant Pau Hospital, thickness excisional biopsy with a minimal side margin. The
Autonomous University of Barcelona, Barcelona, Spain
histology report should include information of the maximum
e-mail: mestorch@santpau.cat

646 M. Estorch
thickness in millimetres (Breslow), mitotic rate in case of a Table 25.1 TNM staging categories for cutaneous melanoma
tumour thickness below 1.0 mm, presence of ulceration, T
presence and extent of regression, clearance of the surgical Classification Thickness (mm) Ulceration status/mitoses
margins, and melanoma type. In addition, information on Tis NA NA
T1 <1.00 A: without ulceration and
anatomical site, including extra-cutaneous sites, and degree
mitosis <1/mm2
of sun damage is necessary. B: with ulceration or mitoses
The outcome of melanoma depends of the stage at presen- >1/mm2
tation. In general, prognosis is excellent (5-year survival in T2 1.01–2.00 A: without ulceration
more than 90% of patients) for localized primary tumours B: with ulceration
with thickness of 1.0 mm or less, whereas for tumours with T3 2.01–4.00 A: without ulceration
thickness >1.0 mm, survival rates range from 50% to 90%. B: with ulceration
Melanoma is a tumour with an unpredictable pattern of dis- T4 >4.00 A: without ulceration
B: with ulceration
semination, and it is important to stage it accurately to plan
N N. metastatic nodes Nodal metastatic burden
appropriate treatment and to estimate prognosis. This issue is
N0 0 NA
especially important in high-risk primary melanoma (i.e. N1 1 A: micrometastasisa
stage III and above), which can metastasize to regional B: macrometastasisb
lymph nodes and to visceral organs. Whereas, in patients N2 2–3 A: micrometastasisa
with stage I, low-risk melanoma, no other investigations, B: macrometastasisb
except physical extensive examination, are necessary. The C: in transit metastases/
tumour may recur in nearly 50% of patients with high risk, satellites without metastatic
nodes
being the risk of first recurrence greatest in the initial years
N3 4+ Metastatic nodes, or matted
after diagnosis: 20% of all first recurrences occur locally, nodes, or in transit metastases/
50% occur in the regional lymph nodes, and 30% occur at a satellites with metastatic
distance. nodes
Ultrasound imaging is useful to detect locoregional lym- M Site Serum LDH
phatic metastases and CT or [18F]FDG PET/CT for distant dis- M0 No distant NA
metastases
ease. Indications of these studies are primary disease staging, M1a Distant skin, Normal
surgical planning, and surveillance in high-risk patients and subcutaneous, or
monitoring response to systemic or locoregional therapy [4, 5]. nodal metastases
The use of imaging studies varies depending on clinical M1b Lung metastases Normal
guidelines, physician preferences, availability, and cost. In M1c All other visceral Normal
metastases
several studies whole-body [18F]FDG PET/CT has proven to
Any distant Elevated
be superior to CT alone in the detection of extracerebral mel- metastasis
anoma metastases [5, 6]. However, higher test accuracy does NA not applicable, LDH lactate dehydrogenase
not necessarily translate into a change of management or in a a
Micrometastases are diagnosed after sentinel lymph node biopsy
better patient outcome. This lack of evidence has led to reim- b
Macrometastases are defined as clinically detectable nodal metastases
bursement restrictions for PET/CT in several countries, with confirmed pathologically
the result that PET/CT is not generally accepted as a stan-
dard imaging modality for all stages/phases of melanoma. staging. The Clark level is not recommended as a staging
criterion, since it is not an independent prognostic factor
when mitotic rate is included in the analysis. Elevated serum
25.2 M
elanoma Staging, Classification, LDH is an independent and highly significant predictor of
and Prognostic Factors survival or outcome of stage IV patients, independent of
other factors [8]. Distant metastases cannot be incorporated
The current version of the American Joint Committee on into the staging system because of the variability of imaging
Cancer (AJCC) for melanoma staging and classification, studies used to diagnose them.
which includes sentinel node staging, is the only internation- The clinical practice guideline of the European Society
ally accepted classification system [7]. The TNM staging for Medical Oncology (ESMO) for melanoma recommends
categories and the stage groupings for cutaneous melanoma ultrasound for locoregional lymphatic node metastases in
are shown in Tables 25.1 and 25.2, respectively. tumour stages pTIb–pTIIIa and CT or whole-body [18F]FDG
According to the AJCC, histological features of the pri- PET/CT scan in stages >pTIIIa, to achieve an earlier diagno-
mary melanoma (tumour thickness, mitotic rate, and ulcer- sis of regional or systemic disease before surgical treatment
ation) are important features of melanoma prognosis and and sentinel node biopsy [5, 9, 10].
25 Hybrid Imaging of Melanoma and Other Cutaneous Malignancies 647
Table 25.2 Anatomic stage groupings for cutaneous melanoma

Clinical staginga Pathologic stagingb
T N M T N M
0 Tis N0 M0 0 Tis N0 M0
IA T1a N0 M0 IA T1a N0 M0
IB T1b N0 M0 IB T1b N0 M0
T2a N0 M0 T2a N0 M0
IIA T2b N0 M0 IIA T2b N0 M0
T3a N0 M0 T3a N0 M0
IIB T3b N0 M0 IIB T3b N0 M0
T4a N0 M0 T4a N0 M0
IIC T4b N0 M0 IIC T4b N0 M0
III Any T N > N0 M0 IIIA T1-4a N1a M0
T1-4a N2a M0
IIIB T1-4b N1a M0
T1-4b N2a M0
T1-4a N1b M0
T1-4a N2b M0
T1-4a N2c M0
IIIC T1-4b N1b M0
T1-4b N2b M0
T1-4b N2c M0
Any T N3 M0
IV Any T Any N M1 IV Any T Any N M1
a
Clinical staging includes microstaging of the primary melanoma and clinical/radiologic evaluation for metastases. By convention, it should be
used after complete excision of the primary melanoma with clinical assessment for regional and distant metastases
b
Pathologic staging includes microstaging of the primary melanoma and pathologic information about the regional lymph nodes after partial (i.e.
sentinel node biopsy) or complete lymphadenectomy. Pathologic stage 0 or stage IA patients are the exception; they do not require pathologic
evaluation of their lymph nodes
The last version of the National Comprehensive Cancer CT is the standard imaging modality for tumour staging
Network (NCCN) Clinical Practice Guidelines in Oncology in patients with advanced melanoma. Nevertheless, it has
for melanoma includes imaging studies that should be per- recently been shown that [18F]FDG PET/CT is clearly
formed at any stage of disease [11]. This clinical guideline superior to CT alone for the detection of extracerebral
suggests CT, [18F]FDG PET/CT, and MR, in patients with metastases, due to the hypermetabolic nature of melanoma
stage III or above, with the recommendation to obtain a base- cells [5] (Fig. 25.1). However, the clinical usefulness of
line imaging study for staging. [18F]FDG PET/CT in newly diagnosed high-risk mela-
Available scientific literature recommends that addi- noma is still controversial, and guidelines do not identify
tional staging with [18F]FDG PET/CT must be offered to all which patients should have PET/CT for staging. Moreover,
patients found to have metastatic melanoma (stage III) there is disagreement on the cost-effectiveness of [18F]
according to sentinel node biopsy, as this is the most sensi- FDG PET/CT as a part of the diagnostic workup, under-
tive way to detect distant metastases [5, 12]. Moreover, scoring the limited usefulness of PET/CT in the early stage
there is consensus in [18F]FDG PET/CT being the best of melanoma and in clinically node-negative head and
modality to identify melanoma metastases in patients with neck cutaneous melanoma [16–19]. Medicare does not
high-risk metastatic melanoma (stages III–IV) [5, 12–14] cover initial staging of regional lymph nodes for mela-
since this examination can detect unrecognized metastases, noma patients, whereas it covers all indications for initial
which may lead to altered management in 10–19% of antitumour treatment strategy [20].
patients [14–16]. In advanced stages of melanoma, [18F]FDG PET/CT
allows detection of unexpected metastases that are missed or
not seen with conventional imaging. In early melanoma
25.2.1 [18F]FDG PET/CT in Initial Melanoma stages, the scan has a limited diagnostic role, due to its low
Staging sensitivity to detect microscopic lymphatic disease [21].
Recently, it has been shown that [18F]FDG PET/CT per-
The major objective of initial melanoma staging is treatment formed before lymphadenectomy, when the sentinel node is
planning and prognosis. positive, may provide false-positive findings that may trigger
648 M. Estorch
a b
Fig. 25.1 A 71-year-old woman with primary malignant melanoma of transaxial-fused PET/CT images with a single metastatic lesion on the
left thigh (pT2aN2bM0, stage IIIB) at diagnosis. [18F]FDG PET/CT right lung (b) and spread bone disease (c)
shows (a) whole-body image with spread metastatic disease; (b and c)
additional, predominantly invasive, procedures which do not

have an impact on the therapeutic strategy [22]. Key Learning Points
A recent systematic review and meta-analysis concluded • The major objective of initial melanoma staging is
that [18F]FDG PET has a high sensitivity, specificity, and treatment planning and estimation of prognosis.
overall accuracy in stage IIIb/IIIc melanoma [23]. These • [18F]FDG PET/CT is clearly superior to CT alone
conclusions have been confirmed by a retrospective study for the detection of extracerebral metastases.
that identifies four factors predictive of PET/CT positivity • However, the clinical utility of [18F]FDG PET/CT
(mitotic rate >3/mm2, thickness >4 mm, lymphadenopathy, in newly diagnosed high-risk melanoma is contro-
and bleeding) [24]. A combinatorial index obtained from versial, and guidelines do not identify which
these four factors allows identification of those patients with patients should have a staging [18F]FDG PET/CT
the highest risk of [18F]FDG PET/CT positivity: 5.8% of scan.
patients with 0 to 2 factors at the high-risk level had positive • [18F]FDG PET/CT has a limited diagnostic value
PET/CT, whereas this fraction increased to 47.4% in patients for the early melanoma stages, whereas in advanced
with three factors and was 100% when all four high-risk fac- stages PET/CT allows detection of unexpected
tors were present. Results obtained from another recent ret- metastases that are missed or not seen with conven-
rospective study are similar, showing that primary staging by tional imaging.
[18F]FDG PET/CT leads to therapy change in 59% of patients • [18F]FDG PET/CT has a high sensitivity, specificity,
with stage III/IV of malignant melanoma, with a 5-year sur- and accuracy in stage IIIb/IIIc melanoma.
vival rate of 33.6% [6].
25.2.2 [18F]FDG PET/CT in Follow-Up

of Melanoma Key Learning Points
• The goal of melanoma follow-up is to detect recur-
The goal of melanoma follow-up is to detect recurrence and rence and metastases among high-risk patients as
metastases in high-risk patients as early as possible, because early as possible.
this may be a potential chance for curative surgical or medi- • [18F]FDG PET/CT is the best modality to iden-
cal treatment. tify metastases in high-risk melanoma (stages
Sensitivity of [18F]FDG PET/CT in detecting systemic III–IV).
metastases varies and depends on the number and frequency • Sensitivity of [18F]FDG PET/CT in detecting sys-
of scans or on how long follow-up is needed, because recur- temic metastases varies and depends on the number
rent melanoma cannot be ruled out by one-time imaging. and frequency of the scans or on how long follow-
When [18F]FDG PET/CT is repeated several times during up is to be extended.
follow-up, the sensitivity increases. • It is strongly discouraged to perform [18F]FDG
Patients with high-risk melanoma have a life-long risk of PET/CT as routine surveillance without clinical
recurrence, and, therefore, the sensitivity of [18F]FDG PET/ indication, because the impact on patient manage-
CT correlates with the length of follow-up. Unfortunately, ment is limited.
each patient has an unpredictable disease progression pattern • [18F]FDG PET/CT is recommended in all patients
[23, 25, 26]. As recently reported, the sensitivity of a single found to have metastatic malignant melanoma
[18F]FDG PET/CT scan to detect occult disease varies from (stage III) according to sentinel node diagnostics
78.6%, when the follow-up is 6 months, to 29.5% when it is and in patients with high-risk metastatic melanoma
5 years [26]. In order to perform a correct follow-up, it is also (stages III–IV), to identify melanoma metastases.
important to define who is a high-risk patient, as there is dis-
agreement in the scientific literature. It is obvious that
patients with ulcerated intermediate-thickness melanoma
(stage IIIb) or with any thick melanoma (stage IVa–b) should 25.2.3 [18F]FDG PET/CT in Preoperative
be regarded as high-risk patients, but when only sentinel- Evaluation of Metastatic Melanoma
positive stage III patients are considered, [18F]FDG PET/CT
has minimal clinical value in identifying unsuspected recur- The prognosis of metastatic melanoma is bleak, being the
rence [27, 28]. overall 1-year survival about 33% [7]. It is important to iden-
The clinical indications and impact of [18F]FDG PET/CT tify patients that could be treated by complete surgical resec-
on management in different primary tumours including tion of metastases, because there is a clear survival benefit
malignant melanoma have been reported [29, 30]. It has been when complete metastasectomy is achieved (2-year survival
shown that the fourth and subsequent follow-up PET/CT rates up to 50%) [4, 33]. Furthermore, when metastases are
scan performed at the end of primary treatment leads to a non-resectable, it is important to start systemic treatment
change in management in 31.6% of patients when the scans without delay.
are acquired for appropriate clinical indications. Whereas, it [18F]FDG PET/CT has a high impact on clinical manage-
is strongly discouraged to perform [18F]FDG PET/CT as rou- ment of patients with metastatic melanoma who are candi-
tine surveillance without clinical indication, because the dates for surgical resection, resulting in frequent management
impact on patient management is limited. changes and avoiding unnecessary surgery in a high number
The following literature-based recommendations to of patients [5, 6, 10, 34]. Recently, the clinical impact of [18F]
perform [18F]FDG PET/CT at follow-up have been pro- FDG PET/CT when complete metastasectomy is planned
posed [12]: has been described [35]: PET/CT may exclude the presence
of metastases and thus spares unnecessary surgery; reveals
–– All patients found to have metastatic malignant mela- more unresectable metastases than conventional imaging,
noma (stage III) according to sentinel node diagnostics thus avoiding futile surgery and enabling timely start of sys-
[5, 31, 32]. temic treatment; or allows a more precise planning of sur-
–– Patients with high-risk metastatic melanoma (stages III– gery including enlargement or reduction of the surgical field
IV), to identify melanoma metastases [5, 13, 14, 32]. (Fig. 25.2).
650 M. Estorch
a b
c d
Fig. 25.2 (a) Contrast-enhanced CT image shows an enlarged suspi- [18F]FDG PET/CT image (d) shows another soft-tissue metastasis in the
cious lymph node in the left inguinal region. (b) [18F]FDG PET/CT left popliteal fossa not seen on the CT image (c) (Schüle et al. Eur J
image confirms the lymph node metastasis in the left inguinal region. Nucl Med Mol Imaging. 2016;43:482–8)
Changes of clinical management do not necessarily mean an 25.2.4 [18F]FDG PET/CT in Therapy Monitoring
improvement on melanoma patient outcome, but complete of Melanoma
metastasectomy has been shown to be beneficial, and it has been
reported that [18F]FDG PET/CT performed before surgical treat- New molecularly targeted therapies and immunotherapies
ment is cost-effective, also providing a survival benefit [36, 37]. have changed the clinical scenario of advanced melanoma,
improving overall survival of patients undergoing systemic
treatment from 25% to over 70% [38].
Molecularly targeted therapies include small molecule
Key Learning Points inhibitors for BRAF-mutated melanomas, which specifi-
• Patient candidates to complete surgical resection of cally inhibit the most common oncogenic driver mutation
metastases should be identified correctly, because responsible for melanoma cell proliferation and survival.
there is a clear survival benefit when complete Immunotherapies include monoclonal antibodies tar-
metastasectomy is achieved. geting immune checkpoint receptors such as cytotoxic
• When metastases are non-resectable, it is important T-lymphocyte-associated protein 4 (CTLA4) and pro-
to start systemic treatment without delay. grammed death-1 (PD1). These novel agents have distinct
• [18F]FDG PET/CT may exclude the presence of mechanisms of action compared to traditional cytotoxic
metastases, thus sparing unnecessary surgery. chemotherapy, and imaging specialists need to understand
• [18F]FDG PET/CT reveals more unresectable these mechanisms to properly interpret the [18F]FDG PET/
metastases than conventional imaging, thus avoid- CT findings [39].
ing futile surgery and enabling timely start of sys- While reduction in viable cells with resultant shrinkage of
temic treatment. lesions is usually seen in melanoma that is responding to
• [18F]FDG PET/CT allows a more precise planning BRAF inhibition, reduction in [18F]FDG uptake occurs much
of surgery. more rapidly. This is because effective downregulation of
extracellular signal-regulated kinase, the terminal gene of
the MAPK pathway, suppresses glycolysis via a network of

transcriptional regulators of glycolysis [40]. Thus, lack of Key Learning Points
early reduction in [18F]FDG uptake can indicate primary • Molecularly targeted therapies have distinct mecha-
refractory disease sites, as demonstrated by the lack of PET nisms of action that must be understood for proper
response in rare variant BRAF mutations that are not respon- interpretation of [18F]FDG PET/CT findings.
sive to MAPK pathway inhibition, whereas persistence or • Since both melanoma and immune infiltrates exhibit
reactivation of [18F]FDG uptake allows identification of pri- high [18F]FDG uptake, it is challenging for [18F]
mary refractory disease. FDG PET to differentiate patients with progression
For immunotherapy with anti-CTLA4, early imaging may from those with “pseudoprogression”.
be compromised by immune flare responses, but it enables
early detection and treatment of immune-related toxicities.
For anti-PD1/PDL1 therapy, imaging later in treatment may
provide better assessment of benefit [41] (Fig. 25.3). 25.3 Noncutaneous Melanoma
Since both melanoma and immune infiltrates can exhibit
high [18F]FDG uptake, it is challenging for [18F]FDG PET to Compared to the cutaneous melanoma, noncutaneous mela-
differentiate patients with progression from those experienc- noma is relatively rare (ocular melanomas accounting for
ing “pseudoprogression”. With experience, however, signs of 5.5% and mucosal melanomas accounting for 1.3% of all
immune-related inflammatory responses may be visualized melanomas) and has distinct epidemiologic characteristics,
on [18F]FDG PET/CT. If these features are seen, a follow-up clinical aspects, pattern of behaviour, and radiologic appear-
scan is recommended to assess for temporal change before ance. Each noncutaneous primary melanoma has its own
defining progression [42]. specific diagnostic and management challenges, depending
A: MAY 2017 B: AUGUST 2017
Fig. 25.3 A 73-year-old woman with primary cutaneous nasal malig- observed. (B) Fused transaxial [18F]FDG PET/CT images show
nant melanoma that was treated by immunotherapy due to disease pro- response to immunotherapy at bone (a) and right suprarenal metastatic
gression. [18F]FDG PET/CT studies were performed for therapy lesions (b) and progression at left suprarenal (b) and mesenteric meta-
monitoring prior to start treatment as a baseline evaluation (A) and after static lesions (c) (white arrow)
6 cycles of pembrolizumab (B): a dissociate response to treatment was
652 M. Estorch
on the anatomic location where they arise. Primary frequent References

localizations for noncutaneous melanoma are ocular (uveal,
choroidal), mucosal (head and neck, gastrointestinal and 1. Leiter U, Eigentler T, Garbe C. Epidemiology of skin cancer. Adv
Exp Med Biol. 2014;810:120–40.
genitourinary tracts), and unknown primary [43]. The man- 2. Ekwueme DU, Guy GP, Li C, Rim SH, Parelkar P, Chen SC. The
agement of noncutaneous melanoma also differs from cuta- health burden and economic costs of cutaneous melanoma mor-
neous, being adjuvant radiotherapy post-primary resection tality by race/ethnicity-United States, 2000 to 2006. J Am Acad
the first treatment recommended. The role of chemotherapy Dermatol. 2011;65:S133–43.
3. Colantonio S, Bracken MB, Beecker J. The association of indoor
and immunotherapy is not established, and there are no well- tanning and melanoma in adults: systematic review and meta-
established guidelines for staging. Data on the use of [18F] analysis. J Am Acad Dermatol. 2014;70:847–57.
FDG PET/CT in noncutaneous melanoma are limited and 4. Tsao H, Atkins MB, Sober AJ. Management of cutaneous mela-
often conflicting and should be interpreted with caution due noma. N Engl J Med. 2004;351:998–1012.
5. Xing Y, Bronstein Y, Ross MI, Askew RL, Lee JE, Gershenwald
to the small number of patients included in the studies [44]. JE, et al. Contemporary diagnostic imaging modalities for the stag-
ing and surveillance of melanoma patients: a meta-analysis. J Natl
Cancer Inst. 2011;103:129–42.
Key Learning Point 6. Schüle S-C, Eigentler TK, Garbe C, la Fougère C, Nikolaou K,
Pfannenberg C. Influence of 18F-FDG PET/CT on therapy manage-
• Data on the use of [18F]FDG PET/CT in noncutane- ment in patients with stage III/IV malignant melanoma. Eur J Nucl
ous melanoma are limited and often conflicting and Med Mol Imaging. 2016;43:482–8.
should be interpreted with caution due to the small 7. Balch CM, Gershenwald JE, Song SJ, Thompson JF, Atkins MB,
number of patients included in the studies. Byrd DR, et al. Final versions of 2009 AJCC melanoma staging and
classification. J Clin Oncol. 2009;27:6199–206.
8. Bedikian AY, Johnson MM, Warneke CL, Papadopoulos NE, Kim
K, Hwu WJ, et al. Prognostic factors that determine the long-term
survival of patients with unresectable metastatic melanoma. Cancer
Investig. 2008;26:624–33.
9. Dummer R, Hauschild A, Lindenblatt N, Pentheroudakis G,
25.4 Cutaneous Non-melanoma Keilholz U, on behalf of the ESMO Guidelines Committee.
Malignancies Cutaneous melanoma: ESMO clinical practice guidelines for
diagnosis, treatment and follow-up. Ann Oncol. 2015;26(S5):
The role of [18F]FDG PET/CT in the management of cutane- 126–35.
10. Bastiaannet E, Wobbes T, Hoekstra OS, van der Jagt EJ, Brouwers
ous non-melanoma malignancies is not clear, even it seems AH, Koelemij R, et al. Prospective comparison of [18F]fluorodeoxy-
to be useful in the management of squamous cell carcinoma glucose positron emission tomography and computed tomography
and cutaneous lymphoma [45, 46]. When treating aggressive in patients with melanoma with palpable lymph node metasta-
skin cancers, pre- and post-operative imaging provides ses: diagnostic accuracy and impact on treatment. J Clin Oncol.
2009;27:4774–80.
important information for treatment planning and multidisci- 11. Coit DG, Thompson JA, Algazi A, Andtbacka R, Bichakjian CK,
plinary cooperation of care [47, 48]. Carson WE III, et al. Melanoma, Version 2.2016. J Natl Compr
A retrospective study evaluating the use of diagnostic CT, Canc Netw. 2016;14:450–73.
[18F]FDG PET/CT, and MRI in the initial lymph node stag- 12. Petersen H, Holdgaard PC, Madsen PH, Knudsen LM, Gad D,
Gravergaard AE, et al. FDG PET/CT in cancer: comparison of
ing of 99 patients with Merkel cell carcinoma found a sensi- actual use with literature-based recommendations. Eur J Nucl Med
tivity of 85% and specificity of 95% with PET, 47% and 97% Mol Imaging. 2016;43:695–706.
for CT, and 0% and 86% for MRI (histopathology of lymph 13. Gao G, Gong B, Shen W. Meta-analysis of the additional value
nodes was used for determining sensitivity and specificity). of integrated 18FDG PET-CT for tumor distant metastasis staging:
comparison with 18FDG PET alone and CT alone. Surg Oncol.
Another study found that [18F]FDG PET has a significant 2013;22:195–200.
management impact in patients with Merkel cell cancer, 14. Bourgeois AC, Chang TT, Fish LM, Bradley YC. Positron emis-
changing staging in 33% and management in 43% [49]. sion tomography/computed tomography in melanoma. Radiol Clin
North Am. 2013;51:865–79.
15. Krug B, Crott R, Lonneux M, Baurain JF, Pirson AS, Vander Borght
T. Role of PET in the initial staging of cutaneous malignant mela-
Key Learning Point noma: systematic review. Radiology. 2008;249:836–44.
16. Bastiaannet E, Uyl-de Groot CA, Brouwers AH, van der Jagt EJ,
• The role of [18F]FDG PET/CT in the management Hoekstra OS, Oyen W, et al. Cost-effectiveness of adding FDG-
of non-melanoma cutaneous malignancies is not PET or CT to the diagnostic work-up of patients with stage III
clear, although pre- and post-operative imaging melanoma. Ann Surg. 2012;255:771–6.
provides important information for treatment plan- 17. Wagner T, Meyer N, Zerdoud S, Julian A, Chevreau C, Payoux P,
Courbon F. Fluoro-deoxyglucose positron emission tomography
ning and multidisciplinary cooperation for optimal fails to detect distant metastases at initial staging of melanoma
care. patients with metastatic involvement of sentinel lymph node. Br J
Dermatol. 2011;164:1235–40.
18. Wagner JD, Schauwecker D, Davidson D, Logan T, Coleman JJ 34. Singnurkar A, Wang J, Joshua AM, Langer DL, Metser U.

III, Hutchins G, et al. Inefficacy of F-18 fluorodeoxy-D-glucose- 18
F-FDG-PET/CT in the staging and management of melanoma: a
positron emission tomography scans for initial evaluation in early- prospective multicenter Ontario PET registry study. Clin Nucl Med.
stage cutaneous melanoma. Cancer. 2005;104:570–9. 2016;41:189–93.
19. Bikhchandani J, Wood J, Richards AT, Smith RB. No benefit in 35. Forschner A, Olthof S-C, Gückel B, Martus P, Vach W, la Fougère
staging FDG-PET in cN0 head and neck cutaneous melanoma. C, et al. Impact of 18F-FDG-PET/CT on surgical management in
Head Neck. 2014;36:1313–6. patients with advanced melanoma: an outcome based analysis. Eur
20. Decision Memo for Positron Emission Tomography (FDG) for J Nucl Med Mol Imaging. 2017;44:1312–8.
Solid Tumors (CAG-00181R4). June 2013. www.cms.gov 36. Krug B, Crott R, Roch I, Lonneux M, Beguin C, Baurain JF, et al.
21. Jimenez-Requena F, Delgado-Bolton RC, Fernandez-Perez C,
Cost-effectiveness analysis of FDG PET-CT in the management
Gambhir SS, Schwimmer J, Perez-Vazquez JM, et al. Meta-analysis of pulmonary metastases from malignant melanoma. Acta Oncol.
of the performance of 18F-FDG PET in cutaneous melanoma. Eur J 2010;49:192–200.
Nucl Med Mol Imaging. 2010;37:284–300. 37.
Dalrymple-Hay MJ, Rome PD, Kennedy C, Fulham M,
22. Frary EC, Gad D, Bastholt L, Hess S. The role of FDG-PET/CT in McCaughan BC. Pulmonary metastatic melanoma – the survival
preoperative staging of sentinel lymph node biopsy positive mela- benefit associated with positron emission tomography scanning.
noma patients. EJNMMI Res. 2016;6:73. https://doi.org/10.1186/ Eur J Cardiothorac Surg. 2002;21:611–4.
s13550-016-0228-1. 38. Ugurel S, Rohmel J, Ascierto PA, Flaherty KT, Grob JJ, Hauschild
23. Rodriguez Rivera AM, Alabbas H, Ramjaun A, Meguerditchian A, et al. Survival of patients with advanced metastatic melanoma:
AN. Value of positron emission tomography scan in stage III cuta- the impact of novel therapies. Eur J Cancer. 2016;53:125–34.
neous melanoma: a systematic review and meta-analysis. Surg 39. Wong AN, McArthur GA, Hofman MS, Hicks RJ. The advantages
Oncol. 2014;23:11–6. and challenges of using FDG PET/CT for response assessment in
24. Danielsen M, Kjaer A, Wu M, Martineau L, Nosrati M, Leong melanoma in the era of targeted agents and immunotherapy. Eur J
S, et al. Prediction of positron emission tomography/computed Nucl Med Mol Imaging. 2017;44(Suppl 1):67–77.
tomography (PET/CT) positivity in patients with high-risk primary 40. Parmenter TJ, Kleinschmidt M, Kinross KM, Bond ST, Li J,

melanoma. Am J Nucl Med Mol Imaging. 2016;6:277–85. Kaadige MR, et al. Response of BRAF-mutant melanoma to BRAF
25. Wasif N, Etzioni D, Haddad D, Gray RJ, Bagaria SP, Pockaj
inhibition is mediated by a network of transcriptional regulators of
BA. Staging studies for cutaneous melanoma in the United States: glycolysis. Cancer Discov. 2014;4:423–33.
a population-based analysis. Ann Surg Oncol. 2015;22:1366–70. 41. Tavaré R, Escuin-Ordinas H, Mok S, McCracken MN, Zettlitz KA,
26. Koskivuo I, Kemppainen J, Giordano S, Seppänen M, Veräjänkorva Salazar FB, et al. An effective immuno-PET imaging method to
E, Vihinen P, Minn H. Whole body PET/CT in the follow-up of monitor CD8-dependent responses to immunotherapy. Cancer Res.
asymptomatic patients with stage IIB-IIIB cutaneous melanoma. 2016;76:73–82.
Acta Oncol. 2016;55:1355–9. 42. Guldbrandsen KF, Hendel HW, Langer SW, Fischer BM. Nuclear
27. Baker JJ, Meyers MO, Frank J, Amos KD, Stitzenberg KB, Ollila molecular imaging strategies in immune checkpoint inhibi-
DW. Routine restaging PET/CT and detection of initial recurrence tor therapy. Diagnostics. 2017;7 https://doi.org/10.3390/
in sentinel lymph node positive stage III melanoma. Am J Surg. diagnostics7020023.
2014;207:549–54. 43. Wong VK, Lubner MG, Menias CO, Mellnick VM, Kennedy TA,
28. Horn J, Lock-Andersen J, Sjostrand H, Loft A. Routine use of Bhalla S, Pickhardt PJ. Clinical and imaging features of noncutane-
FDG-PET scans in melanoma patients with positive sentinel node ous melanoma. AJR Am J Roentgenol. 2017;208:942–59.
biopsy. Eur J Nucl Med Mol Imaging. 2006;33:887–92. 44. Murphy G, Hussey D, Metser U. Noncutaneous melanoma:

29. Bichakjian CK, Halpern AC, Johnson TM, Foote Hood A, Grichnik is there a role for 18F-FDG PET-CT? Br J Radiol. 2014;87:
JM, Swetter SM, et al. Guidelines of care for the management of 20140324.
primary cutaneous melanoma. American Academy of Dermatology. 45. Duncan JR, Carr D, Kaffenberger BH. The utility of positron

J Am Acad Dermatol. 2011;65:1032–47. emission tomography with and without computed tomography
30. Taghipour M, Marcus C, Sheikhbahaei S, Mena E, Prasad S, Jha in patients with nonmelanoma skin cancer. J Am Acad Dermatol.
AK, et al. Clinical indications and impact on management: fourth 2016;75:186–96.
and subsequent posttherapy follow-up 18F-FDG PET/CT scans in 46. Belkin DA, Wysong A. Radiographic imaging for skin cancer.
oncology patients. J Nucl Med. 2017;58:737–43. Semin Cutan Med Surg. 2016;35:42–8.
31. Schröer-Günther MA, Wolff RF, Westwood ME, Scheibler FJ,
47. MacFarlane D, Shah K, Wysong A, Wortsman X, Humphreys

Schürmann C, Baumert BG, et al. F-18-fluoro-2-deoxyglucose TR. The role of imaging in the management of patients with non-
positron emission tomography (PET) and PET/computed tomog- melanoma skin cancer: diagnostic modalities and applications. J
raphy imaging in primary staging of patients with malignant Am Acad Dermatol. 2017;76:579–88.
melanoma: a systematic review. Syst Rev. 2012;1:62. https://doi. 48. Humphreys TR, Shah K, Wysong A, Lexa F, MacFarlane D. The
org/10.1186/2046-4053-1-62. role of imaging in the management of patients with nonmelanoma
32. Mohr P, Eggermont AM, Hauschild A, Buzaid A. Staging of cuta- skin cancer: When is imaging necessary? J Am Acad Dermatol.
neous melanoma. Ann Oncol. 2009;20(Suppl 6):vi14–21. 2017;76:591–607.
33. Howard JH, Thompson JF, Mozzillo N, Nieweg OE, Hoekstra 49. Hawryluk EB, O’Regan KN, Sheehy N, Guo Y, Dorosario A,

HJ, Roses DF, et al. Metastasectomy for distant meta- Sakellis CG, et al. Positron emission tomography/computed tomog-
static melanoma: analysis of data from the first Multicenter raphy imaging in Merkel cell carcinoma: a study of 270 scans in 97
Selective Lymphadenectomy Trial (MSLT-I). Ann Surg Oncol. patients at the Dana-Farber/Brigham and Women’s Cancer Center.
2012;19:2547–55. J Am Acad Dermatol. 2013;68:592–9.
Therapy in Hemato-oncology 26
Paola Anna Erba, Martina Sollini, Roberto Boni,
and Sara Galimberti
Contents
26.1 Lymphomas 655
26.1.1 Hodgkin’s Lymphoma 656
26.1.2 Non-Hodgkin’s Lymphomas 658
26.1.3 Multimodality Imaging in Lymphomas 661
26.1.4 Radioimmunotherapy of Lymphomas 672
26.2 Multiple Myeloma 674
26.2.1 Epidemiology and Etiology 674
26.2.2 Patients’ Workup 675
26.2.3 Treatment 675
26.2.4 Imaging for Staging Multiple Myeloma 676
26.2.5 Diagnostic Workup at Diagnosis 692
26.2.6 Assessment of Response to Treatment 692
26.2.7 Radionuclide Therapy of Multiple Myeloma 697
References 698
Learning Objectives • Understand the use of PET/CT imaging in the workout of

• Summarize the main features regarding epidemiology, patients with lymphoma and MM.
etiology, histopathology, and most common clinical • Summarize the basic modalities of performing, reading,
presentation of hematological malignancies. and interpreting PET/CT scan with [18F]FDG-PET/CT in
• Understand the use of PET/CT imaging in each subtype patients with lymphoma and MM.
of lymphoma and in MM. • Learn the basis for a clinical multidisciplinary
discussion of imaging results in lymphoma and MM.
P. A. Erba (*)
26.1 Lymphomas
Research and Advanced Technologies in Medicine, Lymphomas are hematologic malignancies characterized by
University of Pisa, Pisa, Italy abnormal proliferation of lymphoid cells, most commonly
e-mail: paola.erba@unipi.it
arising in lymph nodes, but potentially involving any organ
M. Sollini or tissue in the body.
Department of Biomedical Sciences, Humanitas University,
Milan, Italy
Several classification systems have been proposed for
lymphomas. Classification of a lymphoma under a certain
R. Boni
Nuclear Medicine Service, “Papa Giovanni XXIII” Hospital,
category can affect treatment and prognosis. Common fea-
Bergamo, Italy tures considered for classifying lymphomas generally
S. Galimberti
include (1) whether or not it is a Hodgkin lymphoma; (2)
Hematology Unit, Department of Clinical and Experimental whether the abnormally replicating cell is a T cell or B cell;
Medicine, University of Pisa, Pisa, Italy and (3) site where the malignant cell arises.

26.1.1 Hodgkin’s Lymphoma In lymphocyte-rich cHL (about 5% of cHLs), patients

usually present early-stage disease and predominant periph-
26.1.1.1 Epidemiology and Etiology eral adenopathy without bulky mediastinal involvement.
The first descriptions of what came to be known as Hodgkin’s Lymphocyte-depleted cHL is the rarest cHL subtype in
disease (or Hodgkin lymphoma, HL) date back to 1832, developed countries (<1% of cases). It is often associated
when the eminent British pathologist Thomas Hodgkin with HIV infection and is more aggressive than other cHL
described an autopsy case series of patients with lymphade- subtypes.
nopathy and splenic enlargement [1]. The HL cancer cells Overall, patients with lymphocyte-depleted cHL or
(that arise from germinal center or post-germinal center B mixed-cellularity cHL have a significantly worse prognosis
cells) form a minority of the tumor mass, most of which is than patients with NSCHL, patients with lymphocyte-rich
composed of a reactive inflammatory milieu comprising cHL having the best prognosis.
lymphocytes, eosinophils, neutrophils, histiocytes, and Typically, HRS cells in cHL express the CD30 antigen
plasma cells. Multinucleate giant cells or large mononuclear and, in about 80–90% of cases, also CD15 [3]. The malig-
cells, referred to as Hodgkin and Reed-Sternberg (HRS) nant lymphocytic and histiocytic cells in LPHL express the
cells, are pathognomonic of the disease. antigen CD20, but variable frequencies (20–35%) of CD20
HL accounts for approximately 10% of the newly expression on malignant RS cells in cHL have been reported
diagnosed lymphomas in the United States and 0.5% of [3]. Both cHL and NLPHL are monoclonal B-cell disorders
all newly diagnosed cancers and for about 5% of the lym- arising from germinal centers with different patterns of sur-
phoma-related deaths per year—although the death rate face antigen expression [3]. NLPHL is generally a much
is declining [2]. Across all stages at diagnosis, the 5-year more indolent lymphoma, usually asymptomatic, and always
survival of patients with HL is currently 70–85% (https:// EBV-negative. The malignant cells in NLPHL are CD15-
seer.cancer.gov/csr/). Median age at diagnosis is negative, CD30-negative, CD45-positive germinal center B
39 years, although cases in the 20–34-year range make cells and invariably express CD20. They exhibit a nodular
up almost one-third of new diagnoses; there is a small growth pattern and popcorn-like, lymphocytic-histiocytic
prevalence in men (56% of newly diagnosed cases) ver- malignant cells without important fibrosis.
sus women [2].
Although etiology of HL is not well understood, some 26.1.1.3 C linical Presentation, Evaluation,
factors have been associated with a higher risk, such as the Staging, and Prognosis
industrial development, socioeconomic status, immune sup- HL most commonly presents with painless lymphadenopa-
pression, age, Epstein-Barr herpes virus infection, and thy enlarging over months, especially in the supraclavicular
familiar history [3]. Immunosuppression (especially if due and low-neck lymph nodes. Enlargement of lymph nodes in
to HIV infection) increases the risk of HL [4]. The inci- the mediastinum and/or neck is seen in about 60% of patients.
dence of HL increases also after solid organ transplantation Other sites include splenic, axillary, abdominal, hilar, or
and in patients with autoimmune conditions, such as rheu- inguino-femoral lymph nodes. Mediastinal masses can grow
matoid arthritis, systemic lupus erythematosus, and quite large before a diagnosis is made; bulky disease (trans-
sarcoidosis. verse diameter exceeding 10 cm) confers poor prognosis.
Approximately 30% of patients also have systemic “B”
26.1.1.2 Pathology and HL Subtypes symptoms such as fevers, chills, night sweats, or unexplained
There are two major subclasses of HL, classical HL (cHL, weight loss >10% of body weight. Symptoms are more fre-
95% of the cases) and nodular lymphocyte-predominant HL quent in patients with advanced-stage or bulky disease and
(NLPHL, 5% of the cases). The HRS cells in cHL and lym- confer poor prognosis, similarly as leukocytosis/neutrophilia
phocytic and histiocytic cells in NLPHL represent only and anemia.
1–5%, of the tumor mass. Four variants of cHL have been The Ann Arbor staging system with Cotswolds modifica-
defined: nodular sclerosing, mixed cellularity, lymphocyte- tion has been in use since 1989 but includes somewhat obso-
rich, and lymphocyte-depleted [5]. lete/invasive procedures, such as liver biopsy, laparotomy,
Nodular sclerosing cHL (NSCHL) accounts for about and bone marrow biopsy. In 2014, the Lugano classification
70% of cHL cases in the developed world. Mediastinal ade- updated staging for lymphomas, and [18F]FDG-PET/CT was
nopathy is seen in 80% of cases, and bulky lymph nodes in formally incorporated into standard staging. Treatment is
about 50% of patients [3]. Association with EBV infection is based on classification of patients as having limited (stages I
less frequent. and II, non-bulky) or advanced (stage III or IV) disease;
In mixed-cellularity cHL (MCCHL, 20–25% of cHL in stage II bulky disease is stratified as limited or advanced dis-
developed countries), the association with EBV infection ease based on histology and other prognostic factors
(75% of the cases) is more frequent than in NSCHL. (Table 26.1) [6, 7]. Generally, unfavorable disease is
26 Hybrid Imaging and Radionuclide Therapy in Hemato-oncology 657
c onsidered with larger tumor burden (either bulky mediasti- 0–19 years is 96.4% and 89.8% for those diagnosed at ages
nal adenopathy, more nodal groups involved), older age, and 20–64 years (2007–2013 Surveillance, Epidemiology, and
“B” symptoms. Approximately 65% of patients with early- End Results data) (https://seer.cancer.gov/csr/).
stage HL have unfavorable disease. A combination regimen with doxorubicin (Adriamycin),
The International Prognostic Score (IPS) is commonly bleomycin, vinblastine, and dacarbazine (ABVD) has
used for patients with advanced-stage III/IV disease become the standard therapy regimen for cHL in the United
(Table 26.2) [7]. The 5-year progression-free survival States. Patients with early-stage cHL treated with this com-
decreases from 84% for patients with an IPS of 0–42% for bination have 5-year survival rates >90%.
those with an IPS of 5–7 [8]. NLPHL is more prevalent in Two cycles of ABVD followed by 20 Gy of involved-
men, with a strong familial risk [9]. Almost 80% of newly field radiation is currently the optimal approach to deliver
diagnosed NHPHL patients present with stage I/II lymph- combined modality treatment in early-stage, favorable prog-
adenopathy and have an excellent long-term prognosis [10]. nosis cHL [12]. Other initial chemotherapy options are
Patients who have advanced-stage (stage III/IV) NLPHL Stanford V and bleomycin, etoposide, doxorubicin, cyclo-
have a worse prognosis than both those who have early-stage phosphamide, vincristine, prednisone, and procarbazine
NLPHL and those with advanced cHL [11]. (BEACOPP). A radiation-free approach is feasible for
patients with early-stage, favorable-risk HL who have dis-
26.1.1.4 Treatment ease at sites vulnerable to late radiation toxicities, such as
Chemotherapy and radiation are the mainstays of cHL treat- near the breast and heart [13].
ment. With current combination chemotherapies, the 5-year Imaging is used to optimize the radiation fields. Interim
relative survival for patients diagnosed with cHL at ages PET is also employed to define subgroups of patients who
Table 26.1 Revised staging of primary nodal lymphomas: Lugano classification

Limited disease
Stage I 1 LN region or lymphoid structure
Stage II ≥2 LN regions on same side of diaphragm
Advanced disease
Stage III LN regions on both sides of the diaphragm or above diaphragm with splenic involvement
Stage IV LN regions and noncontiguous extra-nodal sites
All stages A—no symptoms
(only HL) B—fever (>38 °C), night sweats, 10% body weight loss
Stages I, II E—involvement of extra-nodal site
Designation E is used for only patients with limited disease and not for advanced disease
For bulky disease, documenting the diameter is recommended and not the designation X
Ann Arbor Description
stage
I Involvement of a single lymphatic site (i.e., nodal region, Waldeyer’s ring, thymus, or spleen) (I); or localized involvement of a
single extralymphatic organ or site in the absence of any lymph node involvement (IE)
II Involvement of two or more lymph node regions on the same side of the diaphragm (II) or localized involvement of a single
extralymphatic organ or site in association with regional lymph node involvement with or without involvement of other lymph
node regions on the same side of the diaphragm (IIE)
III Involvement of lymph node regions on both sides of the diaphragm (III), which also may be accompanied by extralymphatic
extension in association with adjacent lymph node involvement (IIIE) or by involvement of the spleen (IIIS) or both (IIIE,S)
IV Diffuse or disseminated involvement of one or more extralymphatic organs, with or without associated lymph node
involvement; or isolated extralymphatic organ involvement in the absence of adjacent regional lymph node involvement, but in
conjunction with disease in distant site(s). Stage IV includes any involvement of the liver or bone marrow, lungs (other than by
direct extension from another site), or cerebrospinal fluid
Designations applicable to any stage
A No symptoms
B Fever (temperature >38 °C), drenching night sweats, unexplained loss of >10% of body weight within the preceding 6 months
E Involvement of a single extranodal site that is contiguous or proximal to the known nodal site
S Splenic involvement
Reproduced with permission from: Jacene HA, Tirumani SH, Wahl RL. Diagnostic applications of nuclear medicine: lymphomas. In: Strauss HW,
Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical Applications. New York, NY: Springer; 2017:
353–94, and from: Hodgkin and non-Hodgkin lymphomas. In: Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A III, Eds. AJCC
Cancer Staging Manual. 7th ed. New York, NY: Springer, 2010, pp 607–11
Table 26.2 International Prognostic Score (IPS) and Indices (IPI) for advanced cHL) is refractory to or relapses after initial ther-
HL and NHL apy. Single-agent chemotherapy, local irradiation, or reduced
Factorsa Risk group intensity of chemotherapy followed by allogeneic stem cell
IPS Male gender Risk increasing with each transplant can be considered in these patients, though most
increase in number of adverse
factors
often with palliative intent only.
Age >45 Novel approaches use antibody-drug conjugates or immu-
Stage IV notherapy, while promising results have been reported with
Hemoglobin <10.5 g/L other novel agents such as immune checkpoint inhibitors,
WBC count e.g., anti-programmed death receptor-1 (PD-1) monoclonal
>15 × 109/L antibodies and allogeneic stem cell transplantation (ASCT).
Lymphocyte count Radical surgical excision without adjuvant therapy has excel-
<0.6 × 109/L or <8%
WBC lent efficacy in early-stage disease. Non-bulky, early-stage
Serum albumin <40 g/L (I/II) lpHL without B symptoms is typically treated with
IPI Age >60 Low: 0–1 radiation alone [14]. Current guidelines from the International
Elevated serum LDH Low intermediate: 2 Lymphoma Radiation Oncology group recommend radiation
ECOG performance High intermediate: 3 therapy of involved lymph nodes or involved sites using
status ≥2 30–36 Gy to treat early-stage LPHL [13]. Patients with
Ann Arbor clinical High: 4–5
NLPHL have an increased risk of transformation to aggres-
stage III or IV
>1 involved extra-nodal
sive, diffuse large B-cell lymphoma, an aggressive NHL.
disease sites Other early and advanced stages may be treated with combi-
Revised Same as IPI Very good: 0 nations of chemotherapy, radiation, and/or rituximab.
IPI Good: 1–2 Multiagent combinations like ABVD and CHOP (cyclophos-
Poor: ≥3 phamide, doxorubicin, vincristine, and prednisone) in com-
FLIPI Age >60 Low: 0–1 bination with rituximab are the most commonly used
Elevated serum LDH Intermediate: 2 chemotherapeutic regimens in the United States for symp-
Ann Arbor clinical High: ≥3
tomatic patients with advanced LPHL [15].
stage III or IV
>4 involved nodal
disease sites 26.1.1.5 Follow-Up After Therapy
Hemoglobin level Generally, history, physical exam, and laboratory evaluations
<12.0 g/dL are performed at increasingly spaced intervals up to 5 years
Reproduced with permission from: Jacene HA, Tirumani SH, Wahl after completion of initial treatment. NCCN guidelines rec-
RL. Diagnostic applications of nuclear medicine: lymphomas. In: ommend very limited imaging (i.e., CT) once during the first
Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
Oncology – From Pathophysiology to Clinical Applications. New York, 12 months and then as clinically needed. The use of [ F]
18
NY: Springer; 2017: 353–94 FDG-PET/CT in follow-up is discussed further below.

WBC white blood cell, LDH lactate dehydrogenase, FLIPI IPI for fol- Increased risk of death persists for more than 25 years in
licular NHL patients with HL, due to late-term treatment complications
a
One point given for each factor present
rather than to HL itself. Strategies using interim PET for
risk-adapted therapy may guide dose intensification in
would benefit from the omission of radiotherapy after short- patients who are more likely to relapse and dose de-
duration ABVD (see below). intensification in those responding well.
Four cycles of MOPP-ABV chemotherapy combined
with involved-field radiotherapy has become the standard of
care for patients with unfavorable-risk cHL. NCCN guide- 26.1.2 Non-Hodgkin’s Lymphomas
lines recommend escalated BEACOPP (which is more
widely used in Europe) as an option for advanced HL with 26.1.2.1 Epidemiology and Etiology
IPS >4, although this regimen is not recommended for NHL (the sixth most common cancer in men and the fifth in
elderly patients with advanced HL because of acute women) ranks as the ninth and sixth most common cause of
toxicity. cancer deaths for men and women, respectively [2]. Although
Treatment for advanced cHL has shifted from using the incidence of NHL has been rising, the death rate is
higher doses and intensity of cytotoxic agents to a risk- decreasing. Most cases of NHL occur in individuals older
adapted approach, based on the result of PET/CT (see than 60 years. Immune suppression (e.g., post-transplant)
below). and immune stimulation (e.g., Sjogren’s syndrome,
Most patients with cHL are cured by first-line therapy, but rheumatoid arthritis) have been associated with increased
about one-third of patients with HL (especially those with risk of NHL. Viral-lymphoma associations have also been
suggested. Environmental exposures linked to NHL include two groups: those of B-cell origin and those of T-cell/natural
organochlorine-based pesticides, organic solvents, and wood killer (NK)-cell origin.
products.
Follicular Lymphoma
26.1.2.2 Pathology and NHL Subtypes Follicular lymphoma (FL) is one of the most common B-cell
Although initial classifications of NHL were based primarily lymphomas (15–20%). Most patients have an indolent clini-
on morphology, newer immunologic, molecular, and genetic cal course, slowly progressing over many years with a his-
knowledge has led to update and revise the classification sys- tory of waxing and waning adenopathy. Median survival is
tems [16]. 10–14 years under current therapeutic regimens. In a small
The three most commonly used classifications for NHL fraction of patients, the disease progresses rapidly in less
are the National Cancer Institute’s Working Formulation than 1 year and/or transforms into aggressive diffuse lym-
(IWF), the Revised European-American Classification of phoma with poor outcome. Due to the insidious and often
Lymphoid Neoplasms (REAL), and the WHO classification. asymptomatic features of FL growth, patients are frequently
The Working Formulation, originally proposed in 1982, clas- diagnosed at an advanced stage. FL currently remains virtu-
sified lymphomas according to morphology and clinical ally incurable or at best therapeutically converted into a
behavior (i.e., low, intermediate, or high grade) with ten sub- chronic disease—with possible toxicity issues for long-term
groups from A to J. Although considered obsolete, this clas- survivors.
sification is still used mainly for historical data comparisons. FL results from the malignant transformation of GC-like
In 1994, the REAL classification applied immunophenotypic B cells blocked in their capacity to further differentiate into
and genetic features to identify distinct clinico-pathologic memory and plasma cells. The t(14;18)(q32;q21) transloca-
NHL entities. The WHO classification, first introduced in tion (>85% of cases), the hallmark and most recurrent fea-
2001 and updated in 2008, further elaborates upon the REAL ture of FL, occurs in bone marrow pre-B cells and leads to
approach (Table 26.3). This classification divides NHL into BCL2 overexpression. Mutations are acquired throughout
years, if not decades, since an FL precursor already “com-
mitted” to originate FL can be present in bone marrow and/
Table 26.3 World Health Organization classification of NHL
or blood even 10–20 years before diagnosis [17].
B-cell lymphomas T/NK-cell lymphoma
Precursor B lymphoblastic Precursor T-cell lymphoblastic
lymphoma/leukemia lymphoma iffuse Large B-Cell Lymphoma
D
Mature B-cell lymphoma Mature T-cell lymphoma Diffuse large B-cell lymphoma (DLBCL) is one of the most
Chronic lymphocytic Adult T-cell lymphoma/ common and aggressive forms of B-cell lymphoma, account-
leukemia/small cell leukemia ing for 30–40% of newly diagnosed NHLs. The median age
lymphoma at diagnosis is in the seventh decade, but it may also occur in
Lymphoplasmacitic Mycosis fungoides (Sezary
young adults and children. DLBCL usually arises de novo
lymphoma syndrome)
Splenic marginal zone Primary cutaneous anaplastic
but can also derive from an indolent form of lymphomas
lymphoma large cell lymphoma such as CLL, FL, MZL, or HL. It may also be secondary to
MALT-lymphoma Anaplastic large cell various primary or acquired immunodeficiency disorders
lymphoma [18].
Normal marginal zone Peripheral T cell lymphoma,
lymphoma unspecified
Mantle cell lymphomas
Follicular lymphoma Angioimmunoblastic T-cell
lymphoma Mantle cell lymphoma (MCL) is an aggressive mature CD5+/
Mantle cell lymphoma Extranodal NK/T cell CD23− B-cell malignancy that accounts for 3–10% of NHLs.
lymphoma, nasal type Patients are typically men (about 2.5:1) with a median age at
Diffuse large B-cell Enteropathy-type T-cell diagnosis of 68 years; they usually present with extensive
lymphoma lymphoma disease, including widespread lymphadenopathy, bone mar-
Mediastinal (thymic) large Hepatosplenic T-cell
row involvement, splenomegaly, circulating tumor cells, and
B-cell lymphoma lymphoma
Intravascular large B-cell Subcutaneous panniulitis-like bowel infiltration. The disease remains incurable, although
lymphoma T cell lymphoma median overall survival has recently increased from 2.5 years
Primary effusion Blastic NK cell lymphoma to 5–7 years.
lymphoma
Burkitt’s lymphoma/leukemia hronic Lymphocytic Leukemia
C
Reproduced with permission from: Jacene HA, Tirumani SH, Wahl Chronic lymphocytic leukemia, most common in adults, is
RL. Diagnostic applications of nuclear medicine: lymphomas. In:
Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
characterized by the accumulation of CD5+/CD23+ neoplas-
Oncology – From Pathophysiology to Clinical Applications. New York, tic B cells in secondary lymphoid organs, bone marrow, and
NY: Springer; 2017: 353–94 peripheral blood. Median age at diagnosis is 71 years and
men are more frequently affected than women. The disease patients. Core or excisional biopsy is recommended. Initial
is clinically and biologically heterogeneous [19]. work-up is similar to that of HL. History and physical exam,
A recently proposed prognostic index, CLL-IPI, includes laboratory tests, and imaging are performed for staging and
five factors: age, clinical, serum β2-microglobulin level, assessing organ function prior to treatments.
IGHV mutational status, and TP53 aberrations. Patients are
divided into four risk groups: high, intermediate, low, and 26.1.2.4 Evaluation and Staging
very low risk [20]. The modified Ann Arbor staging system and the Lugano
classification are reported in Table 26.1. In the Lugano clas-
arginal Zone Lymphomas
M sification, four stages remain, but the subscript E is used for
Marginal zone lymphoma (MZL) comprises three entities: only when there is limited extranodal disease, and it is not
extranodal marginal zone lymphoma of mucosa-associated relevant for patients with advanced-stage disease. The desig-
lymphoid tissue (MALT lymphoma), splenic marginal zone nation “X” is replaced by largest tumor diameter for describ-
lymphoma(SMZL), and nodal marginal zone lymphoma ing bulky disease, and “B” symptoms are not considered [6].
(NMZL). Altogether, they account for about 10% of all lym- Staging is best performed with PET for NHL subtypes
phoid neoplasms, as follows: 7% MALT lymphoma, 2% that are typically [18F]FDG-avid (i.e., DLBCL), whereas
SMZL, and 1.5% NMZL. These lymphomas are associated indolent NHLs may sometimes have low metabolic activity.
with chronic microbial infection and/or autoimmune disor- Variability of [18F]FDG uptake and staging with PET in NHL
ders, suggesting a possible role of antigenic stimulation in are discussed further below. Bone marrow biopsy is no
their pathogenesis. For instance, gastric MALT lymphoma longer required for all patients with NHL for staging, but is
and SMZL may regress following treatment for Helicobacter reserved for those with DLBCL and a negative
pylori and HCV infection, respectively. Most of these MZLs [18F]FDG-PET/CT in the marrow or those with low-[18F]
present as an indolent disease and respond well to conven- FDG-avidity lymphomas [6].
tional therapies. A small proportion of these lymphomas,
particularly SMZL (10–20%), may show an aggressive 26.1.2.5 Prognosis
clinical course, which may be associated with high-grade Histopathologic subtype determines prognosis for patients
transformation. with NHL. Within each subtype of NHL, disease-specific
genetic alterations and molecular markers may provide addi-
T-Cell NHL tional information for predicting outcomes and guiding man-
T- and natural killer (NK)-cell lymphomas arise from agement. The IPI index (Table 26.2) [20] identifies four risk
T-lymphocytes and NK cells, respectively. To distinguish it groups with decreasing 5-year overall survival rates (73%,
from immature T-cell neoplasms such as lymphoblastic leu- 51%, 43%, 26%, respectively). The enhanced IPI (NCCNIPI)
kemia/lymphoma, mature T-cell lymphoma is often called was recently reported to discriminate low- and high-risk sub-
peripheral T-cell lymphoma (PTCL). PTCL accounts for groups better than IPI [21]. Most T-cell lymphomas are clini-
approximately 15% of all NHL cases worldwide. cally aggressive, and with a few exceptions (anaplastic large
Enteropathy-associated T-cell lymphoma (EATL) includes cell, early-stage T-/NK cell, nasal), most are incurable.
two subtypes: (1) type I EATL usually occurs following
long-standing celiac disease (CD), while (2) type II EATL 26.1.2.6 Treatment and Follow-Up
(also called monomorphic epitheliotropic intestinal T-cell Recommendations for detailed management strategies are
lymphoma) occurs without antecedent CD. These two highly provided in the NCCN guidelines. Rituximab alone or associ-
aggressive diseases show transmural growth, which often ated with chemotherapy improves overall response rates and
leads to GI bleeding or perforation, whereas another form PFS for patients with indolent NHL in both the first-line and
grows superficially along the GI mucosa. Indolent T-cell maintenance settings, with improved overall survival also in
lymphoproliferative disorders of the GI tract usually present the long term. Patients who respond to first-line therapy can
with chronic diarrhea, weight loss, and malnutrition, mim- be treated with maintenance rituximab up to 2 years.
icking the symptoms of inflammatory bowel disease. The Consolidation with radioimmunoconjugates in patients ini-
course of this disease is usually indolent, although some tially treated with chemotherapy alone improves the quality
cases of transformation have been reported. of remission, but not better than initial chemoimmunotherapy
alone. Novel anti-CD20 monoclonal antibodies are still under
26.1.2.3 Clinical Presentation investigation. Anthracycline-containing regimens CHOP or
Similar to HL, NHL most commonly presents as painless CHOP plus etoposide (CHOEP) are recommended as the
adenopathy but spreads in a manner less predictable than HL first-line treatments for PTCLs. Complete response (CR)
(more through the hematogenous route than by contiguous rates range 30–70%, and 5-year OS ranges from 20% to 60%.
nodal spread) and is more likely to involve extranodal sites Current consolidation autologous stem cell transplantation
than HL. Systemic “B” symptoms are found in 40% of (ASCT) is recommended in patients with chemosensitive
PTCL [22]. Follow-up for patients with NHL treated with burden to patients. Furthermore, the 2-day interval between
curative intent includes history and physical exam, laboratory injection of 67Ga-citrate and scanning is not clinically prac-
testing, and imaging employed with decreasing frequency tical. Currently, 67Ga-citrate has been largely replaced with
over the first 5 years and then annually. [18F]FDG-PET/CT, which offers higher-resolution images,
Enteropathy-associated T-cell lymphoma (EATL) higher sensitivity (particularly for lower-grade lympho-
mas), lower radiation doses, and a shorter time from injec-
tion until imaging [23, 24]. Bone scanning has also been
used for detection of lymphomatous involvement in bone;
Key Learning Points
however, these marrow lesions may be falsely negative on
• Non-Hodgkin lymphoma (NHL) represents a het-
the conventional bone scan with 99mTc-labeled
erogeneous group of malignancies of different biol-
bisphosphonates.
ogy and prognosis.
The first widely adopted National Cancer Institute-
• Each NHL subtype has distinct epidemiologies, eti-
sponsored Working Group response criteria were published
ologies, morphologic, immune-phenotypic, genetic,
in 1999 [25]. These guidelines standardized the definitions
and clinical features.
for complete (CR) and partial response (PR), stable, progres-
• Responses to therapy differ in different NHL
sive, and relapsed disease (RD). In addition, they incorpo-
subtypes.
rated the category of unconfirmed complete remission, first
• The NCCN guidelines for NHL provide general rec-
introduced in the Cotswolds Classification [26] to designate
ommendations on classification, differential diagno-
patients with a large tumor mass before treatment, with a
sis and supportive care, as well as specific guidance
persistent mass with size reduction following treatment
for management of the most common subtypes.
which, particularly in patients with HL and (DLBCL), may
• The ESMO has published separate guidelines for
be the result of tumor fibrosis, rather than residual, viable
management and treatment of follicular lymphoma,
disease.
mantle cell lymphoma, diffuse large B-cell lym-
phoma, primary cutaneous lymphoma, and gastric
26.1.3.1 [ 18F]FDG-PET for Detection
marginal zone lymphoma of the mucosa-associated
and Histologic Grading of Lymphoma
lymphatic tissue (MALT).
Most lymphomas accumulate [18F]FDG, although in general
• The three most commonly used classification
low-grade lymphomas exhibit lower [18F]FDG uptake than
schemes for NHL are the National Cancer Institute’s
high-grade lymphomas [27]. [18F]FDG-avid disease is
Working Formulation (IWF), the revised European-
observed in almost all patients with HL (94%) and aggres-
American Classification of Lymphoid Neoplasms
sive HNL (83%), whereas it is less frequent in patients with
(REAL), and the WHO classification.
small lymphocytic lymphoma (83%), enteropathy-type
• The IWF, originally proposed in 1982, classified
T-cell lymphoma (67%), extranodal marginal zone lym-
and grouped lymphomas by morphology and clini-
phoma (67%), and MALT marginal zone lymphoma (40%).
cal pattern (low, intermediate, or high grade) with
However, anatomic location of lymphomatous lesions may
ten subgroups labeled A to J.
interfere with this pattern, regardless of histopathologic sub-
• In 1994, the REAL classification attempted to apply
type. In particular, cutaneous lesions are most likely to be
immunophenotypic and genetic features in identify-
non-[18F]FDG-avid, accounting for about one-third of the
ing distinct clinico-pathologic NHL entities.
non-[18F]FDG-avid cases. The other most common locations
• The WHO classification, first introduced in 2001 and
of non-[18F]FDG-avid disease are the bowel, bone marrow,
updated in 2008, further elaborates upon the REAL
effusions, and mucosal surfaces [27]. It should also be noted
approach, by dividing NHL into those of B-cell origin
that the [18F]FDG-PET scan can be falsely negative in
and those of T-cell/natural killer (NK)-cell origin.
patients with low tumor burden.
• Although considered obsolete, the IWF classification
Although visual assessment using the Deauville criteria is
is still used mainly for historical data comparisons.
strongly recommended for assessing response to treatment in
all [18F]FDG-avid lymphoma histotypes (see below,
Table 26.4), there are concerns about the high rate of false-
26.1.3 Multimodality Imaging in Lymphomas positives and interobserver variability.
The standardized uptake value (SUV) is the most fre-
Imaging in lymphoma patients has evolved greatly over the quently used semiquantitative PET index for estimating
past several decades, from the obsolete lymphangiogram to tumor glucose metabolism. It is defined as the ratio of the
CT and MRI for morphologic, predominantly size-based decay-corrected [18F]FDG concentration in a volume of
evaluation. Scintigraphy with 67Ga-citrate is limited by poor interest (VOI) to the injected dose normalized to the patient’s
spatial resolution and by concerns regarding high radiation body weight. In addition to body weight, other corrections
Table 26.4 Deauville 5-point scale

1 No uptake Key Learning Points
2 Uptake ≤ mediastinal blood pool • [18F]FDG-PET/CT is currently the modality of
3 Uptake > mediastinal blood pool but ≤ liver choice for staging, restaging, and assessment of
4 Uptake moderate > liver treatment response in [18F]FDG-avid lymphoma.
5 Uptake markedly > liver and/or new lesions • [18F]FDG-avid disease is observed in almost all
X New areas of uptake unlikely to be related to lymphoma
patients with HL and aggressive HNL, and it’s less
Reproduced with permission from: Jacene HA, Tirumani SH, Wahl frequent in other lymphoma subtypes.
RL. Diagnostic applications of nuclear medicine: lymphomas. In:
Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear • Visual assessment using the Deauville criteria is
Oncology – From Pathophysiology to Clinical Applications. New York, strongly recommended for assessing response to
NY: Springer; 2017: 353–94 treatment in all [18F]FDG-avid lymphoma
histotypes.
can also be applied, e.g., correction for lean body mass [28]. • SUVmax can be used to identify with high likelihood
In a clinical trial, an SUV >10 identified with high likelihood aggressive NHLs and low-grade lymphomas, as
aggressive NHLs and excluded low-grade lymphomas with well as to identify transformation of a low-grade
81% specificity [29]. Also NLPHL is [18F]FDG-avid, lymphoma into a high-grade lymphoma.
although SUVs are generally lower than those observed in • Quantitative algorithms of image feature extraction
cHL. NLPHL patients also had lower SUVmax than patients that may characterize the intratumor heterogeneity
with T-cell/histiocyte-rich large B-cell lymphoma (mean of [18F]FDG uptake have been developed and are
SUVmax 6.9 versus 16.6). currently under investigation in clinical trials.
However, a substantial overlap in [18F]FDG uptake occurs,
as some low-grade tumors exhibit high [18F]FDG uptake.
Therefore, in an individual patient, the level of [18F]FDG
uptake is probably not an adequate single surrogate marker 26.1.3.2 [ 18F]FDG-PET/CT for Initial Staging
for histologic grading of lymphoma. Nevertheless, dispro- of Lymphoma
portionately high accumulation of [18F]FDG in a single [18F]FDG-PET is clearly superior to 67Ga-citrate scintigra-
lesion compared to the others in a given patient may indicate phy [23, 24] and improves primary staging of lymphoma as
the transformation of a low-grade lymphoma to a high-grade compared to CT [35]. For nodal status, this is primarily due
lymphoma in that given patient. [18F]FDG-PET can identify to the ability of [18F]FDG-PET to detect lymphoma in nor-
transformation of CLL to DLBCL, the so-called “Richter’s” mal-sized lymph nodes that would be considered benign by
transformation. It has also been shown that SUVmax ≥10 is CT size criteria alone. [18F]FDG-PET is also more accurate
inversely correlated with OS and identifies transformation in than CT in detecting extranodal involvement. The addition of
CLL patients. Similarly, an SUVmax >10 identifies transfor- [18F]FDG-PET to lymphoma staging procedures changes the
mation from FL to DLBCL, a subtype that requires a more stage (upstaged in most cases) impacting on patients’ man-
aggressive treatment approach than de novo FL [30]. agement in 8–48% of cases [36]. Although this has been
[18F]FDG-PET identified aggressive transformation better shown to be true also for low-grade follicular NHL, the role
than PET with the proliferation marker 3′-deoxy-3′-[18F]- of [18F]FDG-PET in staging other indolent NHLs (i.e., CLL/
fluorothymidine ([18F]FLT). SLL) may be more limited [37].
[18F]FDG-PET and PET/CT can also guide biopsies of The Lugano classification extends the use of [18F]FDG-
lesion with aggressive transformation, the optimal threshold PET/CT as the standard for staging to all [18F]FDG-avid his-
being SUVmax 14 in patients with indolent NHL [31]. tologies. CLL or SLL, lymphoplasmacytic lymphoma,
Quantitative algorithms of image feature extraction that marginal zone lymphomas, and mycosis fungoides remained
may characterize the intratumor heterogeneity of [18F]FDG excluded because of their variable [18F]FDG avidity. In the
uptake have been developed [32]. The term “radiomics” refers modified version of the Ann Arbor staging system, size of the
to a comprehensive evaluation of the entire tumor volume spleen to be considered splenomegaly was standardized at
using quantitative evaluation of tumor imaging characteristics 13 cm, and routine chest X-rays were no longer recom-
[32]. Aerts et al. reported that radiomics decoded a general mended. Figure 26.1 shows examples of the use of [18F]FDG
prognostic phenotype existing in multiple cancer types by for staging patients with lymphoma.
revealing associations with the underlying gene expression Bone marrow biopsy is no longer considered essential for
patterns [33]. Lartizien et al. used texture analysis to develop routine staging of HL and in most patients with DLBCL,
a decision system to perform automatic discrimination of except when [18F]FDG-PET is negative in DLBCL patients;
malignant from benign lesions based on [18F]FDG-PET imag- in such instance, it is important to ascertain if a discordant
ing in B-cell lymphoma or HL patients [34]. histology is present in the bone marrow [6]. This
more likely to detect bone marrow disease when >30% of the

marrow is involved (Fig. 26.2). False-negative [18F]FDG-
PET scans occur with <30% bone marrow involvement,
lower-grade histologies, primary non-[18F]FDG-avid nodal
disease. Often, the CT component of a PET/CT scan may
appear normal (in up to 50% of the cases) [40].
Misinterpretation of diffuse [18F]FDG uptake in the marrow
might be related to concurrent systemic illness or recent
administration of hematopoietic growth factors. Therefore,
careful interrogation of clinical history is mandatory to
reduce interpretation errors.
Key Learning Points

• Adding [18F]FDG-PET to lymphoma staging proce-
dures changes the stage (upstaged in most cases),
thus impacting on patients’ management in 8–48%
of cases.
• The Lugano classification extends the use of [18F]
FDG-PET/CT as the standard for staging to all [18F]
FDG-avid histologies.
• [18F]FDG-PET/CT is highly sensitive and specific
for the detection of bone marrow in HL and DLBCL;
therefore, bone marrow biopsy is used for standard
staging only when [18F]FDG-PET is negative in
DLBCL patients.
26.1.3.3 [ 18F]FDG-PET/CT for Assessing

Fig. 26.1 [18F]FDG-PET in a 65-year-old man with chronic lympho-
Response to Treatment
cytic leukemia/small lymphocytic lymphoma. There is mild to moder- In 1999 Juweid et al. [41] incorporated [18F]FDG-PET into
ate [18F]FDG uptake in bilateral cervical, axillary, para-aortic, iliac, and the lymphoma response criteria that were standard at that
inguinal lymph nodes consistent with indolent NHL (arrowheads). time. They noted that, in patients with DLBCL, long-term
Within the packet of lymph nodes in the left groin, there is a focus of
more intense [18F]FDG uptake (arrow, SUVmax = 10.6). Biopsy of this
outcome was similar regardless of whether the patient had
hypermetabolic lymph node revealed Richter’s transformation. Another CR or PR based on CT criteria, as long as the mass was
area of more moderate uptake, suspicious for early transformation, is not [18F]FDG-avid. Thus, unconfirmed complete remission
detectable in the right axilla (arrow) (reproduced with permission from: was eliminated as a response category. This observation,
Jacene HA, Tirumani SH, Wahl RL. Diagnostic applications of nuclear
medicine: lymphomas. In: Strauss HW, Mariani G, Volterrani D, Larson
along with a large body of data accumulated thereafter, led
SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical to develop the revised criteria for lymphoma assessment
Applications. New York, NY: Springer; 2017: 353–94) published in 2007 [42]. These criteria incorporated PET
into response assessment because of its superior sensitivity
and specificity compared with CT. The 2007 International
recommendation is based on several studies showing that Working Group recommendations were primarily designed
[18F]FDG-PET/CT detects involvement of bone marrow in for HL and DLBCL, due to the limited available data at the
HL with 54–74% sensitivity and 91–95% specificity [38]. time for other histologic subtypes. For the same reason, at
For DLBCL, the Lugano classification recommends bone that time PET had not yet been formally incorporated into
marrow biopsy only in patients without marrow involvement staging. Following a workshop held at the 11th International
on [18F]FDG-PET/CT. Indeed, although [18F]FDG-PET/CT Conference on Malignant Lymphoma (in Lugano,
might underdiagnose low-volume disease in DLBCL, it has Switzerland, June 2011) and a subsequent workshop at the
been consistently shown that [18F]FDG-PET or PET/CT has 12th International Conference on Malignant Lymphoma in
better diagnostic performance than bone marrow biopsy for 2013, the Lugano classification was developed to improve
diagnosing marrow involvement [39]. [18F]FDG-PET is both the staging and response criteria for HL and NHL [6].
Fig. 26.2 [18F]FDG-PET/CT performed for staging in a 34-year-old spleen (b, c, arrowheads). Based on these findings, the patient was
woman with newly diagnosed Hodgkin lymphoma. The diagnostic upstaged to stage IV (for a, b, and c: left, CT; center, PET; right, fused
contrast-enhanced CT scan obtained prior to the PET scan showed PET/CT) (reproduced with permission from: Jacene HA, Tirumani SH,
enlarged lymph nodes in the mediastinum extending to but not below Wahl RL. Diagnostic applications of nuclear medicine: lymphomas. In:
the aortic arch (not shown); this finding was consistent with stage II Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
disease. The PET/CT scan shows intense [18F]FDG uptake in the medi- Oncology – From Pathophysiology to Clinical Applications. New York,
astinal mass (a). In addition, there are numerous [18F]FDG-avid lymph NY: Springer; 2017: 353–94)
nodes below the diaphragm (b, arrows) and in the bone marrow and
For response assessment, the Lugano classification recom- baseline or as progressive disease if uptake is greater than at
mended the Deauville qualitative 5-point scale (that com- baseline. On end-of-therapy PET, a score of 4 or 5 is consid-
pares the tumor uptake to the mediastinal blood pool and/or ered as failure of treatment irrespective of the intensity of
liver uptake) (see Tables 26.4 and 26.5) for interpretation of uptake compared to baseline. Similarly, a score of 3 is usu-
[18F]FDG-PET [43] and considered as metabolic complete ally considered negative on both interim and end-of-therapy
remission those cases with a persistent mass that was no scans but as an inadequate response in risk-adapted trials
longer [18F]FDG-avid. Its prognostic value for interim [18F] evaluating de-escalation strategies. Classification of response
FDG-PET/CT assessment has been shown in both HL and based on CT for non [18F]FDG-avid lymphomas is similar to
NHL [44, 45]. the original IWG 1999 criteria or the revised IWG 2007
Four categories of response using [18F]FDG-PET/CT for criteria.
[ F]FDG-avid lymphomas and CT for non [18F]FDG-avid
18
Concurrent with publication of the Lugano classification,
lymphomas are possible. Interpretation of the score depends the consensus report from the imaging subcommittee of the
on timing of the scan and on the clinical scenario. A score of International Harmonization Project in Lymphoma on the
1 or 2 is interpreted as complete response on both interim acquisition and interpretation of PET scans that accompa-
and end-of-therapy scans. A score of 4 or 5 is considered a nied the revised IWC 2007 criteria [46] was also updated [7].
partial response on interim scans if uptake is less than at Figures 26.3 and 26.4 show two examples of [18F]FDG-PET
Table 26.5 Lugano criteria for response assessment in lymphoma

Response PET/CTa CT
Complete response Score 1, 2, 3 with or without residual Complete disappearance of all sites of disease or decrease in size of nodes to
mass in nodes or extra-nodal sites ≤1.5 cm in LDi
Partial response Score 4 or 5 with decreased uptake ≥50% decrease in PPD (single lesion) or SPD of six measurable target nodes
compared to baseline and residual and/or extra-nodal sites (multiple lesions); decrease in spleen size >50% in
masses of any size length beyond normal
No response/stable Score 4 or 5 with no obvious change <50% decrease in PPD (single lesion) or SPD of six measurable target nodes
disease in FDG uptake compared to baseline and/or extra-nodal sites (multiple lesions) and no criteria for progressive disease
Progressive disease Score 4 or 5 with increase in New or increased adenopathy (nodes 1.5 cm > LDi, PPD > 50% from nadir, LDi
intensity of FDG uptake compared or SDi increase from nadir by 0.5 cm for lesions ≤2 cm and by >1.0 cm for
to baseline or new foci of FDG lesions >2.0 cm), increase in splenic length by >50% of its prior increase from
uptake compatible with lymphoma baseline, or by at least 2 cm from baseline if no prior splenomegaly, new or
recurrent splenomegaly, new nodal or extranodal sites or increase of preexisting
nonmeasurable lesions
Reproduced with permission from: Jacene HA, Tirumani SH, Wahl RL. Diagnostic applications of nuclear medicine: lymphomas. In: Strauss HW, Mariani
G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical Applications. New York, NY: Springer; 2017: 353–94
PET/CT: The interpretation of the score depends on the timing of the scan and the clinical scenario. A score of 1 or 2 is interpreted as complete
response on both interim and end of therapy scans. A score of 4 or 5 is considered a partial response on interim scans if the uptake is less than
baseline or as progressive disease if the uptake is greater than baseline. On end of therapy PET, a score of 4 or 5 is considered as failure of treatment
irrespective of the intensity of uptake compared to baseline. Similarly, a score of 3 is usually considered negative on both interim and end of
therapy scans, but as an inadequate response in risk-adapted trials evaluating de-escalation strategies
CT: A measurable node should be 1.5 cm in longest transverse diameter (LDi) and a measurable extranodal site should be 1.0 cm in longest diameter.
The longest and shortest (SDi) diameters on the transverse plane of each lesion are multiplied to obtain product of the perpendicular diameters (PPD).
Six largest nodes as well as extra-nodal sites which are measurable in two dimensions should be identified from different body regions. The overall
disease burden at baseline is quantified by adding the product of the diameters of these six lesions to obtain the sum of the product of the diameter (SPD)
a
Deauville 5-point scale
a d
Fig. 26.3 [18F]FDG-PET/CT performed for staging in a 55-year-old of lymph node disease was a 1.5 × 1.2 cm left para-aortic lymph node
man with newly diagnosed diffuse large B-cell lymphoma. There is a with intense [18F]FDG uptake (c, arrows). However, bone marrow dis-
focus of intense [18F]FDG uptake corresponding to lytic changes in ease was found in the left rib (d, maximum intensity projection image,
the right T11 vertebral body (a). Biopsy of this lesion confirmed dif- arrow). Additional extranodal disease was found in a scrotal mass (d,
fuse large B-cell lymphoma; whereas, bone marrow biopsy was nega- maximum intensity projection image, arrowhead). The overall find-
tive for lymphomatous involvement, and accordingly the PET scan ings were consistent with stage IV disease (for a, b, and c: left, CT;
showed normal marrow uptake at the site of biopsy (b). The only site right, PET)
a b c
Fig. 26.4 [18F]FDG-PET/CT performed for staging in a 26-year-old arrows). Thus, the PET/CT scan downstaged the patient from stage III to
man with newly diagnosed Hodgkin lymphoma. A diagnostic contrast- stage II, and he received a short course of chemotherapy and radiotherapy
enhanced CT scan shows the large mediastinal mass (a, top) and a border- (for c: top, transaxial PET; bottom fused PET/CT) (reproduced with per-
line abnormal size celiac axis lymph node suspicious for stage III disease mission from: Jacene HA, Tirumani SH, Wahl RL. Diagnostic applica-
(a, bottom—arrow). The PET/CT scan shows abnormal uptake in the tions of nuclear medicine: lymphomas. In: Strauss HW, Mariani G,
mass and lymph nodes above the diaphragm (b, coronal section), but no Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology
abnormal uptake in the celiac axis node below the diaphragm (c— to Clinical Applications. New York, NY: Springer; 2017: 353–94)
scans obtained for assessment of response to treatment in cells within the mass from fibrosis/necrosis. The informa-
patients with lymphoma. tion gained from post-therapy [18F]FDG-PET also provides
prognostic information [47], as patients with PET-positive
disease post-therapy (Fig. 26.5) have lower 1- and 2-year
progression-free and overall survival rates (0–54%) than
Key Learning Points
those who are PET negative (83–95%) (Fig. 26.6), regard-
• PET/CT is currently included into the revised crite-
less of the presence of a residual mass on CT. Therefore, the
ria for lymphoma assessment because of its supe-
results of post-therapy [18F]FDG-PET scan have important
rior sensitivity and specificity compared with CT.
therapeutic implications, as PET positivity predicts subse-
• The Lugano classification is currently used in stag-
quent relapse.
ing and response criteria for HL and NHL.
However, a negative post-therapy PET scan does not
• The Lugano classification recommended the
exclude the presence of residual microscopic disease with
Deauville qualitative 5-point scale.
size below the resolution power of PET imaging that in the
• Classification of response based on CT for non [18F]
modern scanners is between 0.5 and 1.0 cm; lesions of this
FDG-avid lymphomas is similar to the original IWG
size contain about 108–109 tumor cells [48]. Patients who are
1999 criteria or the revised IWG 2007 criteria.
PET negative with a residual mass are more likely to recur
than those who are PET negative without residual mass.
Another crucial issue concerns the minimum interval after
26.1.3.4 P ost-therapy Response Assessment by chemotherapy recommended to avoid false-negative scans due
[18F]FDG-PET/CT to the early treatment effect of “stunning,” which is currently a
Residual masses on anatomic imaging in patients with lym- minimum of 10 days [49]. Longer and more variable times post-
phoma are common post-therapy, but do not always indicate external beam radiation have been suggested [50], while opti-
the presence of active malignancy. Instead, [18F]FDG-PET mal interval times post- radioimmunotherapy are being
has the advantage to distinguish functionally viable tumor investigated [51].
Fig. 26.5 [18F]FDG-PET/CT scans obtained pre- and post-therapy (six and Deauville 5-point scale, this outcome would be rated at score of 5.
cycles of R-CHOP) in a patient with diffuse large B-cell lymphoma. A positive post-therapy PET scan is associated with poor prognosis (for
The pre-therapy scan shows intense [18F]FDG uptake in a left upper a and b: left, CT; right, PET) (reproduced with permission from: Jacene
quadrant mass, adjacent nodules, and left paravertebral regions (a, top HA, Tirumani SH, Wahl RL. Diagnostic applications of nuclear medi-
row). The post-therapy scan shows persistence of a 2.3 cm nodule in the cine: lymphomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM,
left upper quadrant with intense [18F]FDG uptake (b, arrows), consis- Eds. Nuclear Oncology – From Pathophysiology to Clinical
tent with residual active lymphoma. In the new Lugano classification Applications. New York, NY: Springer; 2017: 353–94)
26.1.3.5 Interpretation of Interim

Key Learning Points Therapy [18F]FDG-PET/CT Scans
• [18F]FDG-PET/CT is able to distinguish function- Most risk-adapted therapy trials have employed qualitative
ally viable tumor cells within the residual mass analyses (i.e., visual scales) for interpreting interim PET
from fibrosis/necrosis. scans as positive or negative. Several scales have been
• The information gained from post-therapy [18F] proposed over the years [46, 52–54]. The Deauville criteria
FDG-PET/CT also provides prognostic are currently used in the interim setting.
information. When [18F]FDG uptake in tumor is much less or much
• A negative post-therapy PET/CT scan does not greater than background, this is a clear-cut negative or posi-
exclude the presence of residual microscopic disease tive case. However, the definition of [18F]FDG “minimal
with size below the spatial resolution. residual uptake” (MRU) and “background” (and thus the cut-
• The minimum interval after chemotherapy recom- off point for a positive versus a negative scan) in the interim
mended to avoid false-negative scans is currently setting are critical. MRU was first defined in 2000 as
10 days. low-grade uptake on posttreatment PET within an area of
previous disease, reported as likely inflammation but where
Fig. 26.6 [18F]FDG-PET/CT scans obtained pre- and post-therapy apy inflammation. In the new Lugano classification and Deauville
(sixcycles of R-CHOP) in a patient with diffuse large B-cell lym- 5-point scale, this outcome would be rated at score of 2 (for a and b:
phoma. The pre-therapy PET/CT scan shows intense [18F]FDG uptake left, CT; center, PET; right, fused PET/CT) (reproduced with permis-
in a right psoas mass (a, arrows), while the post-therapy scan is nega- sion from: Jacene HA, Tirumani SH, Wahl RL. Diagnostic applica-
tive for active lymphoma with [18F]FDG uptake equal to background tions of nuclear medicine: lymphomas. In: Strauss HW, Mariani G,
at the prior tumor site (b, arrows). Although residual microscopic dis- Volterrani D, Larson SM, Eds. Nuclear Oncology – From
ease cannot be excluded, a negative post-therapy [18F]FDG-PET scan Pathophysiology to Clinical Applications. New York, NY: Springer;
is associated with good prognosis even if a residual mass remains on 2017: 353–94)
the CT scan. The residual [18F]FDG uptake may be due to post-ther-
a small volume of malignancy cannot be excluded [55]. This 5-point scale (5-PS) among experts and between experts and
definition has been more recently modified. local readers [58].
Kasamon et al. [56] found that 56% of patients with Some uncertainty does remain in the optimal cutoff for
aggressive NHL had a positive interim PET scan using medi- modifying therapy based on interim PET. The use of semi-
astinal blood pool as a comparator. In the follow-up ECOG quantitative parameters such as SUV may be helpful in the
study (E3404), the rate of positive interim PET scans was interim setting to help reduce the variability, provided that
lower than expected, and reproducibility of interim PET PET is performed using standardized procedures. Changes in
interpretation was re-evaluated [52]. Only moderate repro- therapy based on interim therapy PET (except for progres-
ducibility among the experts was found for interpreting the sive disease) should currently be considered only within
interim scans (k = 0.44 for ECOG criteria and k = 0.50 for clinical trials.
London criteria). The sources of disagreement were lesions The current NCCN guidelines recommend interim
in the para-aortic region, bone, and spleen [52], considering PET/CT scan assessment with Deauville criteria in both
that the spleen and bone marrow can have considerably het- early-stage favorable and advanced-stage disease. In a large
erogeneous uptake post-therapy due to treatment effects. On retrospective multicenter study, the interim PET scans were
the other hand, Barrington et al. have reported very good positive as per Deauville criteria in 17% patients. At 3 years
agreement among four centers in Europe for interpretation of PFS was 28% for interim PET-positive patients compared to
interim PET using the London criteria (RATHL trial) [57]. 95% for PET-negative patients resulting in 73% sensitivity
Agreement was slightly lower for independent reads when and 94% specificity for predicting outcome. There was
the category of “uptake greater than mediastinal blood pool strong concordance among reviewers using the Deauville
but less than liver” was considered separately (from negative criteria [59].
or positive). The updated results of the RATHL study reached Multiple consensus reports using PET for response
similar conclusions. There was good agreement in the inter- assessment (interim and post-therapy) in lymphoma and
pretation of the interim PET scans using the Deauville other tumors stress the importance of obtaining high-quality,
standardized PET and CT scans on dedicated PET/CT sys- 26.1.3.6 I nterim [18F]FDG-PET/CT and
tems. Detailed discussions on how to obtain such scans are Risk-Adapted Therapy
found in reports by the National Cancer Institute [60] and the Numerous studies have confirmed the prognostic value of
Netherlands multicenter trial group [61] and reviewed in the interim PET in the management of patients with lympho-
revised IWC [51] and PET Response Criteria in Solid mas. Event-free survival rates are consistently lower in
Tumors (PERCIST) [62]. patients with positive interim PET scans (0–46%) than in
Despite improvements in standardization, visual assess- those with negative interim scans (62–98%), despite vari-
ment of imaging by the Deauville criteria has the pitfall of ability in the study designs, histologies, and treatments
potential suboptimal discrimination between response cate- employed. As previously discussed, the IPS for HL and IPI
gories, due to oversimplification. The inter-patient variabil- for NHL are both currently used to guide management
ity and intra-patient fluctuations of hepatic [18F]FDG uptake decisions. These two indices are based solely on pre-treat-
during therapy are also heavily debated [63]. Furthermore, ment parameters and do not consider the inherent “chemo-
the role of D-5PS is not as well defined in rare lymphoma sensitivity” of the tumor or the rate of tumor kill which are
subsets. Continuous metrics as provided by quantitative both related to outcome, parameters that can both be
analysis would be more suitable to assess changes in tumors assessed by PET.
over the course of the treatment, compared with traditional A negative PET scan at the end of treatment means
dichotomous response categorization. either that all cells died under the effect of the chemo-
Advanced software programs allow determination of met- therapy or that residual microscopic disease is beyond the
abolically active tumor volumes. Thus, MTV as a measure of spatial resolution of the scanner. The rationale of per-
the viable tumor fraction or TLG (the product of MTV and forming a PET scan for interim evaluation of response to
mean SUV within the volume) may reduce the rate of false- therapy (or soon post-therapy) is based on the concept
positive results, increase reproducibility, and maximize sta- that chemotherapy administered during each cycle is
tistical power, hence leading to better prediction of survival expected to kill the same fraction of cells (first-order
than the existing methods [64]. Some textural image features kinetics) [67]; therefore, a true-negative scan implies that
(e.g., entropy and maximum probability) perform better than the rate of chemotherapy killing is fast enough to achieve
SUVmax in predicting morphologic changes of radiotracer cure by the end of therapy, whereas a positive interim
uptake regions longitudinally. In a cohort of patients with therapy scan does not [48, 68].
bulky HL and NHL, it was observed that the combination of For risk-adapted therapy, the prognostic information
common prognostic factors with appropriately chosen tex- derived from an interim PET scan is used not only to pre-
tural and shape parameters evaluated on baseline PET/CT dict treatment outcome but also to improve therapeutic
improves the prediction of interim metabolic response in strategies based on the individual patient’s risk. Therefore,
bulky lymphoma [65]. a positive interim PET scan can justify intensification of
In another pilot study, heterogeneity of the intratumoral therapy to improve the chance of cure, whereas a negative
distribution of 111In-ibritumomab tiuxetan was correlated interim scan can justify to discontinue or to de-intensify
with tumor response in NHL patients [66], and pre-therapy therapy. This second option is particularly attractive in
SUVmax predicted tumor response to 90Y-ibritumomab tiux- early-stage HL, to reduce the long-term toxicities of ther-
etan therapy—at discordance with a prior report [51]. apy in this curable population. In this regard, the imple-
Although routine use of [18F]FDG-PET for monitoring mentation of PET reduces dramatically the number of
the response of low-grade follicular and other indolent lym- patients submitted to radiation therapy, with no reduction in
phomas is debatable, it is gaining growing acceptance in the cure rate [69]. On the other hand, a positive interim
clinical practice [6, 7]. [18F]FDG-PET/CT scan may justify treatment intensifica-
tion, as suggested by the results of different clinical inves-
tigations [70–72].
Key Learning Points
• Most risk-adapted therapy trials have employed
qualitative analyses (i.e., visual scales) for inter- Key Learning Points
preting interim PET scans as positive or negative. • PET has a prognostic significance in the manage-
• The Deauville criteria are currently used in the ment of patients with lymphomas.
interim setting. • For risk-adapted therapy, the prognostic informa-
• The current NCCN guidelines recommend interim tion derived from an interim PET scan is used to
PET scan assessment with Deauville criteria in both predict treatment outcome and to improve therapeu-
early-stage favorable and advanced-stage disease. tic strategies based on the individual patient’s risk.
26.1.3.7 C linical Trials of Risk-Adapted Therapy 26.1.3.8 T reatment Response After

for Aggressive NHL Immunotherapy
Early enthusiasm for interim [18F]FDG-PET in predicting Immunotherapy, most notably with checkpoint inhibitors,
outcome in DLBCL has been tempered by considerable dis- has remarkable efficacy in a variety of solid tumors, includ-
crepancies in the results of different studies, reflecting vari- ing hematologic malignancies [79]. A flare reaction occurs in
ability in patient populations, therapy, use of rituximab, about 15% of patients, which might be interpreted as pro-
equipment, and image interpretation [73, 74]. Whereas the gressive disease and thus lead to premature discontinuation
negative predictive value is high in most studies, the positive of potentially effective therapy. In this regard, immune
predictive value is disappointingly low, most likely as a result response criteria developed for solid tumors [80] may not be
of false-positive results caused by inflammation and tumor considered suitable for lymphomas [81], which often have
necrosis [75]. The crucial issue therefore remains whether or disease that is not detectable in the CT scan (e.g., bone and
not altering treatment on the basis of an interim PET can bone marrow infiltration, soft tissue involvement). Moreover,
favorably impact outcome. According to currently available discordant results between [18F]FDG-PET and CT may be
clinical evidence, biopsy is recommended to confirm interim encountered with a residual CT lesion (i.e., PR by immune
positive PET scan findings prior to treatment intensification response criteria) that is no longer [18F]FDG-avid (i.e., CR
[75]. Based on the results obtained so far, it can be concluded by Lugano criteria). Such potential clinically significant
that the poor outcome of patients with interim PET positivity problems led to a workshop co-sponsored by the Lymphoma
despite intensification of therapy indicates innate resistance Research Foundation and the Cancer Research Institute,
to treatment; therefore, until further evidence is available, which resulted in the elaboration of Lymphoma Response to
treatment modification in DLBCL patients based on early Immunomodulatory Therapy Criteria [82]. These recom-
PET should remain restricted to clinical trials [76]. mendations use a provisional term of indeterminate response
The current NCCN guidelines also state that treatment to categorize the different types of flare reactions. Patients
modification of DLBCL patients should not be guided by suspected of a flare reaction may remain on treatment until
interim PET scan. Any modifications should be performed progressive disease is confirmed by biopsy or by continued
only after confirming real positivity on repeat biopsy of growth or evidence of deterioration. In the future, circulating
residual masses [77]. Demonstration of true benefit of ther- DNA assays may also be useful in distinguishing flare reac-
apy intensification for those with a positive interim PET scan tions from progressive disease in NHL patients on immuno-
is limited by comparison to historical controls and the lack of therapeutic agents such as checkpoint inhibitors [83].
randomization to standard versus experimental therapy. However, these criteria must be validated in large imaging
Whereas the use of the D5PS has improved standardiza- data sets.
tion of [18F]FDG-PET/CT interpretation and comparability
among studies, a number of critical issues still remain. To
improve on the discrimination power of [18F]FDG-PET/CT, Key Learning Points
more quantitative methods of interpretation have been • Immunotherapy may be associated with a flare met-
explored. Lin et al. [78] retrospectively analyzed interim abolic reaction (associated with enhanced [18F]FDG
scans in patients treated for DLBCL. Their observations sug- uptake) in about 15% of patients.
gested that patients with an SUVΔ reduction of greater than • Immune response criteria developed for solid tumors
two-thirds experienced a greater improvement in PFS than may not be considered suitable for lymphomas.
those with lesser reduction, in part resulting from a reduction • Specific Lymphoma Response to Immunomodulatory
in the number of false-positive studies associated with visual Therapy Criteria have been developed.
assessment. In HL, Kostakoglu et al. demonstrated the ben- • In these criteria, patients suspected of a flare reac-
efit of combining [18F]FDG-PET with reduction in tumor tion may remain on treatment until progressive dis-
size determined from contrast-enhanced CT studies [54]. ease is confirmed by biopsy or by continuing growth
or until evidence of deterioration.
• The Lymphoma Response to Immunomodulatory
Key Learning Points Therapy Criteria are now under validation in large
• The current NCCN guidelines state that treatment clinical trials.
modification of DLBCL patients should not be
guided by interim PET/CT scan: biopsy is recom-
mended to confirm interim positive PET/CT scan
findings prior to treatment intensification. 26.1.3.9 [ 18F]FDG-PET/CT for External Beam
• The use of the D5PS has improved standardization Radiation Therapy Treatment Planning
of [18F]FDG-PET/CT interpretation and compara- External beam radiation therapy (EBRT) is an important
bility among studies. component of treatment for lymphomas both for primary
treatment and for treatment of relapsed/refractory disease or
for palliative care. However, since exposure of normal tis- lungs, heart, heart valves, and coronary arteries [88–90]. To
sues to radiation may cause serious long-term side effects, perform RT planning using optimal imaging, the pre-
the volume of normal tissue that is exposed to radiation must chemotherapy staging PET/CT scan should also be per-
be minimized. Modern radiotherapy (RT) is highly “confor- formed in DIBH to allow image fusion with the
mal,” that is, the volume receiving the prescribed radiation post-chemotherapy DIBH planning CT scan. Respiration,
dose conforms almost precisely to the target volume that is both during imaging and during RT, is monitored by the
contoured [84]. Except for palliation purposes, the target available respiration monitoring methods. The patient’s
volume must be very accurately defined to create highly con- inspiration level is displayed to the patient using a screen or
formal treatment plans where only the defined volume with video display glasses compatible with the scanner and the
very narrow margins will receive the prescribed radiation treatment machine. Therefore, along with the whole-body
dose. PET/CT scans can be used for EBRT planning, as sev- PET/CT scan in free breathing, additional PET/CT scan over
eral studies have demonstrated significant changes in man- the mediastinal region should be acquired in DIBH (one bed
agement and RT volume definition with the addition of PET position).
information to CT information [85–87].
In patients with early-stage HL, [18F]FDG-PET helped
pinpoint undetected involved lymph nodes in 70% of Key Learning Points
patients, resulting in an increased of the target volume in • Use of PET-derived volumes for RT planning is
65% of cases and thus enabling far better precision in imple- advantageous since they yield more accurate defini-
menting highly conformal RT [85]. In another study, [18F] tion of the extent and location of the lymphoma
FDG-PET/CT modified the target volume in 25% of patients, lesions than morphological imaging based on radio-
as the RT plan defined with only CT information had inade- logical techniques.
quate coverage of the target defined by [18F]FDG-PET [86]. • Some technical issues are crucial in this clinical
Optimal use of the PET scan for defining and contouring setting.
the target volume for RT requires spatial co-registration of
the PET scan to the planning CT scan using a combined
PET/CT scanner. The PET scan must therefore be acquired
with the patient in the treatment position, on a flat couch top, 26.1.3.10 S urveillance [18F]FDG-PET/CT
with the use of appropriate immobilization devices, and with for Lymphoma
skin markers that are visible in the images. Contrast-enhanced Although the current literature on the use of [18F]FDG-PET
CT is important to differentiate lymph nodes from vessels. for follow-up of patients with lymphoma is limited, it does
For patients with abdominal and pelvic lymphoma involve- confirm the ability of PET to detect relapse before the patient
ment, oral contrast is also used to differentiate lymphoma becomes symptomatic and before anatomic imaging becomes
masses from bowel. positive [91]. PET is also useful for detecting relapse in the
Use of PET-derived volumes for RT planning is advanta- setting of a persistent, unchanging residual mass on CT dur-
geous since they yield more accurate definition of the extent ing follow-up [92]. For patients who achieve CR after ther-
and location of the lymphoma lesions. This is a crucial factor apy, the negative predictive value of PET for excluding
with highly conformal RT, since only the contoured target relapse in the follow-up is 99–100% [91–93]. The major
will receive the prescribed radiation dose. However, there is limitation of surveillance PET is the low positive predictive
also a risk that [18F]FDG-avid sites that are not lymphoma value (between 11% and 53%) [91–93]. In the pediatric pop-
may be included in the target volume, increasing the risk of ulation, false-positive results may occur due to reactive neck
long-term complications of treatment. Ideally, biopsies lymph nodes, rebound thymic hyperplasia, or activated
should be obtained from [18F]FDG-avid lesions that are not brown fat.
clearly lymphoma, but this is rarely feasible. Therefore, Zinzani et al. prospectively performed follow-up PET
expert nuclear medicine support for contouring the PET- scans in 421 adult patients with lymphoma who achieved
positive lymphoma volume is highly recommended, in order CR by PET after therapy (160 with HL, 183 with aggressive
to avoid as far as possible the inclusion in the target volume NHL, 78 with indolent NHL). For HL and aggressive NHL,
of non-lymphoma [18F]FDG-avid lesions or areas of physio- the number of true-positive scans decreased and the true-
logical uptake/accumulation [84]. negative scans increased with time. As expected, the inci-
Movement of tumors and normal structures during imag- dence of relapse was stable over time for those with indolent
ing and RT has important consequences for the delivered NHL. Compared to the prior studies in pediatric patients,
radiation. However, motion may also be used as an advan- the false-positive scan rate was low (~1% of 1789 scans).
tage. For lymphomas located in the mediastinum, RT in deep However, as much as 8% of patients had inconclusive-
breath-in hold (DIBH), by anatomically separating the medi- positive PET scans, two-thirds with concurrent negative CT
astinal lymphoma mass from the critical normal structures, scans; biopsy was important to confirm disease presence in
offers significant reductions in the radiation doses to the these patients. Most patients with inconclusive negative
scans did not relapse [94]. These authors also found that may limit its ability to detect the presence of disease in
PET positivity was more common in groups with a higher these tissues/organs [99].
risk of recurrence, based on a combination of prognostic The prognostic power of [18F]FLT PET and [18F]FDG-
factors, for all lymphoma types. The usefulness of PET for PET has been prospectively explored in HL and NHL.
surveillance in HL in remission after first-line therapy (with However, the combined use of these two tracers did not result
98% positive predictive value for detecting recurrence con- in any signficant benefit [100].
firmed histologically) may be related to risk factors at the There are several ongoing clinical trials investigating the
time of scanning [95]. In particular, patients with a residual role of [18F]FLT PET in patients with lymphoma, particularly
mass were more likely to exhibit recurrence on a PET scan for evaluating response to treatment. In a recent prospective
performed in the first 24 months of follow-up; afterward, study of 65 patients with aggressive lymphoma treated with
symptoms at the time of PET scanning were better predictor four cycles of R-CHOP and three cycles of ICE, interim [18F]
of recurrence. FLT PET had high predictive value for PFS and OS in
In summary, current evidence does not support the rou- patients with good prognosis. However, the positive predic-
tine use of [18F]FDG-PET/CT for follow-up of patients tive value, though slightly better than [18F]FDG-PET, was
with lymphoma, as also recognized by current clinical still very low [101].
guidelines [6, 73]. For indolent lymphomas, the Lugano
criteria recommend judicious use of follow-up PET/CT
scans [6]. Key Learning Points
• [11C]MET performed equally well as [18F]FDG for
staging lymphoma, although its uptake in untreated
Key Learning Points patients was not correlated with outcome.
• Current evidence does not support the routine use • [18F]FLT discriminated indolent from aggressive
of [18F]FDG-PET/CT for follow-up of patients with NHL better than [18F]FDG.
lymphoma. • There are several ongoing clinical trials investigat-
• For indolent lymphomas, the Lugano criteria rec- ing the role of [18F]FLT-PET/CT in patients with
ommend judicious use of follow-up PET/CT scans. lymphoma.
26.1.4 Radioimmunotherapy of Lymphomas

26.1.3.11 P ET Radiopharmaceuticals Other
than [18F]FDG In radioimmunotherapy (RIT), the cytotoxicity caused by an
Although [ F]FDG is the only current PET tracer approved
18
ionizing radiation is conveyed to the neoplastic cells through
and routinely used for the evaluation of patients with binding of a monoclonal antibody (acting as carrier of a suit-
lymphoma, positron-emitting amino acid tracers have also able radionuclide) to tumor-specific antigens [102].
been investigated. Efforts in the past 30 years or so have focused on the so-
called molecular therapy to identify possible tumor antigens
[11C]Methionine ([11C]MET) against which to address the monoclonal antibody, as selec-
As reported in a limited number of studies, uptake of [11C] tion of the appropriate antigen plays a key role for the efficacy
MET is observed in all grades of lymphoma, either HL or of treatment [103]. The ideal antigen should be densely and
NHL, with different degrees of uptake in low- and high- uniformly expressed on the surface of malignant cells (but not
grade NHL [96]. [11C]MET performed equally well as [18F] in normal cells), it should form a stable complex with the anti-
FDG for staging lymphoma, although its uptake in untreated body, and the antigen/antibody complex should not be inter-
patients was not correlated with outcome [96, 97]. nalized (although the latter property is not so crucial) [104].
Another important issue concerns the origin and structure
3′-18F-3′-Deoxythimidine ([18F]FLT) of these antibodies; many monoclonal antibodies were in
Uptake of [18F]FLT uptake has been reported in indolent fact initially developed from murine cells, thus being intrin-
and aggressive NHL, although with an intensity slightly sically heterologous proteins with immunogenic activity
lower than [18F]FDG uptake [98, 99]. Nonetheless, when administered to humans. This immunogenicity is much
[18F]FLT discriminated indolent from aggressive NHL lower for the chimeric antibodies (made up of components
better than [18F]FDG, its uptake being highly correlated for 60% human and 40% murine) or for the humanized
with the proliferation index Ki67 [99]. However, physio- antibodies (with 95% of human components and generally
logic high uptake of [18F]FLT imaging in several normal only the binding site murine) developed through genetic
tissues/organs (hematopoietic bone marrow, liver, spleen) engineering techniques. In fact, the formation in the patient
receiving these preparations of human anti-mouse antibodies with 90Y-Ibritumomab tiuxetan, in patients with mild throm-
(HAMA), anti-chimeric antibodies (HACA), or antihuman bocytopenia (platelet count between 100,000 and 149,0000/
antibodies (HAHA) can alter the pharmacokinetic profiles of μL), the administered activity should be reduced to 11 MBq/
the monoclonal antibody or predispose to allergic reactions kg (0.3 mCi/kg) [110].
upon repeat administration [105]. Radioimmunotherapy with 90Y-ibritumomab tiuxetan has
The radiolabeled anti-CD20 antibody 90Y-ibritumomab largely been demonstrated to be safe and effective for the
tiuxetan (Zevalin®) has been approved for the treatment of treatment of NHL, with response rates as high as 74–83% in
relapsed or refractory, low-grade, or follicular B-cell NHL or patients with relapsed or refractory low-grade, follicular, or
treatment of previously untreated follicular NHL in patients CD20+ transformed NHL—as assessed using the
who achieve a partial or complete response to first-line chemo- International Workshop response criteria [111–113]. The
therapy (www.zevalin.com). 90Y is chelated to tiuxetan, which favorable response rate to treatment has been reported at a
is covalently linked to the antibody via arginine and lysine. very high level (83%) also for patients in whom prior immu-
The rationale for therapy with 90Y-ibritumomab tiuxetan notherapy with the unlabeled andti-CD20 antibody (the same
is that the CD20 antigen is expressed on the surface of nor- constituent of 90Y-ibritumomab tiuxetan) had failed. Rates of
mal B-lymphocytes (except pre-B cells and secretory B complete response to RIT as high as even >50% have been
cells) and of more than 90% of B-cell lymphomas. Unlabeled reported (see example in Fig. 26.7), with duration of response
monoclonal antibody (rituximab) is given as part of up to 16 months [114, 115].
90
Y-ibritumomab tiuxetan therapy in order to block CD20 The potential of radioimmunotherapy with
antigens on normal B cells and the spleen and to facilitate 90
Y-ibritumomab tiuxetan has been explored not only in
deeper penetration into the tumor [106, 107]. The unlabeled patients with recurrent disease or after failure of other treat-
monoclonal antibody itself induces several mechanisms of ments but also as a frontline therapy and as a consolidation
tumor cell killing [108, 109]. therapy after favorable response to conventional chemo-
90
Y-ibritumomab tiuxetan is administered as a slow therapy regimens. Also in these clinical settings RIT with
i.v. infusion over 10 min; the pharmacokinetics of 90
Y-ibritumomab tiuxetan proved to be a safe and highly
90
Y-Ibritumomab tiuxetan is characterized by a plasma effective treatment, ensuing significant benefits to the
half- life of 28 h. The administered activity is generally patients particularly in terms of progression-free survival and
15 MBq/kg body weight (0.4 mCi/kg), with a maximum conversion from clinical to molecular complete. Moreover,
limit of 1.2 GBq (32 mCi). While a platelet count <100,000/ although 90Y-ibritumomab tiuxetan is not approved for repeat
μL is an absolute contraindication to radioimmunotherapy treatment, several case reports on tolerance and clinical
Fig. 26.7 [18F]FDG-PET/CT scans obtained at baseline (top row) and finding is consistent with complete response to RIT; the duration of
then repeated 12 weeks after RIT with 90Y-ibritumomab tiuxetan (bot- response was 10 months (left, CT; center, PET; right, fused PET/CT)
tom row). This 62-year-old man had received five regimens of chemo- (reproduced with permission from: Jacene HA, Tirumani S, Wahl
therapy for low-grade NHL and then progressed with transformation to RL. Radionuclide therapy of lymphomas. In: Strauss HW, Mariani G,
diffuse large B-cell lymphoma after R-CHOP. The patient received Volterrani D, Larson SM, Eds. Nuclear Oncology – From
15 MBq/kg (0.4 mCi/kg) 90Y-ibritumomab tiuxetan. The baseline PET/ Pathophysiology to Clinical Applications. New York, NY: Springer;
CT scan shows intense [18F]FDG uptake in left supraclavicular lymph 2017: 1141–56)
nodes that resolved in the scan obtained 12 weeks after therapy. This
benefit indicate that repeat treatment might be considered in 26.2 Multiple Myeloma
selected patients. Finally, RIT with 90Y-ibritumomab tiuxetan
has demonstrated its efficacy also as a conditioning regimen 26.2.1 Epidemiology and Etiology
for autologous stem cell transplantation in combination with
high-dose chemotherapies, with the potential to overcome Multiple myeloma (MM) accounts for about 1.8% of all can-
resistance to chemotherapy [115–119]. cers and about 10% of all hematologic malignancies; it is
Dose-limiting toxicity for 90Y-ibritumomab tiuxetan is also the most common primary bone cancer. The number of
bone marrow suppression, which is generally reversible. new cases estimated in 2016 is >30,300 with over 12,650
However, no statistically significant correlation was noted deaths in the United States, with an incidence higher in men
between the development of hematological toxicity and (61%) and in the Afro-American population. Some catego-
pharmacokinetic parameters of the compound [120, 121]. ries of workers (i.e., farmers, workers in wood mills, paper
Since 90Y does not emit γ-rays (and on the other hand, scin- mills, and furniture manufacturing plants) seem to be more
tigraphic visualization of the secondary X-rays emitted by prone to develop the disease. Incidence of the disease peaks
the reaction of Bremsstrahlung is not satisfactory), between 45 and 85 years of age, with a median of 69 years,
biodistribution can be evaluated by administering a and rarely patients are younger than 40 years (<2%) (http://
diagnostic-level dose of the surrogate radiopharmaceutical seer.cancer.gov/statfacts).
111
In-ibritumomab tiuxetan [122]. The early images (up to The clonal proliferation of malignant plasma cells in the
about 24 h) typically show activity in the blood pool, with bone marrow may result both in local growth and in systemic
significant uptake in the liver and spleen, while accumula- effects due to the overproduction of a monoclonal protein
tion in the lung and bone is generally low. Dosimetric stud- (M-protein). Most commonly the M-protein is an IgG (52%)
ies have demonstrated an up to 850-fold greater radiation or IgA (21%), rarely IgD or IgE. Biclonal plasma proteins
dose to the tumor lesions than to normal organs, and cumu- can also be detected, such as IgG and IgA (33%) or IgG and
lative urinary excretion is only 7% of administered activity IgM (24%); non-secretive (negative both serum and urine
at 7 days post-administration. immunofixations) and micromolecular (only light chains
Another anti-CD20 monoclonal antibody (tositumomab) secreted) subtypes have been also described (1%). Almost all
labeled with 131I has been approved for clinical use in the patients evolve from an asymptomatic premalignant stage,
United States in 2003, with similar indications as Zevalin® in termed monoclonal gammopathy of undetermined signifi-
patients with NHL. 131I-tositumomab (Bexxar®) is directed cance (MGUS), through the smoldering myeloma (SMM),
against a different epitope of the same CD20 surface antigen, and eventually to “malignant” MM [125].
and the reported response rate is about 70% [123, 124]. MM differs from MGUS and SMM by the presence of
However, it has recently withdrawn from the market because end-organ damage associated with a complex syndrome
of declining sales. named CRAB [126]: hypercalcemia (C), renal failure (R),
anemia (A), and bone lesions (B). MM, MGUS, and smol-
dering myeloma (SMM) are defined according to the updated
Key Learning Points version of the criteria for the diagnosis of plasma cell prolif-
• In radioimmunotherapy, the cytotoxicity caused by erative disorders established by the International Myeloma
an ionizing radiation is conveyed to the neoplastic Working Group (IMWG) [126, 127].
cells through binding of a suitably radiolabeled High-risk SMM should be treated, because of the very
monoclonal antibody. high probability (about 80%) of developing a CRAB sign in
• The radiolabeled anti-CD20 antibody 90Y-ibritumomab a short period [127]. In patients with SMM and a free light
tiuxetan (Zevalin®) has been approved for the treat- chain ratio >100 (about 15% of th43 cases), the risk of pro-
ment of relapsed or refractory, low-grade, or follicu- gression within 2 years is 64–72% [128].
lar B-cell NHL or treatment of previously untreated Excess bone marrow plasma cells, M-protein, osteolytic
follicular NHL in patients who achieve a partial or bone lesions (present in >90% of patients), renal disease, and
complete response to first-line chemotherapy. immunodeficiency [129] are the pathophysiologic bases of
• Dose-limiting toxicity for 90Y-ibritumomab tiuxetan the clinical manifestations of MM. Severe bone pain, patho-
is bone marrow suppression. logic fractures (>70%), spinal cord compression (2–3%),
• Another anti-CD20 monoclonal antibody (tositu- and hypercalcemia (15%) are caused by lytic bone lesions
momab) labeled with 131I has been approved for clini- (spinal cord compression may also occur as a consequence
cal use in the United States in 2003, with similar of the collapse of a vertebral body due to the disappearance
indications as Zevalin®. However, it has recently with- of a lesion in response to therapy) [129].
drawn from the market because of declining sales. Bone lesions in MM can occur in any bone. At diagnosis,
almost 10% of the patients present diffuse osteopenia or
osteoporosis. MM bone disease involves more than 50% of a 26.2.2 Patients’ Workup
bone surface and osteoblastic activity is drastically depressed
or absent [129]. Initial investigation of a patient with suspected MM should
Transformed malignant plasma cells are responsible, both include family history, past medical history, a complete
directly and indirectly, of local bone resorption, through acti- blood count with differential, and a complete biochemistry,
vation of osteoclast activity and inhibition of the osteoblasts. including liver function tests, renal function tests, electro-
The local proliferation of myeloma cells may lead to the lytes, calcium, and albumin [125, 132]. The course of MM is
appearance of a solid lesion involving the soft tissue primar- highly variable, and the clinical pattern is remarkably hetero-
ily arising outside the bone or extending from a preexistent geneous. Many studies have identified prognostic factors
bone lesion. In the absence of systemic spread, this situation capable of predicting this heterogeneity in survival (serum
is named solitary plasmacytoma [127, 129]. β2-microglobulin, albumin, C-reactive protein, and lactate
Excess protein produced by myeloma can increase the dehydrogenase) [125].
viscosity of plasma. Hyperviscosity is associated with The standard work-up of MM is based on a number of
impairment of the microcirculation, particularly in the eye laboratory tests used for risk stratification. The International
and central nervous system causing visual impairment, dizzi- Staging System (ISS) is a powerful and reproducible three-
ness, headache, and hearing loss. Rarely there is spontaneous stage classification, associated with the outcome. Gene-
bleeding [130] due to platelet dysfunction. Patients with expression profiling may stratify patients as having standard
myeloma are at increased risk of infection due to reduction or high-risk disease, but this is not yet an established routine
of normal plasma cell function. Neuropathy is usually sym- practice [125].
metric, distal sensory, or sensory-motor with axonal degen- The fraction of plasma cells infiltrating the bone mar-
eration [131]. In 1% of the cases, myeloma occurs in a rare row is evaluated on May-Grünwald Giemsa stained
syndrome (POEMS syndrome) [129] characterized by poly- smears. Cytology of bone marrow aspirates remains the
neuropathy and monoclonal plasma cell proliferative disor- standard method for quantitating plasma cell infiltration.
der (almost always λ) associated with one or more major The bone marrow plasma cell labeling index (PCLI), a
criteria (sclerotic bone lesions, Castleman’s disease, or ele- parameter of the DNA synthesis rate derived from in vitro
vated levels of VEGF) and one or more minor criteria incubation with tritiated thymidine, predicts survival
(organomegaly, extravascular volume overload, endocrinop- [129]. This index is usually low at diagnosis (<1%) in
athy, skin changes, papilloedema, thrombocytosis/polycy- MGUS and SMM, but it raises at relapse correlating with
themia) [127]. neoangiogenesis, and inversely with survival, regardless
The presence of lytic bone lesions is highly suggestive for of the tumor mass. Forty percent of patients with symp-
MM, but it is not by itself sufficient to establish the diagno- tomatic MM have normal PCLI.
sis. The minimum criteria for MM diagnosis are either the
detection of at least 10% abnormal plasma cells in the bone
marrow (after biopsy) or of M-protein levels >30 g/L in
Key Learning Point
serum or >1 g excreted in the 24-h urine [126].
• The International Staging System (ISS) is a power-
ful and reproducible three-stage classification for
MM, associated with clinical outcome.
Key Learning Points
• MM differs from MGUS and SMM because of the
presence of end-organ damage associated with a
complex syndrome named CRAB. 26.2.3 Treatment
• The pathophysiologic bases of the clinical manifes-
tations of MM are excess bone marrow plasma Despite all treatment options, MM is still considered an
cells, M-protein, osteolytic bone lesions, renal dis- incurable disorder with a median survival of about 2–3 years
ease, and immunodeficiency. in high-risk patients, compared to 5–7 years with high-dose
• Bone lesions in MM can occur in any bone therapy followed by autologous stem cell bone marrow
segment. transplantation (ASCT) in standard-risk patients.
• The minimum criteria for MM diagnosis are either Cytogenetic and FISH studies reveal chromosome abnor-
the detection of at least 10% abnormal plasma cells malities of prognostic significance in 33% and in 90%,
in the bone marrow (on bone marrow biopsy) or of respectively, of the patients with MM [129]. Translocations
M-protein levels >30 g/L in serum or >1 g excreted and some chromosomal alteration are associated with poor
in the 24-h urine. prognosis (median survival 25 months). An elevated PCLI
also confers a negative prognosis [126]. Table 26.6 reports
the 5-year survival rate of MM patients according to stage of [18F]FDG-PET/CT, and MRI [136]. These techniques enable
the disease [133]. to evaluate the total number of bone lesions, to distinguish
MGUS or SMM from active myeloma, as well as to better
discriminate between stage II and III disease. The original
Key Learning Point system based on conventional X-ray only is still employed in
• Despite all treatment options, MM is still consid- areas where access to advanced imaging modalities is
ered an incurable disorder with a median survival of limited.
about 2–3 years in high-risk patients. An accurate staging system is crucial, since MM has
extremely heterogeneous outcomes and treatment is strongly
dependent on the disease onset. Based on the stage and risk
factors, risk-adapted therapeutic strategies can be defined.
26.2.4 Imaging for Staging Multiple Myeloma MGUS does not require any treatment (except long-term
observation), and early treatment of patients with SMM
Imaging is used to evaluate the extent and severity of bone seems not to prolong survival. The diagnosis of active
involvement (intramedullary and/or extramedullary, site and symptomatic MM requiring therapy is based on end-organ
number of lesions) at baseline, including disease-related effects of the disease.
complications, to assess response to treatment, and provide
follow-up surveillance [134]. The standard skeletal survey
reveals punched-out lytic lesions, osteopenia, or fractures in Key Learning Points
approximately 80% of patients at diagnosis. Due to the lack • Imaging is used to evaluate the extent and severity
of the osteoblastic response, bone scintigraphy with 99mTc- of bone involvement at baseline, to assess response
diphosphonates is insensitive for the detection of many to treatment, and to provide follow-up surveillance.
myelomatous bone lesions. • Durie and Salmon introduced a clinical staging sys-
The clinical staging system for MM (International Staging tem based on the presence of bone lesions to grade
System, ISS, Table 26.7), introduced over 25 years ago, is severity of the disease.
mainly based on serum β2-microglobulin and albumin levels. • The Durie and Salmon PLUS system was released
ISS has proven to adequately estimate tumor burden and risk in 2005 to improve the accuracy of staging using
stratification, also enabling to distinguish MGUS and SMM advanced imaging modalities such as [18F]FDG-
from MM [126]. The CRAB criteria are instead preferred to PET (preferably PET/CT) and MRI.
establish MM-related organ dysfunction [134]. In 1975
Durie and Salmon introduced a clinical staging system based
on the presence of bone lesions to grade severity of the dis-
ease [135]. The original Durie and Salmon system was 26.2.4.1 N on-radionuclide Imaging of Disease
essentially based on the use of planar X-ray. A newer ver- Burden
sion, the Durie and Salmon PLUS system, was released in Appropriate use of imaging techniques is essential for deter-
2005 (see Table 26.8) to improve the accuracy of staging mining the severity of bone involvement and characterizing
using advanced imaging modalities such as [18F]FDG-PET, skeletal complications. Imaging is also critical for detecting
extramedullary foci, for diagnosing infection and other
complications, and for evaluating progression of the dis-
Table 26.6 Five-year survival rates of MM patients according to the
stage of disease ease. Recently, the European Myeloma Network provides
5-year
survival
Table 26.7 International staging system (ISS) for MM
Stage rate (%)
Stage I 82 Stage Criteria
ISS stage I (serum albumin >3.5, serum beta-2- Stage I Serum β2-microglobulin <3.5 mg/L and albumin ≥3.5 g/
microglobulin <3.5) and dL
No high-risk cytogenetics Stage II No stage I or III
Normal LDH There are two categories:
Stage II 62 – Serum β2-microglobulin <3.5 mg/L but serum
Neither stage I or III albumin <3.5 g/dL;
Stage III 40 – Serum β2-microglobulin ranging from 3.5 to
ISS stage III (serum beta-2-microglobulin >5.5) and <5.5 mg/L irrespective of the serum albumin level
High-risk cytogenetics [t(4;14), t(14;16), or del(17p)] Stage III Serum β2-microglobulin ≥5.5 mg/L
or elevated LDH (modified from: Kyle RA, Rajkumar V. Criteria for diagnosis, staging,
(modified from: Rajkumar SV. Updated diagnostic criteria and staging risk stratification and response assessment of multiple myeloma.
system for multiple myeloma. ASCO Educational Book 2016;418–23) Leukemia. 2009;23:3–9)
Table 26.8 Overview of Durie and Salmon staging system (A) and the Durie and Salmon PLUS staging system (B)
A: Original Durie and Salmon staging system B: Durie and Salmon PLUS staging system
Measured myeloma cell Classification
mass in whole body
(myeloma cells in
Disease Criteria billions/m2) MRI and/or 18[F]FDG-PET
MGUS Stage I (low cell mass) 600 billion All negative
All of the following:
– Hemoglobin value >10 g/dL
– Serum calcium value normal or <10.5 mg/dL
– Normal bone structure (scale 0) or solitary bone
plasmacytoma only at bone X-ray
– Low M-protein production rates:
IgG value <5.0 g/dL
IgA value <3.0 g/dL
Urine light chain M-protein <4 g/24 h
Smoldering Stage II (intermediate cell mass) 600–1200 billion Stage I A Can have single
or indolent No Stage I or Stage III plasmacytoma and/or
myeloma limited disease on imaging
MM Stage III (high cell mass) 1200 billion Stage I B <5 focal lesions
One or more of the following: Mild diffuse disease
– Hemoglobin value <8.5 g/dL Stage II A/B 5–20 focal lesions
– Serum calcium value >12 mg/dL Moderate diffuse disease
– Advanced lytic bone lesions at bone X-ray Stage III A/B 20 focal lesions
(scale 3) Severe diffuse disease
– High M-protein production rates
IgG value >7.0 g/dL
IgA value >5.0 g/dL
Urine light chain M-protein >12 g/24 h
Subclassification (either A or B)
A: Relatively normal renal function (serum creatinine <2.0 mg/dL) and no extramedullary disease
B: Abnormal renal function (serum creatinine value ≥2.0 mg/dL and extramedullary disease
(modified from: Lütje S, Rooy JWJ, Croockewit S, Koedam E, Oyen WJG, Raymakers RA. Role of radiography, MRI and FDG-PET/CT in diagnos-
ing, staging and therapeutical evaluation of patients with multiple myeloma. Ann. Hematol. 2009;88:1161–8)
recommendations for the management of the most common the early stages (lytic lesions become apparent when as much
complications of MM [137]. as 30–50% of the bone mineral density is lost), and some
For staging purpose, the detection of osteolytic lesions is areas are not well visualized (i.e., lateral and antero-posterior
generally based on conventional planar X-ray (Fig. 26.8) views of the spine are needed for better visualizing the verte-
[138–140]. Lytic lesions on plain X-rays are typically holes, bral bodies). Moreover, it is not possible to differentiate
punched-out lesions with absent reactive sclerosis of the myeloma-related osteoporosis from osteoporosis due to
surrounding bone, typically localized in the flat bones of the other causes.
skull and pelvis. In the long bones, lesions may appear as A major disadvantage of conventional X-ray is the rela-
endosteal scalloping, small lytic lesions (<1 cm), mottled tively long imaging time and the necessity of maintaining
areas of multiple small lesions, or large destructive lesions different positions that not all patients (often elderly and suf-
[141]. These lesions are the consequence of nodular replace- fering pain for previous pathological fractures) are able to
ment of marrow by plasma cells with destruction of the maintain. The potential of low-dose whole-body radio-
entire bone. Conventional X-ray may also reveal diffuse graphic systems that enable to obtain high-quality imaging
osteoporosis, best recognized in the spine [142]. of the bones in shorter time has been explored, although
Radiological and clinical findings (pain) are used in the clinical experience in this clinical setting is still limited.
score system to predict the risk of fracture of long bones Whole-body multi-detector computed tomography
and to identify optimal treatment (Table 26.9). Asymptomatic (WBMD-CT) is more sensitive for the detection of lytic
patients with X-ray evidence of disease (at least one lytic lesions in myeloma compared to conventional radiography
lesion) have a high risk of progression, with a median time (between 70% and 89% for MDCT, depending on stage of
of 8 months. the disease) as it can detect more lesions, with an extremely
Whole-body skeletal survey, traditionally used in the high resolution (very small lesions can be detected, even
Durie and Salmon staging system, is burdened with a rela- <5 mm) and better performance when evaluating areas that
tively high rate of false negatives (30–70%) since it cannot are critical for planar X-ray; it is very quick to perform
always identify the presence of bone lesions, especially in (2 min or less), has a more accurate evaluation of areas with
a c d
Fig. 26.8 Examples of X-ray appearance of lytic lesions in patients with D10–L5 vertebroplasty (reproduced with permission from: Sollini
with MM. (a) Single lytic lesion in the right humerus. (b) Two lytic M, Galimberti S, Boni R, Erba PA. Diagnostic applications of nuclear
lesions with abundant osteoblastic reaction. Postero-anterior (c) and lat- medicine: multiple myelomas. In: Strauss HW, Mariani G, Volterrani D,
eral projection (d) demonstrating multiple lytic lesions at different lev- Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical
els of the dorsal spine (D6–D9) in a patient with MM already treated Applications. New York, NY: Springer; 2017: 395–434)
instability or at risk of fracture, and is superior regarding the with CT. Therefore the use of CT is possible even in patients
planning for radiotherapy or surgical interventions; further- with Bence-Jones proteinuria, who are at risk of cast
more, it can be employed in cases of suspected spinal cord nephropathy and renal failure. Whole-body LDCT advan-
compression when MRI is contraindicated [137, 140, 143]. tages over whole-body X-ray include the following features
Additionally, it can be identified also extramedullary lesions. [143, 146, 147]:
However, specificity of MDCT for assessing certain skeletal
changes often associated with myeloma, such as osteopenia, • Superior diagnostic sensitivity for depiction of osteolytic
is low. Low-dose whole-body CT (WBLD-CT), better than lesions
conventional X-ray in lytic lesion detection [144], may be • Superior performance in estimating fracture risk and bone
used to reduce the radiation burden [134, 145]. Iodinated instability
contrast agent is not required for evaluating the skeleton • Shorter duration of the examination
Table 26.9 Scoring system for diagnosing impending pathological • Production of higher-quality 3D high-resolution images
fractures for planning biopsies and therapeutic interventions
A • Demonstration of unsuspected manifestations of myeloma
Variable Score or other diseases
1 2 3
Site Upper limb Lower limb Peritrochanteric
Major disadvantages of whole-body LDCT include
Pain Mild Moderate Functional
Lesion Blastic Mixed Lytic
increased length of time required for radiologists to report
Size <1/3 diameter 1/3–1/2 diameter >2/3 diameter their findings, lack of availability in several centers, and
B lack of specificity. Furthermore, although exposure to radi-
Total Risk of impending Management ation is much lower compared with standard CT, it contin-
score fracture ues to be higher than whole-body X-ray (mean dose of
≤7 Low (5%) Conservative (chemo/radiotherapy) whole-body LDCT is approximately 3.6 and 2.8 mSv for
8 Suggestive (15%) Conservative or surgery (to be women and men, respectively, compared with 1.2 mSv for
evaluated case by case)
whole-body X-ray [148]). Figure 26.9 shows typical CT
≥9 Diagnostic (33%) Prophylactic bone fixation
examples of lytic lesions of the spine and of a solitary
(adapted from: Winterbottom AP, Shaw AS. Imaging patients with
myeloma. Clin Radiol. 2009;64:1–11)
plasmacytoma.
a b c
d e
Fig. 26.9 Typical CT appearance of lytic lesions of the spine (a, sagit- M, Galimberti S, Boni R, Erba PA. Diagnostic applications of nuclear
tal view) and iliac bone (b and c in coronal and transaxial views, respec- medicine: multiple myelomas. In: Strauss HW, Mariani G, Volterrani D,
tively) with the bone-imaging window. The bottom panel shows the Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical
transaxial (d) and sagittal (e) views of the posterior tract of right rib at Applications. New York, NY: Springer; 2017: 395–434)
the site of a plasmacytoma (reproduced with permission from: Sollini
a b
Fig. 26.10 MRI appearance of MM lesion in the left pubic bone: (a) Erba PA. Diagnostic applications of nuclear medicine: multiple myelo-
T1 image; (b) T2 image. The lesion is hypointense in the T1-weighted mas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
sequence and hyperintense in the T2-weighted (and STIR) sequence Oncology – From Pathophysiology to Clinical Applications. New YorK,
(reproduced with permission from: Sollini M, Galimberti S, Boni R, NY: Springer; 2017: 395–434)
Whole-body magnetic resonance imaging (WB-MRI) response to treatment and with overall survival. Patients
accurately depicts the marrow involvement in MM patients. with advanced disease and normal MRI respond better to
It should be remarked that MRI depicts bone marrow involve- conventional chemotherapy and have better survival than
ment, while CT and skeletal survey reveal lytic lesions. MRI patients with focal/diffuse bone marrow infiltration.
is the method of choice for evaluating bone marrow involve- Nevertheless, patients with positive MRI do not always
ment. MRI has high sensitivity in distinguishing focal or dif- require immediate treatment. The detection of ≥10 spine
fuse infiltration of the bone (especially in the spine), as well lesions in patients with advanced stage is associated with an
as detecting destruction of the mineral bone. Myeloma increased risk (6–11-fold) of fracture than patients with nor-
lesions in the marrow appear typically hypointense in mal MRI or with less than 10 lesions. MRI also enables to
T1-weighted sequences and hyperintense in T2-weighted distinguish osteoporotic from malignant fractures and is the
and STIR sequences (Fig. 26.10). technique of choice when cord compression is suspected
Five MRI patterns of marrow involvement have been rec- providing accurate information about the epidural space, the
ognized in multiple myeloma [138, 139, 144, 149], as level and extension of cord or nerve root compression, and
follows: size of the lesion. The main disadvantages of MRI are the
relatively low availability/access, high cost, and long scan
• Focal (in 30–50% of patients with advanced MM, but also duration (about 45 min). Moreover, claustrophobic patients
in high-risk SMM) or patients with some metallic implant/device cannot
• Diffuse (in 25–40% of patients with MM) undergo the exam. The typical MRI protocol for evaluating
• Combined diffuse and focal (in 10% of patients with MM) MM patients does not include some skeletal segments (i.e.,
• Normal (in 15–25% of patients with MM) the sternum, clavicles, and ribs) and may underestimate the
• Variegated or “salt and pepper” (in 1–5% of patients with disease burden. Whole-body MRI (WB-MRI, Fig. 26.11)
MM) may overcome those limitations since it is more sensitive
than WB-MDCT for assessing bone marrow infiltration
MRI findings reflect pathophysiological processes such (both focal and diffuse) and is recommended immediately
as iron overload, amyloid deposition, or reactive marrow after the X-ray survey. It allows identifying lesions early in
hyperplasia. Patients with MM may exhibit normal MRI their course (before osteolysis occurs) as well as nonsecre-
appearance when the tumor burden is low, while focal or dif- tory and macrofocal myeloma (especially in the spine and
fuse bone marrow infiltration exhibits variable patterns. At pelvis), which is characterized by lesions with low metabo-
diagnosis, bone marrow infiltration is detected by MRI in lism and activity.
29–50% of the patients with Durie-Salmon stage I disease MRI may also be improved with the use of dynamic
and negative plain radiographs. MRI findings correlate with contrast-enhanced sequences (DCE-MRI) that evaluate the
a b
Fig. 26.11 Whole-body MR imaging in a 65-year-old patient with PA. Diagnostic applications of nuclear medicine: multiple myelomas.
MM. There are multiple osteolytic lesions in the cranial vault (a), ster- In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
num and ribs (b), and in the femurs (c). Diffuse infiltration of the spine Oncology – From Pathophysiology to Clinical Applications. New York,
is shown by the reduced signal intensity of the bone marrow (repro- NY: Springer; 2017: 395–434)
duced with permission from: Sollini M, Galimberti S, Boni R, Erba
bone marrow microcirculation caused by myeloma-induced 26.2.4.2 R

adionuclide Imaging of Disease
angiogenesis. The DCE-MRI-derived “A” variable describes Burden
intensity of the signal (therefore angiogenesis), and is found
to be significantly increased in MM patients, in whom it cintigraphy with Bone-Seeking Radiopharmaceuticals
S
correlates with the presence of osteolytic lesions, with the Bone scintigraphy with 99mTc-disphosphonates is not useful
degree of bone destruction, and with the presence of local for diagnosis nor for staging of MM, as the detection of bone
complications. The amplitude of “A” is a negative prognostic lesions is only about 35–60% [145, 150] (Fig. 26.12), with
factor for event-free survival and for overall survival in MM sensitivities generally lower than that of conventional X-rays;
[134, 145, 150, 151]. in particular, increased tracer uptake has been reported in
Diffusion-weighted imaging (DWI) MRI uses the appar- only 44% of regions radiographically abnormal [156].
ent diffusion coefficient values to better evaluate myeloma Lesions that are well defined on the bone scan are the result
burden and infiltration patterns [152], for correlation with of complications in MM, mainly osteoblastic response to a
treatment response. DWI MRI is superior to whole-body compression fracture of a vertebral body or pelvic insuffi-
X-ray for the detection of bone involvement in patients ciency fracture, soft tissue calcifications within a plasmacy-
with relapsed/refractory MM in all areas of the skeleton toma, or tumor-associated amyloidosis. This pattern is
except the skull [153]. Similarly, DWI MRI was found consistent with the prevalent loss of bone mass not associ-
more sensitive than [18F]FDG-PET/CT in detecting ated with osteoblastic activation, typical pattern of the
myeloma lesions (77% and 47%, respectively) in both myeloma bone lesions. Nonetheless, bone scintigraphy may
newly diagnosed and pre-treated MM patients [154]. In a be helpful for evaluating some specific skeletal segments
more recent study, DWI MRI resulted particularly accurate such as the sternum and ribs that can hardly be explored with
in diagnosing diffuse disease (37% of regions imaged on plain X-rays (Fig. 26.13).
whole-body DWI scans compared with only 7% on [18F]
FDG-PET/CT) while sensitive in the detection of focal Scintigraphy with Tumor-Seeking
lesions which was similar [155]. Radiopharmaceuticals
Radionuclide imaging with nonspecific oncotropic radio-
pharmaceuticals, such as 67Ga-citrate, 99mTc(V)-DMSA,
201
Tl-chloride, 99mTc-sestamibi, or [18F]FDG, has all met
greater success in imaging MM than conventional bone
Key Learning Points scanning agents. 67Ga-citrate localizes in areas of active
• Appropriate use of imaging techniques is essential tumor either through primary localization within the tumor
for determining the severity of bone involvement, cells or because of the presence of mononuclear cell and
characterizing skeletal complications and detecting lymphocytic infiltrates that characterize the tumor-induced
extramedullary foci of disease. secondary “inflammatory response” [157, 158]. 67Ga-citrate
• For staging purposes, the detection of osteolytic lesions is rarely used for tumor detection in MM because increased
is generally based on conventional planar X-ray. uptake may be observed in some abnormal soft tissue sites
• In myeloma patients, whole-body multi-detector and in solitary myelomas of bone [156, 159], although with
computed tomography (WBMD-CT) is more sensi- variable degrees. The use of this radiopharmaceuticals is also
tive for the detection of lytic lesions than conven- hampered by certain disadvantages (e.g., multiday scanning
tional radiography. and low resolution) versus other tracers such as 99mTc-
• MRI is the method of choice for evaluating bone sestamibi or [18F]FDG. When infection is suspected in MM
marrow involvement. patients, 67Ga-citrate may be used if 99mTc-HMPAO-
• Five MRI patterns of marrow involvement have leukocytes or 111In-oxine-leukocytes are not available [160].
been recognized in multiple myeloma. 99m
Tc(V)-DMSA has been reported to accumulate in plasma-
• Patients with MM may exhibit normal MRI appear- cytoma, irrespective of the presence of amyloidosis [161].
ance when the tumor burden is low, while focal or dif- 201
Tl-chloride has been used for the detection of MM,
fuse bone marrow infiltration exhibits variable patterns. based on uptake mediated by either increased metabolic
• MRI findings correlate with response to treatment demand of the tumor or secondary inflammatory response
and with overall survival. induced in the marrow [162]. In a study comparing
• DWI MRI uses the apparent diffusion coefficient val- 201
Tl-chloride and 99mTc-labeled bone-seeking agents in
ues to better evaluate myeloma burden and infiltra- patients with MM, 201Tl-chloride was shown to be a promis-
tion patterns, for correlation with treatment response. ing agent for detecting disease [163]; however, in the clinical
routine, this agent is currently replaced by 99mTc-sestamibi,
Fig. 26.12 Bone scintigraphy with 99mTc-MDP (a) and [18F]FDG-PET/ sion from: Sollini M, Galimberti S, Boni R, Erba PA. Diagnostic applica-
CT (b) in a patient with IgG-secreting MM. Conventional bone scintigra- tions of nuclear medicine: multiple myelomas. In: Strauss HW, Mariani G,
phy revealed some abnormality in the right sacroiliac region, but did not Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology
detect lesions in a left rib and in the left sacrum (reproduced with permis- to Clinical Applications. New York, NY: Springer; 2017: 395–434)
which shows similar localization properties, associated how- [164]) is based on extension (E) and intensity (I) of 99mTc-
ever with images of higher quality and lower cost. sestamibi uptake evaluated in planar images of the whole
99m
Tc-sestamibi accumulates in cells with high metabolic body (Fig. 26.14) [165].
requirements and can therefore be employed to assess in a The degree of radiopharmaceutical uptake correlates
semiquantitative manner the degree of bone marrow infiltra- well with the degree of plasma cell infiltration, with the
tion by plasma cells; this technique is both highly sensitive amount of a monoclonal component, with the clinical status
(92%) and specific (96%). The patterns of 99mTc-sestamibi and with stage of the disease. Good correlation was also
uptake typical of MM can be described as normal (N), focal found between 99mTc-sestamibi uptake and other parameters
(F), diffuse (D), and mixed (F + D). Another semiquantita- of disease activity, such as the serum levels of lactate dehy-
tive score (similar to that originally described by Pace et al. drogenase, C-reactive protein, and β2-microglobulin.
a b
Fig. 26.13 Patterns of bone scintigraphy in different patients with left femur (reproduced with permission from: Sollini M, Galimberti S,
MM. (a) Mild and diffusely increased 99mTc-MDP uptake, as more fre- Boni R, Erba PA. Diagnostic applications of nuclear medicine: multiple
quently observed in MM patients. (b) Less frequent pattern of bone myelomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
scintigraphy showing multiple with markedly increased 99mTc-MDP Nuclear Oncology – From Pathophysiology to Clinical Applications.
uptake in the skull, left clavicula, homeri, ribs, spine, left iliac bone, and New York, NY: Springer; 2017: 395–434)
Diffuse moderate or intense 99mTc-sestamibi uptake or focal approach has never reached wide application in the routine
uptake correlates with poor prognosis. 99mTc-sestamibi clinical practice.
imaging is less sensitive than MRI (50.5% versus 100%) [18F]FDG is currently the standard for radionuclide imag-
[166] for evaluating focal lesions localized in the spine ing in patients with MM. It identifies early bone marrow
(especially in patients with early disease stage). SPECT infiltration in patients with apparent solitary plasmacytoma
imaging has been shown to overcome the relatively poor and demonstrates the extent of extramedullary involvement.
spatial resolution of planar 99mTc-sestamibi scintigraphy. It [18F]FDG-PET shows both high sensitivity (86%) and high
is important to emphasize that the 99mTc-sestamibi scan is specificity (92%) [170], with the additional advantage of dis-
always negative in case of MGUS, while false-negative tinguishing between metabolically active disease ([18F]FDG-
results have been described in case of overexpression of positive) and clinical conditions characterized by very low
P-glycoprotein, associated with multidrug-resistant burden, such as MGUS or smoldering disease ([18F]FDG-
myeloma [145, 164, 167]. negative). Active transformed plasma cells are extremely
More recently the use of somatostatin receptor (SSTR) glucose-avid; therefore, [18F]FDG-PET shows focal or dif-
imaging as alternative method to visualize malignant plasma fuse patterns of uptake reflecting the distribution of bone
cells has been proposed [168]. Preliminary data in patients marrow disease. While patients with MGUS generally have a
with relapsing MM indicate that SSTR imaging is very sen- negative [18F]FDG-PET scan (although diffuse low-grade
sitive (abnormal accumulation in 82% of patients versus bone marrow uptake can occasionally be observed), the
33% detection rate using whole-body X-ray). Partial or total appearance of [18F]FDG focal uptake in a patient with MGUS
normalization of the scintigraphic pattern has also been indicates transformation into an active myeloma.
observed in patients responding to treatment. The expression The detection of extramedullary disease at diagnosis
of SSTR could be related to aggressiveness of the disease. using [18F]FDG-PET scan correlates with poor prognosis
Scintigraphy with the serum amyloid P (SAP) labeled (see Figs. 26.15 and 26.16 for different representative pat-
with either 99mTc or 123I has been proposed to confirm the terns of [18F]FDG uptake in patients with plasmacytoma
diagnosis and to quantify the extent of the deposits of light and MM, respectively). [18F]FDG-PET identifies patients
chain (AL) amyloidosis [145, 150, 169]. Nevertheless, this with high-risk disease and can be employed to monitor
a b c d
A C
B D
Fig. 26.14 Different patterns of 99mTc-sestamibi scintigraphy demon- tions of nuclear medicine: multiple myelomas. In: Strauss HW, Mariani
strating different degrees of bone marrow involvement: focal (a, b, d) G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
and diffuse (c). (A, B, C, D) Example of semiquantitative score evalu- Pathophysiology to Clinical Applications. New York, NY: Springer;
ated as I3E3 (I = intensity; E = extension) (reproduced with permission 2017: 395–434)
from: Sollini M, Galimberti S, Boni R, Erba PA. Diagnostic applica-
patients with either nonsecretory myeloma or in complete larger bone lesion characterized by higher uptake. The com-
remission without measurable M-component [171] leading plementary imaging information provided by the CT compo-
to upstage the disease in 31–37.5% of patients (depending nent of a [18F]FDG-PET/CT examination is effective, even if
on the series) and to change treatment management in up CT is performed in most cases in the low-dose mode
to 56% of the cases [172, 173]. Data derived from The (Fig. 26.18). [18F]FDG-PET/CT has been reported to be
National Oncologic PET Registry (NOPR), a large pro- more sensitive than both plain X-ray (89.2% versus 47.4%)
spective program aiming at recording changes of the and CT alone (89.2% versus 70.4%) [175] also when used to
intended clinical management of cancer patients on the localize extramedullary sites of disease, thus resulting in
basis of the [18F]FDG-PET/CT findings, have confirmed a upstaging (additional lesions in almost 30% of the patients
definite impact on management in about 52% of patients who had been diagnosed with solitary plasmacytoma by
when the scan was requested for staging purpose, 46% in MRI [176]) which impact the therapeutic strategies [177,
case of restaging, and 51% in presence of suspected recur- 178]. On the other hand, sensitivity of [18F]FDG-PET for
rence [174]. assessing diffuse disease in the spine and pelvis remains
In patients with MM, the 2.5 threshold for SUV (often almost half that of MRI [167].
considered as the cutoff for other forms of cancer) may not Thus, MRI remains the method of choice for evaluating
be applicable. In fact, many lesions <10 mm cannot be diffuse bone marrow infiltration in the spine, considering
detected on the [18F]FDG scan using this threshold if one that diffusely increased nonspecific [18F]FDG uptake is fre-
does not take into account the recovery coefficient, whereby quently observed in the bone marrow of young people or in
SUV decreases along with decreasing lesion size. patients with anemia [167, 172, 179]. [18F]FDG-PET/CT
Figure 26.17 shows an example of faint abnormal [18F]FDG should be preferred for an accurate initial staging of patients
uptake (SUVmax <2.5) in a bone lesion, with a concomitant with MM (Fig. 26.19), while MRI is to be preferred when
a b
Fig. 26.15 [18F]FDG-PET/CT in a patient with solitary plasmacytoma present in any other bone segments (reproduced with permission from:
of a right rib. The MIP image is depicted in panel (a). (b) Transaxial fused Sollini M, Galimberti S, Boni R, Erba PA. Diagnostic applications of
PET/CT section with corresponding CT component (c) and PET compo- nuclear medicine: multiple myelomas. In: Strauss HW, Mariani G,
nent (d) demonstrating an area of intense tracer uptake in a large lesion Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology
developing from the posterior trait of right rib. No additional lesions are to Clinical Applications. New York, NY: Springer; 2017: 395–434)
assessing the degree of bone marrow infiltration by plasma cases); however, when [18F]FDG-PET/CT and MRI are con-
cells in the early phase of the disease. By combining MRI of cordantly positive, no false-positive results are generally
the spine/pelvis and whole-body PET/CT, the ability to observed, yielding specificity and positive predictive values
detect sites of active MM, both medullary and extramedul- close to 100%. This consideration is of value in aiding clini-
lary, can be as high as 92%. Rarely, active disease may be cians to either continue with or change chemotherapy regi-
missed by both [18F]FDG-PET/CT and MRI (false-negative mens on the basis of persistently positive findings on imaging
studies. The low negative predictive values of [18F]FDG-

PET/CT (50%) and MRI (59%) and also of the combination • The detection of extramedullary disease at diagnosis
of the two (64%) suggest that negative results should always using [18F]FDG-PET/CT correlates with poor prognosis.
be considered with caution and correlated to other markers • [18F]FDG-PET/CT identifies patients with high-risk
of disease activity [180]. disease and can be employed to monitor patients
with either nonsecretory myeloma or in complete
remission without measurable M-component.
• [18F]FDG-PET/CT should be preferred for accurate
initial staging of patients with MM.
Key Learning Points • MRI should be preferred when assessing the degree
• [18F]FDG is currently the standard for radionuclide of bone marrow infiltration by plasma cells in the
imaging in patients with MM. early phase of the disease.
Fig. 26.16 Different patterns of [18F]FDG uptake in patients with more intense uptake corresponding to multiple bone and soft tissue sites
MM. (a) MIP image in top left box, and coronal sections (CT compo- of disease. The bottom row depicts a transaxial section (CT component
nent with mediastinal window, PET component, fused PET/CT) show- on the left, PET component in the middle, fused PET/CT on the right)
ing multiple sites with diffuse and moderate [18F]FDG uptake in bone showing humeral, spine, sternum, and rib involvement, with a lesion of
(as intense as the liver uptake), associated with other sites with focally the lateral right rib involving the chest wall. Of notice, [18F]FDG-PET/
increased tracer uptake, corresponding to multiple bone sites of active CT mimics with better spatial resolution the pattern of homogeneous
myeloma with variable glyco-metabolic activity. The bottom row and diffuse uptake at the humera and femurs, as typically visualized
depicts a transaxial section (CT component on the left, PET component with 99mTc-sestamibi in advanced MM (reproduced with permission
in the middle, fused PET/CT on the right) showing humeral, scapular, from: Sollini M, Galimberti S, Boni R, Erba PA. Diagnostic applica-
spine, and rib involvement. (b) MIP image in top left box and coronal tions of nuclear medicine: multiple myelomas. In: Strauss HW, Mariani
sections (CT component with mediastinal window, PET component, G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
fused PET/CT) showing multiple sites with diffuse and intense [18F] Pathophysiology to Clinical Applications. New York, NY: Springer;
FDG uptake in bone (more intense than the liver) with focal areas of 2017: 395–434)
26.2.4.3 [ 18F]FDG-PET/CT as Prognostic

Indicator Key Learning Points
The prognostic significance of [18F]FDG-PET/CT is not yet • The prognostic significance of [18F]FDG-PET/CT
well established in MM, although it has been suggested that is not yet well established in MM, although it has
survival of MM patients can be improved by altering treat- been suggested that survival of MM patients can be
ment in patients in whom [18F]FDG suppression cannot be improved by altering treatment in patients in whom
achieved after induction therapy [181]. [18F]FDG-PET/CT [18F]FDG suppression cannot be achieved after
also offers the advantage of identifying any superimposed induction chemotherapy.
infectious condition (not rarely associated with MM) or • [18F]FDG-PET/CT should be performed at least
possible treatment-related complications (such as vascular 4 weeks after a surgical procedure or 2–3 months
infections), further impacting on patients’ management. after completion of radiotherapy.
However, this high sensitivity associated with the intrinsi-
cally low specificity of [18F]FDG can increase the false-
positive rate when the scan is performed within a short
interval after radiotherapy or surgery. Therefore it is recom- 26.2.4.4 [ 18F]FDG-PET/CT Pitfalls
mended to perform [18F]FDG-PET/CT at least 4 weeks and Limitations
after a surgical procedure or 2–3 months after completion [18F]FDG-PET/CT limitations include the lack of sensitivity
of radiotherapy [172, 182]. for detecting diffuse bone marrow involvement (MRI
a b
Fig. 26.17 Transaxial sections at different levels from a PET/CT scan CT) (reproduced with permission from: Sollini M, Galimberti S, Boni R,
with [18F]FDG in a patient with MM, showing MM lesions character- Erba PA. Diagnostic applications of nuclear medicine: multiple myelo-
ized by different patterns of [18F]FDG uptake, respectively, with mas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
SUVmax = 2.2 in the right clavicula (a) and SUVmax = 4 in the right iliac Oncology – From Pathophysiology to Clinical Applications. New York,
bone (b) (for each column: top, PET; middle, CT; bottom, fused PET/ NY: Springer; 2017: 395–434)
remains the gold standard in this case) and low-density plas- introduced novel interpretative criteria (IMPeTUs) for [18F]
mocyte infiltration. Moreover, the physiological [18F]FDG FDG-PET/CT in MM [184] (Table 26.10). These criteria
uptake in the brain reduces the sensitivity for small lesions in include the description, using a 5-point scale, of the follow-
the skull [183]. ing features:
The major concerns regarding use of [18F]FDG-PET/CT
in MM patients are (1) the lack of standardized interpretation • Metabolic state of the bone marrow
criteria and (2) controversies regarding the use of SUVmax to • Number and site of focal PET-positive lesions with or
define scan positivity. Recently, an Italian panel of experts without osteolytic characteristics
Fig. 26.18 Example of the complementary imaging information myelomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
obtained at different levels with an [18F]FDG-PET/CT scan in a patient Nuclear Oncology – From Pathophysiology to Clinical Applications.
with MM (reproduced with permission from: Sollini M, Galimberti S, New York, NY: Springer; 2017: 395–434)
Boni R, Erba PA. Diagnostic applications of nuclear medicine: multiple
Fig. 26.19 [18F]FDG-PET/CT performed for staging in a patient with PET/CT (reproduced with permission from: Sollini M, Galimberti S,
IgGk-secreting MM (stage IIIA, ISS1). MIP image (on the left) and Boni R, Erba PA. Diagnostic applications of nuclear medicine: multiple
axial images at two different levels of the pelvis (on the right) show myelomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
increased [18F]FDG uptake in the pelvis and right femoral head (for Nuclear Oncology – From Pathophysiology to Clinical Applications.
each level of transaxial section: top, PET; middle, CT; bottom, fused New York, NY: Springer; 2017: 395–434)
• Presence and site of extramedullary disease The visual degree of uptake is defined for bone marrow,
• Presence of paramedullary disease the target lesion (i.e., the hottest area), and extramedullary
• Presence of fractures lesions according to the schema proposed in the Deauville
Table 26.10 IMPeTUs criteria for the interpretation of PET/CT images in MM

Lesion type Site Number of lesion (x) Grading
Diffuse Bone marrowa Deauville 5-point scaleb
Focal Skull x = 1 (no lesions) Deauville 5-point scaleb
Spine x = 2 (1–3 lesions)
Extraspinal (all the rest) x = 3 (4–10 lesions)
x = 4 (>10 lesions)
Lytic x = 1 (no lesions)
x = 2 (1–3 lesions)
x = 3 (4–10 lesions)
x = 4 (>10 lesions)
Fracture At least one at CT
Paramedullaryc At least one
Extramedullary Nodal (n = 7)/extranodal (n = 5)d At least one Deauville 5-point scaleb
(modified from: Nanni C, Zamagni E, Versari A, Chauvie S, Bianchi A, Rensi M, et al. Image interpretation criteria for FDG-PET/CT in multiple
myeloma: a new proposal from an Italian expert panel. IMPeTUs (Italian Myeloma criteria for PET USe). Eur J Nucl Med Mol Imaging.
2016;43:414–21)
a
Appended if increased uptake in limbs and ribs
b
Deauville 5-point scale: 1 = No uptake at all. 2 ≤ Mediastinal blood pool uptake (SUVmax). 3 > Mediastinal blood pool uptake, ≤ liver uptake.
4 > Liver uptake more than 10%. 5 > Liver uptake (twice)
c
Bone lesion involving surrounding soft tissues with bone cortical interruption
d
Nodal disease (N) plus site: laterocervical (LC), supraclavicular (SC), mediastinal (M), axillary (Ax), retroperitoneal (Rp), mesentery (Mes),
inguinal (In); extranodal disease (EN) plus site: liver (Li), muscle (Mus), spleen (Spl), skin (Sk), other (Oth)
criteria for the evaluation of lymphoma patients. The [11C]methionine PET/CT has been tested in MM patients
IMPeTUs criteria are currently under clinical validation in under the assumption that it is incorporated into immuno-
prospective studies evaluating the resulting impact on clini- globulins produced by malignant plasma cells [187].
cal management of patients. Consistently osteolytic lesions exhibited a high [11C]methi-
onine uptake, similar to that of [18F]FDG. Furthermore,
newly diagnosed osteolytic lesions in untreated patients
were strongly [11C]methionine-positive, while recurring
Key Learning Points
osteolytic lesions were [11C]methionine-negative. Although
• The major concerns regarding use of [18F]FDG-
the series evaluated in this study was relatively small, the
PET/CT in MM patients are (1) the lack of stan-
substantial mismatch between the number of lesions identi-
dardized interpretation criteria and (2) controversies
fied by [11C]methionine PET and those identified by CT
regarding the use of SUVmax to define scan
represented an important added value of metabolic imaging
positivity.
for early detection of lesions not yet obvious as macro-
• Recently, novel interpretative criteria (IMPeTUs)
scopic bone changes; these findings would therefore pro-
for [18F]FDG-PET/CT in MM have been
vide an accurate estimate of tumor burden, particularly for
introduced.
clinically silent and unexpected localizations. Very recently,
the superiority of [11C]methionine for staging and restaging
of both intra- and extramedullary MM lesions has been
26.2.4.5 P ET Radiopharmaceuticals Other than reported in a head-to-head comparison between [11C]methi-
[18F]FDG onine and [18F]FDG [188].
Radiolabeled choline (a precursor of phospholipid synthesis PET/CT with 3′-18F-fluoro-3′-deoxy-l-thymidine ([18F]
whose uptake is increased in proliferating cells) and [18F] FLT) has been proposed to distinguish hematological malig-
FDG have been comparatively evaluated in patients with nancies (specifically highly proliferating syndromes such as
MM, suggesting a higher sensitivity of radiolabeled choline myelodysplasia and myeloproliferative disorders) from other
over [18F]FDG for detecting bone lesions due to the high clinical conditions causing pancytopenia. Although the exact
phospholipid metabolic rate of myelomatous cells [185, mechanism of [18F]FLT uptake in bone marrow remains
186]. The main clinical disadvantages of using radiolabeled unclear, several studies suggest that such event is linked to
choline are its high physiological liver uptake that may mask the cycling activity of hematopoietic cells, thus reflecting the
hepatic lesions and, similarly as for [18F]FDG, the possibility state of the entire bone marrow compartment [134]. Bone
of false-positive results due to the concomitant tumors other marrow uptake of [18F]FLT is intense in patients with myelo-
than MM (e.g., bone metastases from prostate cancer). dysplasia and myeloproliferative disorders, while it is
lower-to-
absent in patients with MM, myelofibrosis, or the underlying plasma cell proliferative disorder. Thus, they
aplastic anemia. Patients with MM may present variable recommend that at least one imaging modality among [18F]
degrees of [18F]FLT uptake in response to bone marrow FDG-PET/CT, low-dose whole-body CT, or MRI of the
expansion into the peripheral bones. whole body or spine (depending on local availability and
Extensive skeletal involvement has also been incidentally resources) should be done in case of suspected
detected during PET with [11C]Acetate [189]; however, it is SMM. Increased [18F]FDG uptake alone is not adequate for
not entirely clear why myelomatous lesions would exhibit the diagnosis of MM, as evidence of the underlying osteo-
such high uptake of [11C]Acetate. lytic bone destruction is needed on the CT portion of the
examination. Care should be taken to avoid over-
interpretation: as with skeletal surveys, if there are doubts
Key Learning Points about the nature of these lesions, a repeat study within
• Radiolabeled choline has higher sensitivity than 3–6 months should be done. Such patients might be followed
[18F]FDG for detecting bone lesions, due to the high up closely at 1–3-month intervals before systemic therapy is
phospholipid metabolic rate of myelomatous cells. started. Furthermore, in view of the incorporation and avail-
• Consistently osteolytic lesions exhibit a high [11C] ability of more sensitive imaging modalities to identify
methionine uptake, similar to that of [18F]FDG. osteolytic bone destruction, the IMWG no longer recom-
• Bone marrow uptake of [18F]FLT is intense in mends the presence of osteoporosis or vertebral compression
patients with myelodysplasia and myeloprolifera- fractures alone, in the absence of lytic lesions, as being suf-
tive disorders, while it is lower-to-absent in patients ficient evidence of bone disease as a diagnostic criterion for
with MM, myelofibrosis, or aplastic anemia. MM. However, if vertebral compression fractures are seen in
younger patients with monoclonal gammopathy, judgment
should be exercised, and additional imaging (CT or [18F]
FDG-PET/CT) should be obtained to clarify that the changes
26.2.5 Diagnostic Workup at Diagnosis are not related to myeloma. Overdiagnosis of MM among
elderly patients with MGUS would be highly likely if osteo-
Although there are some issues that need to be clari- porosis and compression fractures alone were regarded as
fied when using sensitive techniques such as WBLD-CT, sufficient to be identified as CRAB features [127].
PET/CT, or MRI, data so far acquired support the use of
WBLD-CT as novel standard procedure for the diagnosis
of lytic disease in patients with MM (grade 1A) as pointed Key Learning Points
out by the “European Myeloma Network Guidelines for the • Despite the considerable amount of favorable
Management of Multiple Myeloma-related Complications” results reported so far, the “European Myeloma
[137] (Fig. 26.20). Network Guidelines for the Management of
Positive lesions in WBLD-CT are considered those with a Multiple Myeloma-related Complications” do not
diameter of 5 mm or more. Conventional radiography can also recommend [18F]FDG-PET/CT for routinary use in
be used if WBLD-CT is not available. In asymptomatic the diagnosis of MM patients.
patients with no lytic disease in WBLD-CT, whole-body MRI • The “Revised International Myeloma Working
(or spine and pelvic MRI if WB-MRI is not available) must be Group diagnostic criteria for multiple myeloma and
performed: in the presence of more than one focal lesion, the smouldering multiple myeloma” recommend that at
patient is classified as having symptomatic disease that needs least one imaging modality among [18F]FDG-PET/
therapy (grade 1A). Despite the considerable amount of favor- CT, low-dose whole-body CT, or MRI of the whole
able results reported so far, the “European Myeloma Network body or spine (depending on local availability and
Guidelines for the Management of Multiple Myeloma-related resources) should be employed in case of suspected
Complications” does not recommend [18F]FDG-PET/CT for SMM.
routinary use in the diagnosis of MM patients, although they
recognize that it may be useful for better definition of com-
plete or stringent complete response (CR or sCR) and for the
progression of the disease (grade 2B) [137]. 26.2.6 Assessment of Response to Treatment
The “Revised International Myeloma Working Group
diagnostic criteria for multiple myeloma and smouldering Limited data are available concerning the value of imag-
multiple myeloma” reported the detection of one or more ing during treatment and follow-up in MM patients [144].
osteolytic lesions on skeletal radiography, CT, or [18F]FDG- Different criteria can be applied to evaluate response
PET/CT among the evidences of end-organ damage due to to treatment in MM patients, including those set by the
Suspected plasmocytoma
or myeloma
Soft tissue mass WBLD-CT/skeletal Suspected spinal

survey* cord compression
CT scan +/- biopsy

Lytic lesion Urgent MRI/CT scan + appropriate
medical management
Yes No
Risk for fracture WB-MRI or

spinal/pelvic MRI^
Yes No
Focal lesion > 1 0-1 focal lesions normal
or diffuse pattern
Urgent orthopedic Systemic

review +/- radiotherapy therapy
or surgery
Watch-and-wait if no
other criteria of
*If WBLD-CT not available symptomatic MM
^lf WB-MRI not available
Fig. 26.20 Suggested algorithm for the use of imaging in patients with for the management of multiple myeloma-related complications.
MM (modified from: Terpos E, Kleber M, Engelhardt M, Zweegman Haematologica. 2015;100:1254–66)
S, Gay F, Kastritis E, et al. European myeloma network guidelines
European Group for Blood and Bone Marrow Transplant/ with relapsed refractory MM, while specific criteria for
International Bone Marrow Transplant Registry/American progression have been defined for SMM (Table 26.12).
Bone Marrow Transplant Registry, by the Chronic Regardless of the specific system employed to classify
Leukemia-Myeloma Task Force, by the Southwest response to treatment, the main pitfall of these systems is the
Oncology Group (SWOG), and by the Eastern Cooperative method used to assess bone marrow and bone lesions. In par-
Oncology Group (ECOG). In 2006, the International ticular, plain X-ray is of limited value; even if the appearance
Myeloma Working Group (IMWG) recognized the impor- of new bone lesions clearly indicates disease progression,
tance of standardizing such criteria and elaborated the so- lytic bone lesions rarely show radiologic signs of recovery,
called uniform response criteria (Table 26.11); in addition and new compression vertebral fractures may occur either in
to the classical categories (i.e., CR, PR, SD, PD), they intro- case of disease progression or in case of bone loss or in case
duced the new category of “positive response of additional of reduction of the tumor mass that supports the bone cortex
clinical significance” following treatment. The category [126, 134, 167].
“minor response” has been deleted in the IMWG response Similarly, MDCT is not generally used since osteolytic
criteria, since it has been considered to be somewhat lesions rarely recover. In selected cases, characterized by
vague, therefore unreliable. However, more recently this significant soft tissue component, favorable response to ther-
category has been reintroduced by the Joint Symposium apy can be demonstrated by CT (decrease in tumor size). CT
of the American Society of Hematology and the Food and may be of value also in patients with persistent unexplained
Drug Administration on New Drug Approvals for patients symptoms despite therapy, or with risk of impending
Table 26.11 International Myeloma Working Group uniform response criteria for MM
Category Criteria
Complete response Negative immunofixation of serum and urine and disappearance of any soft tissue plasmacytomas and <5% plasma
(CR) cells in bone marrow
Patients with only measurable disease by serum FLC levels: CR is defined as normal FLC ratio (0.26–1.65) in
addition to CR criteria listed above
Stringent complete Negative immunofixation of serum and urine and disappearance of any soft tissue plasmacytomas and <5% plasma
response (sCR) cells in bone marrow and normal FLC ratio and absence of clonal cells in bone marrow by immunohistochemistry
or immunofluorescence
Very good partial Serum and urine M-protein detectable by immunofixation but not on electrophoresis or ≥90% or greater reduction
response (VGPR) in serum M-protein plus urine M-protein <100 mg/24 h
Note: Patients with only measurable disease by serum FLC levels: VGPR is defined as a >90% decrease in the
difference between involved and uninvolved FLC levels
Partial response (PR) ≥50% reduction of serum M-protein and reduction in 24 h urinary M-protein by ≥90% or to <200 mg/24 h and
≥50% reduction in the size of soft tissue plasmacytomas if present at baseline
Note: If the serum and urine M-protein are unmeasurable: ≥50% decrease in the difference between involved and
uninvolved FLC levels
If serum and urine M-protein and serum FLC are unmeasurable: ≥50% reduction in bone marrow plasma cells,
provided baseline percentage was ≥30%
Stable disease (SD) Not meeting criteria for CR, VGPR, PR or PD
Progressive disease >25% from lowest response value in any one or more of the following:
(PD) Serum M-protein (absolute increase must be ≥0.5 g/dL)a
Urine M-protein (absolute increase must be ≥200 mg/24 h)
Only in patients without measurable serum and urine M-protein levels: the difference between involved and
uninvolved
FLC levels (absolute increase must be >10 mg/L)
Bone marrow plasma cell percentage (absolute % must be ≥10%)
Definite development of new bone lesions/soft tissue plasmacytomas or definite increase in size of existing bone
lesions/soft tissue plasmacytomas
Development of hypercalcemia (corrected serum calcium >11.5 mg/dL) ascribable solely to the plasma cell
proliferative disorder
All response categories (CR, sCR, VGPR, and PR) require two consecutive assessments made at any time before the institution of any new therapy;
complete, PR, and SD categories also require no known evidence of progressive or new bone lesions if radiographic studies were performed.
Radiographic studies are not required to satisfy these response requirements. Bone marrow assessments need not be confirmed
(modified from: Kyle RA, Rajkumar V. Criteria for diagnosis, staging, risk stratification and response assessment of multiple myeloma. Leukemia.
2009;23:3–9)
a
For PD serum M-protein increases of ≥1 g/dL are sufficient to define relapse if starting M-protein is ≥5 g/dL
Table 26.12 Additional response criteria for specific disease stages recommended by the International Myeloma Working Group uniform
response criteria for MM
Category Criteria
Relapsed MM At least one prior regimen and not meeting criteria for relapsed and refractory MM
Relapsed and refractory MM Relapse of disease while on salvage therapy or progression within 60 days of most recent therapy
Minor response (MR) in patients with ≥25% but <49% reduction of serum M-protein and reduction in 24-h urine M-protein by 50–89%,
relapsed refractory MM which still exceeds 200 mg/24 h and reduction in size (25–49%) of any soft tissue plasmacytoma and
no increase in size/number of lytic bone lesions (development of compression fracture does not
exclude response)
Progression to active MM in patients Evidence of PD based on the IMWG criteria and any one or more of the following:
with smoldering MM Development of new lesion (soft tissue plasmacytoma/bone lesions)
Hypercalcemia (>11 mg/dL)
Decrease in hemoglobin of ≥2 g/dL
Serum creatinine level ≥2 mg/dL
(modified from: Kyle RA, Rajkumar V. Criteria for diagnosis, staging, risk stratification and response assessment of multiple myeloma. Leukemia.
2009;23:3–9)
fracture, or when no clinico-biochemical response to treat- response to therapy is associated with changes in the MRI
ment is observed. pattern (from diffuse to focal/variegate) or with a reduction
The disappearance of any MRI-detectable bone marrow of the signal on the T2-weighted spin echo images. MRI has
lesion typically represents complete response. A partial limited value for the early assessment of response to
treatment, since complete resolution of bone marrow lesions has shown high specificity (86%) in patients in CR, whereas
may take as long as 9–12 months from the completion of the sensitivity was 45% in presence of PR; these results empha-
therapy. DCE-MRI is superior to standard MRI; favorable size the high performance of 99mTc-sestamibi scintigraphy
response is indicated by the disappearance of gadolinium for identifying absent disease, whereas its sensitivity is lower
enhancement or by the absence of contrast-induced enhance- for the definition of residual disease when response is not
ment in a follow-up study as compared to the baseline scan. complete.
DCE-MRI derives parameters correlated with microvascular The prognostic value of 99mTc-sestamibi imaging for pre-
density in the bone marrow both before and after therapy dicting recurrence of the disease has been extensively inves-
[190, 191]. However, the real added value of this technique tigated by Pace et al. [163]. None of the patients responding
seems to lie in the possibility to predict which patients might to therapy had a posttreatment 99mTc-sestamibi scan showing
benefit from therapy with anti-angiogenetic drugs, e.g., bev- focal uptake or diffuse uptake with a summed score higher
acizumab. Nevertheless, this approach has not yet gained a than 2, whereas the non-responders showed either diffuse
definite role in the evaluation of the course of the disease and bone marrow uptake or areas of focal uptake with a summed
of activity of the myeloma lesions. Preliminary reports sug- score higher than 2. A negative baseline 99mTc-sestamibi
gest that DWI MRI may be used for better definition of imaging showed high prognostic value for clinico-
response to therapy, although this data must be confirmed in biochemical remission. Furthermore, 99mTc-sestamibi imag-
larger studies and in comparison with [18F]FDG-PET/CT ing patterns at follow-up were significantly associated with
[152]. The combination of DWI and DCE-MRI produced a the clinical status evaluated after chemotherapy. A semi-
score with a high agreement (76%) with the IMWG criteria quantitative parameter of 99mTc-sestamibi washout derived
of response in 27 patients with myeloma under treatment from dual point scintigraphic acquisitions has been also
[192]; however, more data are needed to evaluate such score employed to distinguish patients responding to treatment
in larger studies [144]. from non-responders. A fast clearance of 99mTc-sestamibi
Given the heterogeneous nature of the disease and incon- from the bone marrow of patients with MM is associated
sistent response to treatment, along with the relatively small with poor response to subsequent chemotherapy, a finding
number of published studies in this field, MRI has currently most likely related to the expression of the P-glycoprotein
a limited role in treatment response assessment. In general, (Pgp) [198]. Since overexpression of Pgp is one of the pri-
MRI findings of response to therapy (CR or PR) include res- mary mechanisms of MDR in MM as well as in other malig-
olution or decrease of the extent of focal lesions or diffuse nancies [199, 200], it can be assumed that the 99mTc-sestamibi
patterns after conventional chemotherapy or novel anti- washout rate could serve to predict response to chemother-
myeloma agents [193–195]. In a small study with 33 patients apy in MM patients.
who underwent ASCT, MRI resulted in high specificity and However, sufficient data are not available on the impact of
accuracy (86% and 79%, respectively) but only 64% sensi- 99m
Tc-sestamibi scintigraphy on progression-free survival
tivity for the detection of remission because of false-positive and overall survival.
results of nonviable lesions [195]. In case of effective response to therapy, [18F]FDG uptake
Thus, for evaluation of remission, functional imaging declines rapidly (within hours). The persistence of a positive
techniques (e.g., PET/CT or possibly DWI MRI) seem to be [18F]FDG-PET has high prognostic value, as it correlates
more appropriate. Nonetheless, in a retrospective study with early recurrence. Furthermore, [18F]FDG-PET/CT rep-
which compared MRI and [18F]FDG-PET/CT in 210 patients resents the most sensitive imaging technique to evaluate
with MM, MRI achieved better results than PET/CT for the response of the bone lesions to treatment, particularly for
detection of myeloma at diagnosis and at progression, distinguishing treated bone marrow lesions from viable neo-
whereas PET/CT detected response to anti-myeloma therapy plastic tissues [126, 134, 179] (Figs. 26.21 and 26.22).
earlier than MRI [196]. [18F]FDG-PET/CT has been employed in patients with
In patients with SMM and a negative MRI scan at diagno- newly diagnosed MM as part of the “Total Therapy 3” study
sis, the timing of follow-up with MRI and the significance of [181]. The scan was obtained at baseline (together with met-
any findings on these studies have not yet been established. astatic bone survey and MRI) and within 10 days after start-
Instead, a close correlation between the results of 99mTc- ing the first induction cycle of chemotherapy. The follow-up
sestamibi scintigraphy (i.e., disappearance of any area of studies included annual metastatic bone survey, while MRI
uptake) and the conventional criteria used for defining clini- scans were performed before each of two transplantations,
cal response to therapy (i.e., measurement of the M-protein before consolidation and maintenance phases, and semian-
component) has been found [167, 197]. Overall, these find- nually thereafter. All imaging studies were also repeated at
ings support the hypothesis that bone marrow uptake is an relapse. Disappearance of [18F]FDG uptake before
indicator of myeloma activity in bone marrow, even in the transplantation proved to be an independent favorable prog-
absence of focal areas of increased uptake. This procedure nostic factor, demonstrating the importance of complete
Fig. 26.21 Use of [18F]

FDG-PET to assess response a b
to treatment in a patient with
IgGk-secreting MM (stage
IIIb, ISS2). The baseline MIP
image (a) shows multiple
sites of increased [18F]FDG
uptake throughout the
skeleton that resolve virtually
completely after
chemotherapy with
bortezomib and liposomal
doxorubicin (b) (reproduced
with permission from: Sollini
M, Galimberti S, Boni R,
Erba PA. Diagnostic
applications of nuclear
medicine: multiple myelomas.
In: Strauss HW, Mariani G,
Applications. New York, NY:
Springer; 2017: 395–434)
suppression of tumor metabolism in myeloma. While very More recently, 700 patients with symptomatic MM eligible
early [18F]FDG-PET/CT changes (at 10 days) did not affect for high-dose therapy (HDT) have been randomized to
outcome, complete [18F]FDG suppression of extramedullary receive either bortezomib-lenalidomide-dexamethasone
disease before transplantation conferred superior overall and followed by 1-year maintenance with lenalidomide or
event-free survival. Complete [18F]FDG suppression before Revlimid + Velcade + dexamethasone (VRD) followed by
the first transplantation correlated with better outcomes in high-dose therapy and autologous transplantation and then
both low-risk and high-risk gene array-defined risk groups, VRD consolidation and 1-year lenalidomide maintenance.
reaching statistical significance for overall survival in low- Spine and pelvis MRI and whole-body PET/CT at diagnosis,
risk and for event-free survival in high-risk myeloma. after 3 cycles of VRD, and prior to maintenance were per-
Normalization of the PET findings before transplantation formed. At diagnosis, MRI was positive in 94.7% of patients,
was associated with improved outcomes, whereas gene and PET/CT in 91% of cases. Interim-MRI remained positive
array-defined high-risk designation conferred poor overall in 93%, while interim-PET/CT in 55% of patients.
and event-free survival. In the absence of molecular genetic Normalization of MRI after 3 months of induction therapy did
data, the favorable implications of pre-transplantation [18F] not predict neither progression-free survival nor overall sur-
FDG suppression were canceled by elevated serum levels of vival, whereas normalization of PET/CT was associated with
lactate dehydrogenase and β2-microglobulin. Clinical CR or a significant improvement in PFS but not OS. Prior to mainte-
n-CR status before transplantation did not impact post-trans- nance, MRI remained positive in 83%, and PET/CT in 21% of
plantation survival outcomes. The prognosis of high-risk the patients. Normalization of MRI before maintenance was
patients not achieving complete [18F]FDG suppression was not predictive for survivals, whereas normalization of PET/CT
especially poor. was associated with a prolonged PFS and OS, thus suggesting
[18F]FDG-PET/CT scan. The probability of progression to

symptomatic MM within 2 years was 75% for patients with
positive PET/CT and 30% for patients with negative PET/CT,
with a median TTP of 21 months versus 60 months, respec-
tively (P = 0.0008). This probability was higher if hypermeta-
bolic activity was combined with underlying osteolysis (87%
progression rate at 2 years versus 61% in patients with positive
[18F]FDG-PET/CT but without underlying osteolysis) [206].
[18F]FDG-PET/CT has a role also in nonsecretory or oli-
gosecretory myeloma for the detection of active lesions in
the body [207] (Fig. 26.23). In this setting, it may be useful
also to monitoring treatment response.
PET/MRI is a novel imaging modality with attractive
potential clinical applications [144]. Regarding MM, there is
only one prospective study comparing [18F]FDG-PET/MRI
head-to-head with PET/CT in 30 patients with MM. There
was high correlation between the two techniques regarding
both the number of metabolically active lesions and the
degree of uptake, as evaluated by SUV [208].
Key Learning Points

• The main pitfall of the available systems used for
treatment assessment in MM is the method used to
assess bone marrow and bone lesions.
• MDCT is generally not reliable for MM treatment
assessment, since osteolytic lesions rarely recover.
• MRI has limited value for the early assessment of
response to treatment, since complete resolution of
bone marrow lesions may take as long as
Fig. 26.22 [18F]FDG-PET/CT (MIP image) obtained after disease 9–12 months from the completion of therapy.
recurrence in the same patient as in Fig. 26.21 (reproduced with permis- • 99mTc-sestamibi scintigraphy has been used in the
sion from: Sollini M, Galimberti S, Boni R, Erba PA. Diagnostic appli- past for treatment assessment in MM, but it is cur-
cations of nuclear medicine: multiple myelomas. In: Strauss HW, rently replaced by [18F]FDG.
Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
Pathophysiology to Clinical Applications. New York, NY: Springer; • In case of effective response to therapy, [18F]FDG
2017: 395–434) uptake declines rapidly, whereas the persistence of
a positive [18F]FDG-PET has high prognostic value,
as it correlates with early recurrence.
that PET/CT should be incorporated in the follow-up of MM • In SMM 16% of patients with normal whole-body
patients who undergo high-dose treatment [201]. X-ray had positive PET/CT results.
Quantitative data on kinetics and distribution patterns of • [18F]FDG-PET/CT has a role also in nonsecretory
[18F]FDG obtained from dynamic PET/CT in 40 patients or oligosecretory myeloma for the detection of
with MM correlated significantly with percentage of bone active lesions throughout the body.
marrow infiltration by plasma cells [202]. Furthermore, PET/ • PET/MRI is a novel imaging modality with attrac-
CT efficiently detected extramedullary disease in patients tive potential clinical applications.
both at diagnosis and at relapse [203].
In SMM 16% of patients with normal whole-body X-ray
had positive PET/CT results [204]. The median time to pro-
gression for patients with positive PET/CT was shorter than 26.2.7 Radionuclide Therapy of Multiple
for patients with negative PET/CT (1.1 years versus 4.5 year), Myeloma
while the probability of progression at 2 years for patients with
a positive PET/CT scan was 58% [205]. In a study involving There have been several attempts at developing radioimmu-
188 patients with SMM, 39% of patients had a positive notherapy for MM, but very few of the radiopharmaceutical
Fig. 26.23 Variable patterns in the same individual of [18F]FDG uptake bones (b) (reproduced with permission from: Sollini M, Galimberti S,
in a patient with oligosecretory/nonsecretory MM. The PET/CT scan Boni R, Erba PA. Diagnostic applications of nuclear medicine: multiple
shows focally increased [18F]FDG uptake at the site of MM localization myelomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
in the vertebral body of L3 (a). Whereas, no abnormally increased [18F] Nuclear Oncology – From Pathophysiology to Clinical Applications.
FDG uptake is detectable at the sites of osteolytic lesions in the pelvic New York, NY: Springer; 2017: 395–434)
tested have moved from the experimental setting (either therapy with 177Lu- or 90Y-labeled peptides is being contem-
in vitro or in animal models) to the phase of pilot/prelimi- plated [220, 221].
nary clinical investigation [209].
Anti-CD45, anti-CD66, and anti-CD138 monoclonal
antibodies have been explored to this purpose, mostly as
conditioning treatment for achieving myeloablation in Key Learning Point
patients with MM [210, 211]. Alternative approaches to • There have been several attempts at developing
achieve myeloablation have been based on the use of bone- radioimmunotherapy for MM, but very few of the
seeking radiopharmaceuticals such as 153Sm-EDTMP [212– radiopharmaceuticals tested have moved from the
214] and 166Ho-DOTMP [209]. experimental setting to clinical validation.
The potential of RIT with MM-specific monoclonal anti-
bodies has also been explored, using either 213Bi, 131I, or 90Y
as emitters of α++ or β− particles with cell-killing properties
[209]. Anti-CD38 antibodies have been used in most of these References
investigations. Other pilot investigations in MM patients
1. Hodgkin T. On some morbid appearances of the absorbent glands
have been based on the use of the anti-CD138 monoclonal and spleen. Med Chir Trans. 1832;17:68–114.
preparation 131I-B-B4 [215] and of a 131I-labeled antibody 2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer
against the oncofetal ED-B-containing fibronectin isoform— J Clin. 2017;67(1):7–30.
3. Horning SJ. Hodgkin’s lymphoma. In: Abeloff M, Armitage J,
which is abundantly expressed in the extracellular matrix of
Niederhuber J, Kastan MB, McKenna WG, editors. Abeloff’s clini-
several tumor types [216, 217]. cal oncology. 4th ed. Philadelphia: Churchill Livingstone Elsevier;
Encouraging results have also been obtained with the 2008. p. 2353–70.
use of peptides that target the chemokine receptor-4 4. Patel P, Hanson DL, Sullivan PS, et al. Incidence of types of
cancer among HIV-infected persons compared with the general
(CXCR4), which is often overexpressed on the surface of
population in the United States, 1992–2003. Ann Intern Med.
myeloma cells [218]. For therapy, the CXCR4-targeting 2008;148:728–36.
peptides can be labeled with either 177Lu or 90Y [219]. This 5. Cotran RS, Kumar V, Robbins SL. Diseases of white cells, lymph
approach is especially interesting also from the point of nodes, and spleen. In: Cotran RS, Kumar V, Robbins SL, edi-
tors. Pathologic basis of disease. 5th ed. Philadelphia: Saunders
view of theranostics, since the peptide can be labeled with
Company; 1994.
68
Ga for imaging with PET/CT—as a preliminary investiga- 6. Cheson BD, Fisher RI, Barrington SF, Cavalli F, Schwartz
tion for assessing the degree of CXCR4 expression in LH, Zucca E, et al. Recommendations for initial evaluation,
myeloma lesions as well as in other cancers for which staging, and response assessment of Hodgkin and non-Hodgkin
lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27): 23. Friedberg JW, Fischman A, Neuberg D, Kim H, Takvorian T, Ng
3059–68. AK, et al. FDG-PET is superior to gallium scintigraphy in stag-
7. Barrington SF, Mikhaeel NG, Kostakoglu L, Meignan M, ing and more sensitive in the follow-up of patients with de novo
Hutchings M, Mueller SP, et al. Role of imaging in the staging and Hodgkin lymphoma: a blinded comparison. Leuk Lymphoma.
response assessment of lymphoma: consensus of the International 2004;45(1):85–92.
Conference on Malignant Lymphomas Imaging Working Group. J 24. Tsukamoto N, Kojima M, Hasegawa M, Oriuchi N, Matsushima
Clin Oncol. 2014;32(27):3048–58. T, Yokohama A, et al. The usefulness of 18F-fluorodeoxyglucose
8. Hasenclever D, Dihel V. A prognostic score for advanced Hodgkin’s positron emission tomography (18F-FDG-PET) and a comparison of
disease. International Prognostic Factors Project on Advanced 18
F-FDG-PET with 67gallium scintigraphy in the evaluation of lym-
Hodgkin’s Disease. N Engl J Med. 1998;339(21):1506–14. phoma: relation to histologic subtypes based on the World Health
9. Morton LM, Wang SS, Devesa SS, Hartge P, Weisenburger DD, Organization classification. Cancer. 2007;110(3):652–9.
Linet MS. Lymphoma incidence patterns by WHO subtype in the 25. Cheson BD, Horning SJ, Coiffier B, Shipp MA, Fisher RI, Connors
United States, 1992–2001. Blood. 2006;107(1):265–76. JM, et al. Report of an international workshop to standardize
10. Shimabukuro-Vornhagen A, Haverkamp H, Engert A, Balleisen L, response criteria for non-Hodgkin’s lymphomas. NCI Sponsored
Majunke P, Heil G, et al. Lymphocyte-rich classical Hodgkin’s lym- International Working Group. J Clin Oncol. 1999;17(4):1244.
phoma: clinical presentation and treatment outcome in 100 patients 26. Lister TA, Crowther D, Sutcliffe SB, Glatstein E, Canellos GP,
treated within German Hodgkin’s Study Group trials. J Clin Oncol. Young RC, et al. Report of a committee convened to discuss
2005;23:5739–45. the evaluation and staging of patients with Hodgkin’s disease:
11. Xing KH, Connors JM, Lai A, Al-Mansour M, Sehn LH, Villa D, Cotswolds meeting. J Clin Oncol. 1989;7(11):1630–6.
et al. Advanced-stage nodular lymphocyte predominant Hodgkin 27. Weiler-Sagie M, Bushelev O, Epelbaum R, Dann EJ, Haim N, Avivi
lymphoma compared with classical Hodgkin lymphoma: a matched I, et al. 18F-FDG avidity in lymphoma readdressed: a study of 766
pair outcome analysis. Blood. 2014;123(23):3567–73. patients. J Nucl Med. 2010;51(1):25–30.
12. Engert A, Plütschow A, Eich HT, Lohri A, Dörken B, Borchmann 28. Boellaard R, O’Doherty MJ, Weber WA, Mottaghy FM, Lonsdale
P, et al. Reduced treatment intensity in patients with early-stage MN, Stroobants SG, et al. FDG PET and PET/CT: EANM proce-
Hodgkin’s lymphoma. N Engl J Med. 2010;363:2220–7. dure guidelines for tumour PET imaging: version 1.0. Eur J Nucl
13. Specht L, Yahalom J, Illidge T, Berthelsen AK, Constine LS, Eich Med Mol Imaging. 2010;37(1):181–200.
HT, et al. Modern radiation therapy for Hodgkin lymphoma: field 29. Schoder H, Noy A, Gonen M, Weng L, Green D, Erdi YE, et al.
and dose guidelines from the International Lymphoma Radiation Intensity of 18fluorodeoxyglucose uptake in positron emission
Oncology Group (ILROG). Int J Radiat Oncol Biol Phys. tomography distinguishes between indolent and aggressive non-
2014;89:854–62. Hodgkin’s lymphoma. J Clin Oncol. 2005;23(21):4643–51.
14. Wirth A, Yuen K, Barton M, Roos D, Gogna K, Pratt G, et al. 30. Michallet AS, Sesques P, Rabe KG, Itti E, Tordot J, Tychyj-Pinel C,
Long-term outcome after radiotherapy alone for lymphocyte- et al. An 18F-FDG-PET maximum standardized uptake value > 10
predominant Hodgkin lymphoma: a retrospective multicenter study represents a novel valid marker for discerning Richter’s Syndrome.
of the Australasian Radiation Oncology Lymphoma Group. Cancer. Leuk Lymphoma. 2016;57(6):1474–7.
2005;104(6):1221–9. 31. Bodet-Milin C, Kraeber-Bodere F, Moreau P, Campion L, Dupas
15. Hoppe RT, Advani RH, Ai WZ, Ambinder RF, Aoun P, Bello
B, Le Gouill S. Investigation of FDG-PET/CT imaging to guide
CM, et al. Hodgkin Lymphoma Version 1.2017, NCCN clini- biopsies in the detection of histological transformation of indolent
cal practice guidelines in oncology. J Natl Compr Cancer Netw. lymphoma. Haematologica. 2008;93(3):471–2.
2017;15(5):608–38. 32. Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout
16. Jaffe ES. The 2008 WHO classification of lymphomas: implica- RG, Granton P, et al. Radiomics: extracting more information from
tions for clinical practice and translational research. Hematology medical images using advanced feature analysis. Eur J Cancer.
Am Soc Hematol Educ Program. 2009;2009:523–31. 2012;48(4):441–6.
17. Kridel R, Mottok A, Farinha P, Ben-Neriah S, Ennishi D, Zheng 33. Aerts HJ, Velazquez ER, Leijenaar RT, Parmar C, Grossmann P,
Y, et al. Cell of origin of transformed follicular lymphoma. Blood. Carvalho S, et al. Decoding tumour phenotype by noninvasive
2015;126(18):2118–27. imaging using a quantitative radiomics approach. Nat Commun.
18. Swerdlow SH, Campo E, Pileri SA, Harris NL, Stein H, Siebert R, 2014;5:4006.
et al. The 2016 revision of the World Health Organization classifi- 34. Lartizien C, Rogez M, Niaf E, Ricard F. Computer-aided stag-
cation of lymphoid neoplasms. Blood. 2016;127:2375–90. ing of lymphoma patients with FDG PET/CT imaging based on
19. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero textural information. IEEE J Biomed Health Inform. 2014;18(3):
G, Dohner H, et al. Guidelines for the diagnosis and treatment 946–55.
of chronic lymphocytic leukemia: a report from the International 35. Bodet-Milin C, Eugene T, Gastinne T, Bailly C, Le Gouill S, Dupas
Workshop on Chronic Lymphocytic Leukemia updating the B, et al. The role of FDG-PET scanning in assessing lymphoma in
National Cancer Institute-Working Group 1996 guidelines. Blood. 2012. Diagn Interv Imaging. 2013;94(2):158–68.
2008;111(12):5446–56. 36. Schoder H, Meta J, Yap C, Ariannejad M, Rao J, Phelps ME, et al.
20. International CLL-IPI working group. An international prognos- Effect of whole-body 18F-FDG PET imaging on clinical staging and
tic index for patients with chronic lymphocytic leukaemia (CLL- management of patients with malignant lymphoma. J Nucl Med.
IPI): a meta-analysis of individual patient data. Lancet Oncol. 2001;42(8):1139–43.
2016;17(6):779–90. 37. Karam M, Novak L, Cyriac J, Ali A, Nazeer T, Nugent F. Role of
21. Zhou Z, Sehn LH, Rademaker AW, Gordon LI, Lacasce AS,
fluorine-18 fluoro-deoxyglucose positron emission tomography
Crosby-Thompson A, et al. An enhanced International Prognostic scan in the evaluation and follow-up of patients with low-grade
Index (NCCN-IPI) for patients with diffuse large B-cell lymphoma lymphomas. Cancer. 2006;107(1):175–83.
treated in the rituximab era. Blood. 2014;123(6):837–42. 38. El-Galaly TC, d’Amore F, Mylam KJ, de Nully Brown P, Bogsted
22. Horwitz SM, Zelenetz AD, Gordon LI, Wierda WG, Abramson M, Bukh A, et al. Routine bone marrow biopsy has little or no ther-
JS, Advani RH, et al. NCCN guidelines insights: non-Hodgkin’s apeutic consequence for positron emission tomography/computed
lymphomas, Version 3.2016. J Natl Compr Cancer Netw. tomography-staged treatment-naive patients with Hodgkin lym-
2016;14(9):1067–79. phoma. J Clin Oncol. 2012;30(36):4508–14.
39. Adams HJ, Kwee TC, de Keizer B, Fijnheer R, de Klerk JM, computed tomography criteria better predict response than each
Nievelstein RA. FDG PET/CT for the detection of bone mar- test alone? Leuk Lymphoma. 2012;53(11):2143–50.
row involvement in diffuse large B-cell lymphoma: system- 55. Mikhaeel NG, Timothy AR, O’Doherty MJ, Hain S, Maisey MN.
atic review and meta-analysis. Eur J Nucl Med Mol Imaging. 18-FDG-PET as a prognostic indicator in the treatment of aggres-
2014;41(3):565–74. sive Non-Hodgkin’s Lymphoma-comparison with CT. Leuk
40. Nakamoto Y, Cohade C, Tatsumi M, Hammoud D, Wahl RL. CT Lymphoma. 2000;39(5–6):543–53.
appearance of bone metastases detected with FDG PET as part of 56. Kasamon YL, Wahl RL, Ziessman HA, Blackford AL, Goodman
the same PET/CT examination. Radiology. 2005;237(2):627–34. SN, Fidyk CA, et al. Phase II study of risk-adapted therapy of
41. Juweid ME, Wiseman GA, Vose JM, Ritchie JM, Menda Y, newly diagnosed, aggressive non-Hodgkin lymphoma based on
Wooldridge JE, et al. Response assessment of aggressive non- midtreatment FDG-PET scanning. Biol Blood Marrow Transplant.
Hodgkin’s lymphoma by integrated International Workshop 2009;15(2):242–8.
Criteria and fluorine-18-fluorodeoxyglucose positron emission 57. Barrington SF, Qian W, Somer EJ, Franceschetto A, Bagni
tomography. J Clin Oncol. 2005;23(21):4652–61. B, Brun E, et al. Concordance between four European cen-
42. Cheson BD, Pfistner B, Juweid ME, Gascoyne RD, Specht L, tres of PET reporting criteria designed for use in multicentre
Horning SJ, et al. Revised response criteria for malignant lym- trials in Hodgkin lymphoma. Eur J Nucl Med Mol Imaging.
phoma. J Clin Oncol. 2007;25(5):579–86. 2010;37(10):1824–33.
43. Meignan M, Gallamini A, Meignan M, Gallamini A, Haioun 58. Barrington SF, Kirkwood AA, Franceschetto A, Fulham MJ,
C. Report on the First International Workshop on Interim-PET- Roberts TH, Almquist H, et al. PET-CT for staging and early
Scan in Lymphoma. Leuk Lymphoma. 2009;50(8):1257–60. response: results from the Response-Adapted Therapy in Advanced
44. Biggi A, Gallamini A, Chauvie S, Hutchings M, Kostakoglu L, Hodgkin Lymphoma study. Blood. 2016;127(12):1531–8.
Gregianin M, et al. International validation study for interim PET 59. Gallamini A, Barrington SF, Biggi A, Chauvie S, Kostakoglu L,
in ABVD-treated, advanced-stage Hodgkin lymphoma: interpre- Gregianin M, et al. The predictive role of interim positron emis-
tation criteria and concordance rate among reviewers. J Nucl Med. sion tomography for Hodgkin lymphoma treatment outcome is
2013;54(5):683–90. confirmed using the interpretation criteria of the Deauville five-
45. Itti E, Meignan M, Berriolo-Riedinger A, Biggi A, Cashen AF, point scale. Haematologica. 2014;99(6):1107–13.
Vera P, et al. An international confirmatory study of the prognostic 60. Shankar LK, Hoffman JM, Bacharach S, Graham MM, Karp J,
value of early PET/CT in diffuse large B-cell lymphoma: com- Lammertsma AA, et al. Consensus recommendations for the use
parison between Deauville criteria and DeltaSUVmax. Eur J Nucl of 18F-FDG PET as an indicator of therapeutic response in patients
Med Mol Imaging. 2013;40(9):1312–20. in National Cancer Institute Trials. J Nucl Med. 2006;47(6):
46. Juweid ME, Stroobants S, Hoekstra OS, Mottaghy FM, Dietlein 1059–66.
M, Guermazi A, et al. Use of positron emission tomography for 61. Boellaard R, Oyen WJ, Hoekstra CJ, Hoekstra OS, Visser EP,
response assessment of lymphoma: consensus of the Imaging Willemsen AT, et al. The Netherlands protocol for standardisa-
Subcommittee of International Harmonization Project in tion and quantification of FDG whole body PET studies in multi-
Lymphoma. J Clin Oncol. 2007;25:571–8. centre trials. Eur J Nucl Med Mol Imaging. 2008;35(12):2320–33.
47. Spaepen K, Stroobants S, Dupont P, Thomas J, Vandenberghe 62. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to
P, Balzarini J, et al. Can positron emission tomography with PERCIST: evolving considerations for PET response criteria in
[18F]-fluorodeoxyglucose after first-line treatment distinguish solid tumors. J Nucl Med. 2009;50(Suppl 1):122S–50S.
Hodgkin’s disease patients who need additional therapy from oth- 63. Ceriani L, Suriano S, Ruberto T, Zucca E, Giovanella L. 18F-FDG
ers in whom additional therapy would mean avoidable toxicity? uptake changes in liver and mediastinum during chemotherapy
Br J Haematol. 2001;115(2):272–8. in patients with diffuse large B-cell lymphoma. Clin Nucl Med.
48. Kasamon YL, Jones RJ, Wahl RL. Integrating PET and PET/ 2012;37(10):949–52.
CT into the risk-adapted therapy of lymphoma. J Nucl Med. 64. Bagci U, Yao J, Miller-Jaster K, Chen X, Mollura DJ. Predicting
2007;48(Suppl 1):19S–27S. future morphological changes of lesions from radiotracer uptake
49. Engles JM, Quarless SA, Mambo E, Ishimori T, Cho SY, Wahl in 18F-FDG-PET images. PLoS One. 2013;8(2):e57105.
RL. Stunning and its effect on 3H-FDG uptake and key gene 65. Ben Bouallègue F, Tabaa YA, Kafrouni M, Cartron G, Vauchot F,
expression in breast cancer cells undergoing chemotherapy. J Nucl Mariano-Goulart D. Association between textural and morpho-
Med. 2006;47(4):603–8. logical tumor indices on baseline PET-CT and early metabolic
50. Castellucci P, Zinzani P, Nanni C, Farsad M, Moretti A, Alinari L, response on interim PET-CT in bulky malignant lymphomas. Med
et al. 18F-FDG PET early after radiotherapy in lymphoma patients. Phys. 2017;44(9):4608–19.
Cancer Biother Radiopharm. 2004;19(5):606–12. 66. Hanaoka K, Hosono M, Tatsumi Y, Ishii K, Im SW, Tsuchiya N,
51. Jacene HA, Filice R, Kasecamp W, Wahl RL. 18F-FDG PET/CT et al. Heterogeneity of intratumoral 111In-ibritumomab tiuxetan
for monitoring the response of lymphoma to radioimmunotherapy. and 18F-FDG distribution in association with therapeutic response
J Nucl Med. 2009;50(1):8–17. in radioimmunotherapy for B-cell non-Hodgkin’s lymphoma.
52. Horning SJ, Juweid ME, Schoder H, Wiseman G, McMillan A, EJNMMI Res. 2015;5:10.
Swinnen LJ, et al. Interim positron emission tomography scans 67. Skipper HE, Schabel FM Jr, Wilcox WS. Experimental evaluation
in diffuse large B-cell lymphoma: an independent expert nuclear of potential anticancer agents. XIII. On the criteria and kinetics
medicine evaluation of the Eastern Cooperative Oncology Group associated with “curability” of experimental leukemia. Cancer
E3404 study. Blood. 2010;115(4):775–7; quiz 918. Chemother Rep. 1964;35:1–111.
53. Torizuka T, Zasadny KR, Kison PV, Rommelfanger SG, Kaminski 68. Kasamon YL, Wahl RL. FDG PET and risk-adapted therapy in
MS, Wahl RL. Metabolic response of non-Hodgkin’s lymphoma Hodgkin’s and non-Hodgkin’s lymphoma. Curr Opin Oncol.
to 131I-anti-B1 radioimmunotherapy: evaluation with FDG PET. J 2008;20(2):206–19.
Nucl Med. 2000;41(6):999–1005. 69. Kobe C, Kuhnert G, Kahraman D, Haverkamp H, Eich HT, Franke
54. Kostakoglu L, Schoder H, Johnson JL, Hall NC, Schwartz LH, M, et al. Assessment of tumor size reduction improves outcome
Straus DJ, et al. Interim [18F]fluorodeoxyglucose positron emis- prediction of positron emission tomography/computed tomogra-
sion tomography imaging in stage I–II non-bulky Hodgkin lym- phy after chemotherapy in advanced-stage Hodgkin lymphoma. J
phoma: would using combined positron emission tomography and Clin Oncol. 2014;32(17):1776–81.
70. Straus DJ, Pitcher B, Kostakoglu L, et al. Initial results of US 86. Terezakis SA, Schoder H, Kowalski A, McCann P, Lim R,
intergroup trial of response-adapted chemotherapy or chemo- Turlakov A, et al. A prospective study of 18FDG-PET with CT
therapy/radiation therapy based on PET for non-bulky stage I and coregistration for radiation treatment planning of lymphomas and
II Hodgkin lymphoma (CALGB/Alliance 50604). Blood. 2015; other hematologic malignancies. Int J Radiat Oncol Biol Phys.
126:578. 2014;89(2):376–83.
71. Radford J, Illidge T, Counsell N, Hancock B, Pettengell R, Johnson 87. Bird D, Patel C, Scarsbrook AF, Cosgrove V, Thomas E, Gilson D,
P, et al. Results of a trial of PET-directed therapy for early-stage et al. Evaluation of clinical target volume expansion required for
Hodgkin’s lymphoma. N Engl J Med. 2015;372(17):1598–607. involved site neck radiotherapy for lymphoma to account for the
72. André MPE, Girinsky T, Federico M, Reman O, Fortpied C, Gotti absence of a pre-chemotherapy PET-CT in the radiotherapy treat-
M, et al. Early positron emission tomography response-adapted ment position. Radiother Oncol. 2017;124(1):161–7.
treatment in stage I and II Hodgkin lymphoma: final results of 88. Aznar MC, Maraldo MV, Schut DA, Lundemann M, Brodin NP,
the randomized EORTC/LYSA/FIL H10 trial. J Clin Oncol. Vogelius IR, et al. Minimizing late effects for patients with medi-
2017;35(16):1786–94. astinal Hodgkin lymphoma: deep inspiration breath-hold, IMRT,
73. Cashen AF, Dehdashti F, Luo J, Homb A, Siegel BA, Bartlett NL. or both? Int J Radiat Oncol Biol Phys. 2015;92(1):169–74.
18
F-FDG PET/CT for early response assessment in diffuse large 89. Kriz J, Spickermann M, Lehrich P, Schmidberger H, Reinartz
B-cell lymphoma: poor predictive value of international harmoni- G, Eich H, et al. Breath-hold technique in conventional APPA
zation project interpretation. J Nucl Med. 2011;52(3):386–92. or intensity-modulated radiotherapy for Hodgkin’s lymphoma:
74. Pregno P, Chiappella A, Bello M, Botto B, Ferrero S, Franceschetti comparison of ILROG IS-RT and the GHSG IF-RT. Strahlenther
S, et al. Interim 18-FDG-PET/CT failed to predict the outcome Onkol. 2015;191(9):717–25.
in diffuse large B-cell lymphoma patients treated at the diagnosis 90. Petersen PM, Aznar MC, Berthelsen AK, Loft A, Schut DA,
with rituximab-CHOP. Blood. 2012;119(9):2066–73. Maraldo M, et al. Prospective phase II trial of image-guided radio-
75. Moskowitz CH, Schoder H, Teruya-Feldstein J, Sima C, Iasonos A, therapy in Hodgkin lymphoma: benefit of deep inspiration breath-
Portlock CS, et al. Risk-adapted dose-dense immunochemother- hold. Acta Oncol. 2015;54(1):60–6.
apy determined by interim FDG-PET in advanced-stage diffuse 91. Jerusalem G, Beguin Y, Fassotte MF, Belhocine T, Hustinx R,
large B-Cell lymphoma. J Clin Oncol. 2010;28(11):1896–903. Rigo P, et al. Early detection of relapse by whole-body positron
76. Swinnen LJ, Li H, Quon A, Gascoyne R, Hong F, Ranheim EA, emission tomography in the follow-up of patients with Hodgkin’s
et al. Response-adapted therapy for aggressive non-Hodgkin’s disease. Ann Oncol. 2003;14(1):123–30.
lymphomas based on early [18F]FDG-PET scanning: ECOG- 92. Rhodes MM, Delbeke D, Whitlock JA, Martin W, Kuttesch JF,
ACRIN Cancer Research Group study (E3404). Br J Haematol. Frangoul HA, et al. Utility of FDG-PET/CT in follow-up of chil-
2015;170(1):56–65. dren treated for Hodgkin and non-Hodgkin lymphoma. J Pediatr
77. Zelenetz AD, Abramson JS, Advani RH, Andreadis CB, Byrd JC, Hematol Oncol. 2006;28(5):300–6.
Czuczman MS, et al. NCCN clinical practice guidelines in oncol- 93. Levine JM, Weiner M, Kelly KM. Routine use of PET scans
ogy: non-Hodgkin’s lymphomas. J Natl Compr Cancer Netw. after completion of therapy in pediatric Hodgkin disease
2010;8(3):288–334. results in a high false positive rate. J Pediatr Hematol Oncol.
78. Lin C, Itti E, Haioun C, Petegnief Y, Luciani A, Dupuis J, et al. 2006;28(11):711–4.
Early 18F-FDG PET for prediction of prognosis in patients with 94. Zinzani PL, Stefoni V, Tani M, Fanti S, Musuraca G, Castellucci
diffuse large B-cell lymphoma: SUV-based assessment versus P, et al. Role of [18F]fluorodeoxyglucose positron emission
visual analysis. J Nucl Med. 2007;48(10):1626–32. tomography scan in the follow-up of lymphoma. J Clin Oncol.
79. Lesokhin AM, Ansell SM, Armand P, Scott EC, Halwani A, 2009;27(11):1781–7.
Gutierrez M, et al. Nivolumab in patients with relapsed or refrac- 95. Petrausch U, Samaras P, Haile SR, Veit-Haibach P, Soyka JD,
tory hematologic malignancy: preliminary results of a phase Ib Knuth A, et al. Risk-adapted FDG-PET/CT-based follow-up in
study. J Clin Oncol. 2016;34(23):2698–704. patients with diffuse large B-cell lymphoma after first-line ther-
80. Hoos A, Wolchok JD, Humphrey RW, Hodi FS. CCR 20th apy. Ann Oncol. 2010;21(8):1694–8.
Anniversary Commentary: Immune-related response criteria - 96. Nuutinen J, Leskinen S, Lindholm P, Soderstrom KO, Nagren K,
capturing clinical activity in immuno-oncology. Clin Cancer Res. Huhtala S, et al. Use of carbon-11 methionine positron emission
2015;21(22):4989–91. tomography to assess malignancy grade and predict survival in
81. Kirienko M, Sollini M, Chiti A. Hodgkin lymphoma and imag- patients with lymphomas. Eur J Nucl Med. 1998;25(7):729–35.
ing in the era of anti‐PD‐1/PD‐L1 therapy. Clin Transl Imaging. 97. Sutinen E, Jyrkkio S, Varpula M, Lindholm P, Gronroos T,
2018;6(6):417–27. Lehikoinen P, et al. Nodal staging of lymphoma with whole-body
82. Cheson BD, Ansell S, Schwartz L, Gordon LI, Advani R, Jacene PET: comparison of [11C]Methionine and FDG. J Nucl Med.
HA, et al. Refinement of the Lugano classification lymphoma 2000;41(12):1980–8.
response criteria in the era of immunomodulatory therapy. Blood. 98. Buchmann I, Neumaier B, Schreckenberger M, Reske S. [18F]3′-
2016;128(21):2489–96. deoxy-3′-fluorothymidine-PET in NHL patients: whole-body
83. Roschewski M, Dunleavy K, Pittaluga S, Moorhead M, Pepin F, biodistribution and imaging of lymphoma manifestations—a pilot
Kong K, et al. Circulating tumour DNA and CT monitoring in study. Cancer Biother Radiopharm. 2004;19(4):436–42.
patients with untreated diffuse large B-cell lymphoma: a correla- 99. Buck AK, Bommer M, Stilgenbauer S, Juweid M, Glatting G,
tive biomarker study. Lancet Oncol. 2015;16(5):541–9. Schirrmeister H, et al. Molecular imaging of proliferation in
84. Illidge T, Specht L, Yahalom J, Aleman B, Berthelsen AK, malignant lymphoma. Cancer Res. 2006;66(22):11055–61.
Constine L, et al. Modern radiation therapy for nodal non- 100. Kasper B, Egerer G, Gronkowski M, Haufe S, Lehnert T, Eisenhut
Hodgkin lymphoma-target definition and dose guidelines from the M, et al. Functional diagnosis of residual lymphomas after radio-
International Lymphoma Radiation Oncology Group. Int J Radiat chemotherapy with positron emission tomography comparing
Oncol Biol Phys. 2014;89(1):49–58. FDG- and FLT-PET. Leuk Lymphoma. 2007;48(4):746–53.
85. Girinsky T, Auperin A, Ribrag V, Elleuch M, Ferme C, Bonniaud 101. Schoder H, Zelenetz AD, Hamlin P, Gavane S, Horwitz S, Matasar
G, et al. Role of FDG-PET in the implementation of involved- M, et al. Prospective study of 3′-deoxy-3′-18F-fluorothymidine
node radiation therapy for Hodgkin lymphoma patients. Int J PET for early interim response assessment in advanced-stage
Radiat Oncol Biol Phys. 2014;89(5):1047–52. B-cell lymphoma. J Nucl Med. 2016;57(5):728–34.
102. Sfakianakis GN, DeLand FH. Radioimmunodiagnosis and radiotherapy in patients with relapsed follicular non-Hodgkin lym-
immunotherapy, 1982. J Nucl Med. 1982;23:840–50. phoma: final results of a phase II study. Oncology. 2018; https://
103. Rosen ST, Winter JN, Epstein AL. Application of monoclonal doi.org/10.1159/000486788.
antibodies to tumor diagnosis and therapy. Ann Clin Lab Sci. 120. Conti PS, White C, Pieslor P, Molina A, Aussie J, Foster P. The
1983;13:173–84. role of imaging with 111In-ibritumomab tiuxetan in the ibri-
104. Roselli M, Guadagni F, Buonomo O, Belardi A, Ferroni P, Diodati tumomab tiuxetan (Zevalin) regimen: results from a Zevalin
A, et al. Tumor markers as targets for selective diagnostic and Imaging Registry. J Nucl Med. 2005;46:1812–8.
therapeutic procedures. Anticancer Res. 1996;16:2187–92. 121. Vose JM, Bierman PJ, Loberiza FR Jr, Bociek RG, Matso D,
105. Mirick GR, Bradt BM, Denardo SJ, Denardo GL. A review of Armitage JO. Phase I trial of 90Y-ibritumomab tiuxetan in patients
human anti-globulin antibody (HAGA, HAMA, HACA, HAHA) with relapsed B-cell non-Hodgkin’s lymphoma following high-
responses to monoclonal antibodies. Not four letter words. Q J dose chemotherapy and autologous stem cell transplantation.
Nucl Med Mol Imaging. 2004;48:251–7. Leuk Lymphoma. 2007;48:683–90.
106. Anderson KC, Bates MP, Slaughenhoupt BL, Pinkus GS, 122. Kylstra JW, Witzig TE, Huang M, Emmanouilides CE, Hagenbeek
Schlossman SF, Nadler LM. Expression of human B cell- A, Tidmarsh GF. Discriminatory power of the 111-indium scan
associated antigens on leukemias and lymphomas: a model of (111-In) in the prediction of altered biodistribution of radio-
human B cell differentiation. Blood. 1984;63:1424–33. immunoconjugate in the 90-yttrium ibritumomab tiuxetan thera-
107. Tedder TF, Boyd AW, Freedman AS, Nadler LM, Schlossman peutic regimen: meta-analysis of five clinical trials and 9 years of
SF. The B cell surface molecule B1 is functionally linked with B post-approval safety data. J Clin Oncol. 2011;29(15 suppl):8048.
cell activation and differentiation. J Immunol. 1985;135:973–9. 123. Kaminski MS, Zasadny KR, Francis IR, Milik AW, Ross CW,
108. Cardarelli PM, Quinn M, Buckman D, Fang Y, Colcher D, King Moon SD, et al. Radioimmunotherapy of B-cell lymphoma with
DJ, et al. Binding to CD20 by anti-B1 antibody or F(ab′)2 is suf- [131I]anti-B1 (anti-CD20) antibody. N Engl J Med. 1993;329:
ficient for induction of apoptosis in B-cell lines. Cancer Immunol 459–65.
Immunother. 2002;51:15–24. 124. Vose JM, Wahl RL, Saleh M, Rohatiner AZ, Knox SJ, Radford
109. Stashenko P, Nadler LM, Hardy R, Schlossman SF. JA, et al. Multicenter phase II study of iodine-131 tositumomab
Characterization of a human B lymphocyte-specific antigen. J for chemotherapy-relapsed/refractory low-grade and transformed
Immunol. 1980;125:1678–85. low-grade B-cell non-Hodgkin’s lymphomas. J Clin Oncol.
110. Wagner HN Jr, Wiseman GA, Marcus CS, Nabi HA, Nagle CE, 2000;18:1316–23.
Fink-Bennett DM, et al. Administration guidelines for radioim- 125. Moreau P, San Miguel J, Ludwig H, Schouten H, Mohty M,
munotherapy of non-Hodgkin’s lymphoma with 90Y-labeled anti- Dimopoulos M, et al. Multiple myeloma: ESMO clinical practice
CD20 monoclonal antibody. J Nucl Med. 2002;43:267–72. guidelines for diagnosis, treatment and follow-up. Ann Oncol.
111. Witzig TE, Gordon LI, Cabanillas F, Czuczman MS, 2017;28(suppl_4):iv52–61.
Emmanouilides C, Joyce R, et al. Randomized controlled 126. Kyle RA, Rajkumar V. Criteria for diagnosis, staging, risk stratifi-
trial of yttrium-90-labeled ibritumomab tiuxetan radioimmu- cation and response assessment of multiple myeloma. Leukemia.
notherapy versus rituximab immunotherapy for patients with 2009;23:3–9.
relapsed or refractory low-grade, follicular, or transformed 127. Rajkumar SV, Dimopoulos MA, Palumbo A, Blade J, Merlini G,
B-cell non- Hodgkin’s lymphoma. J Clin Oncol. 2002;20: Mateos MV, et al. International Myeloma Working Group updated
2453–63. criteria for the diagnosis of multiple myeloma. Lancet Oncol.
112. Hochster H, Weller E, Gascoyne RD, Habermann TM, Gordon 2014;15:e538–48.
LI, Ryan T, et al. Maintenance rituximab after cyclophosphamide, 128. Waxman AJ, Mick R, Garfall AL, Cohen A, Vogl DT, Stadtmauer
vincristine, and prednisone prolongs progression-free survival in EA, et al. Classifying ultra-high risk smoldering myeloma.
advanced indolent lymphoma: results of the randomized phase III Leukemia. 2014;29:751–3.
ECOG1496 study. J Clin Oncol. 2009;27:1607–14. 129. Raab MS, Podar K, Breitkreutz I, Richardson PG, Anderson
113. Goldsmith SJ. Radioimmunotherapy of lymphoma: Bexxar and KC. Multiple myeloma. Lancet. 2009;374:324–39.
Zevalin. Semin Nucl Med. 2010;40:122–35. 130. Lindsley H, Teller D, Noonan B, Peterson M, Mannik
114. Morschhauser F, Radford J, Van Hoof A, Botto B, Rohatiner AZ, M. Hyperviscosity syndrome in multiple myeloma. A reversible,
Salles G, et al. 90Yttrium-ibritumomab tiuxetan consolidation of concentration-dependent aggregation of the myeloma protein. Am
first remission in advanced-stage follicular non-Hodgkin lym- J Med. 1973;54:682–8.
phoma: updated results after a median follow-up of 7.3 years from 131. Latov N, Sherman WH, Nemni R, Galassi G, Shyong JS, Penn
the international, randomized, phase III first-line indolent trial. J AS, et al. Plasma-cell dyscrasia and peripheral neuropathy with a
Clin Oncol. 2013;31:1977–83. monoclonal antibody to peripheral-nerve myelin. N Engl J Med.
115. Rizzieri D. Zevalin® (ibritumomab tiuxetan): after more than a 1980;303:618–21.
decade of treatment experience, what have we learned? Crit Rev 132. Dimopoulos M, Kyle R, Fermand JP, Rajkumar SV, San
Oncol Hematol. 2016;105:5–17. Miguel J, Chanan-Khan A, et al. Consensus recommendations
116. Shimoni A, Zwas ST. Radioimmunotherapy and autologous stem- for standard investigative workup: report of the International
cell transplantation in the treatment of B-cell non-Hodgkin lym- Myeloma Workshop Consensus Panel 3. Blood. 2011;117:
phoma. Semin Nucl Med. 2016;46(2):119–25. 4701–5.
117. Mondello P, Cuzzocrea S, Navarra M, Mian M. 90 Y-ibritumomab 133. Rajkumar SV. Updated diagnostic criteria and staging system for
tiuxetan: a nearly forgotten opportunity. Oncotarget. multiple. Am Soc Clin Oncol Educ Book. 2016;35:418–23.
2016;7(7):7597–609. 134. Lütje S, Rooy JWJ, Croockewit S, Koedam E, Oyen WJG,
118. Jacene HL, Tirumani S, Wahl RL. Radionuclide therapy of lym- Raymakers RA. Role of radiography, MRI and FDG-PET/CT in
phomas. In: Strauss HW, Mariani G, Volterrani D, Larson SM, diagnosing, staging and therapeutical evaluation of patients with
editors. Nuclear oncology - from pathophysiology to clinical multiple myeloma. Ann Hematol. 2009;88:1161–8.
applications. New York: Springer; 2017. p. 1141–56. 135. Durie BG, Salmon SE. A clinical staging system for multiple
119. Puvvada SD, Guillén-Rodríguez JM, Yan J, Inclán L, Heard K, myeloma. Correlation of measured myeloma cell mass with
Rivera XI, et al. Yttrium-90-Ibritumomab tiuxetan (Zevalin®) presenting clinical features, response to treatment, and survival.
radioimmunotherapy after cytoreduction with ESHAP chemo- Cancer. 1975;36:842–54.
136. Durie BGM. The role of anatomic and functional staging in diffusion-weighted imaging: comparison with x-ray skeletal sur-
myeloma: description of Durie/Salmon plus staging system. Eur J vey by region and relationship with laboratory estimates of dis-
Cancer. 2006;42:1539–43. ease burden. Clin Radiol. 2015;70:614–21.
137. Terpos E, Kleber M, Engelhardt M, Zweegman S, Gay F, Kastritis 154. Sachpekidis C, Mosebach J, Freitag MT, Wilhelm T, Mai EK,
E, et al. European myeloma network guidelines for the manage- Haberkorn U, et al. Application of 18F-FDG PET and diffusion
ment of multiple myeloma-related complications. Haematologica. weighted imaging (DWI) in multiple myeloma: comparison of
2015;100:1254–66. functional imaging modalities. Am J Nucl Med Mol Imaging.
138. Dimopoulos M, Terpos E, Comenzo RL, Tosi P, Beksac M, Sezer 2015;5:479–92.
O, et al. International myeloma working group consensus state- 155. Pawlyn C, Fowkes L, Otero S, Jones J, Boyd K, Davies F, et al.
ment and guidelines regarding the current role of imaging tech- Whole-body diffusion-weighted MRI: a new gold standard for
niques in the diagnosis and monitoring of multiple Myeloma. assessing disease burden in patients with multiple myeloma?
Leukemia. 2009;23:1545–56. Leukemia. 2016;30:1446–8.
139. Terpos E, Moulopoulos LA, Dimopoulos MA. Advances in imag- 156. Hubner KF, Andrews GA, Hayes RL, Poggenburg JK Jr, Solomon
ing and the management of myeloma bone disease. J Clin Oncol. A. The use of rare-earth radionuclides and other bone-seekers in
2011;29:1907–15. the evaluation of bone lesions in patients with multiple myeloma
140. Regelink JC, Minnema MC, Terpos E, Kamphuis MH, or solitary plasmacytoma. Radiology. 1977;125:171–6.
Raijmakers PG, Pieters-van den Bos IC, et al. Comparison of 157. Kanoh T, Ohno T, Uchino H, Yamamoto I, Torizuka K. Avid
modern and conventional imaging techniques in establishing uptake of gallium-67 in multiple myeloma. An additional indica-
multiple m yeloma-related bone disease: a systematic review. Br J tor of the aggressive phase. Clin Nucl Med. 1987;12:482–4.
Haematol. 2013;162:50–61. 158. Roach PJ, Arthur CK. Comparison of thallium-201 and gallium-67
141. Agren B, Bjorkstrand B, Rudberg U, Aspelin PBL, Ågren B, scintigraphy in soft tissue and bone marrow multiple myeloma: a
Björkstrand B, et al. Radiography and bone scintigraphy in bone case report. Australas Radiol. 1997;41:67–9.
marrow transplant multiple myeloma patients. Acta Radiol. 159. Lin WY, Wang SJ. Ga-67 scan findings in bone marrow involve-
1997;38:144–50. ment with plasmablastic myeloma and corresponding Tc-99m
142. Chassang M, Grimaud A, Cucchi JM, Novellas S, Amoretti N, MIBI images. Clin Nucl Med. 2001;26(11):963.
Chevallier P, et al. Can low-dose computed tomographic scan of 160. Bekerman C, Hoffer PB, Bitran JD. The role of gallium-67 in
the spine replace conventional radiography? An evaluation based the clinical evaluation of cancer. Semin Nucl Med. 1984;14:
on imaging myelomas, bone metastases, and fractures from osteo- 296–323.
porosis. Clin Imaging. 2007;31:225–7. 161. Ohnishi T, Noguchi S, Murakami N, Tajiri J, Morita M, Tamaru
143. Pianko MJ, Terpos E, Roodman GD, Divgi CR, Zweegman S, M, et al. Pentavalent technetium-99m-DMSA uptake in a patient
Hillengass J, et al. Whole-body low-dose computed tomography having multiple myeloma without amyloidosis. J Nucl Med.
and advanced imaging techniques for multiple myeloma bone dis- 1991;32:1785–7.
ease. Clin Cancer Res. 2014;20:5888–97. 162. Ishibashi M, Nonoshita M, Uchida M, Kojima K, Tomita N,
144. Terpos E, Dimopoulos MA, Moulopoulos LA. The role of imag- Matsumoto S, et al. Bone marrow uptake of thallium-201
ing in the treatment of patients with multiple myeloma in 2016. before and after therapy in multiple myeloma. J Nucl Med.
Am Soc Clin Oncol Educ Book. 2016;35:e407–17. 1998;39:473–5.
145. Winterbottom AP, Shaw AS. Imaging patients with myeloma. Clin 163. Watanabe N, Shimizu M, Kageyama M, Tanimura K, Kinuya
Radiol. 2009;64:1–11. S, Shuke N, et al. Multiple myeloma evaluated with 201Tl
146. Princewill K, Kyere S, Awan O, Mulligan M. Multiple myeloma scintigraphy compared with bone. J Nucl Med. 1999;40:
lesion detection with whole body CT versus radiographic skeletal 1138–42.
survey. Cancer Investig. 2013;31:206–11.147. 164. Pace L, Catalano L, Pinto A, De Renzo A, Di Gennaro F, Califano
147. Wolf MB, Murray F, Kilk K, Hillengass J, Delorme S, Heiss C, C, et al. Different patterns of technetium-99m sestamibi uptake in
et al. Sensitivity of whole-body CT and MRI versus projection multiple myeloma. Eur J Nucl Med. 1998;25:714–20.
radiography in the detection of osteolyses in patients with mono- 165. Balleari E, Villa G, Garrè S, Ghirlanda P, Agnese G, Carletto
clonal plasma cell disease. Eur J Radiol. 2014;83:1222–30. M, et al. Technetium-99m-Sestamibi scintigraphy in multiple
148. Cretti F, Perugini G. Patient dose evaluation for the whole-body myeloma and related gammopathies: a new tool for the identifi-
low-dose multidetector CT (WBLDMDCT) skeleton study in cation and follow-up of myeloma bone disease. Haematologica.
multiple myeloma (MM). Radiol Med. 2016;121:93–105. 2001;86:78–84.
149. Dimopoulos MA, Hillengass J, Usmani S, Zamagni E, Lentzsch S, 166. Mirzaei S, Filipits M, Keck A, Bergmayer W, Knoll P, Koehn
Davies FE, et al. Role of magnetic resonance imaging in the man- H, et al. Comparison of Technetium-99m-MIBI imaging
agement of patients with multiple myeloma: a consensus state- with MRI for detection of spine involvement in patients with
ment. J Clin Oncol. 2015;33:657–64. multiple myeloma. BMC Nucl Med. 2003;3:2. https://doi.
150. D’Sa S, Abildgaard N, Tighe J, Shaw P, Hall-Craggs M. Guidelines org/10.1186/1471-2385-3-2.
for the use of imaging in the management of myeloma. Br J 167. Fonti R, Salvatore B, Quarantelli M, Sirignano C, Segreto S,
Haematol. 2007;137:49–63. Petruzziello F, et al. 18F-FDG PET/CT, 99mTc-MIBI, and MRI
151. Merz M, Moehler TM, Ritsch J, Bäuerle T, Zechmann CM, in evaluation of patients with multiple myeloma. J Nucl Med.
Wagner B, et al. Prognostic significance of increased bone marrow 2008;49:195–200.
microcirculation in newly diagnosed multiple myeloma: results of 168. Agool A, Slart RHJA, Dierckx RAJO, Kluin PM, Visser L, Jager
a prospective DCE-MRI study. Eur Radiol. 2016;26:1404–11. PL, et al. Somatostatin receptor scintigraphy might be useful
152. Terpos E, Koutoulidis V, Fontara S, Zagouri F, Christoulas D, for detecting skeleton abnormalities in patients with multiple
Koureas A, et al. Diffusion-weighted imaging improves accuracy myeloma and plasmacytoma. Eur J Nucl Med Mol Imaging.
in the diagnosis of MRI patterns of marrow involvement in newly 2010;37:124–30.
diagnosed myeloma: results of a prospective study in 99 patients. 169. Hazenberg BPC, van Rijswijk MH, Piers DA, Lub-de Hooge
Blood. 2015;126:4178. MN, Vellenga E, Haagsma EB, et al. Diagnostic performance of
153. Giles SL, Desouza NM, Collins DJ, Morgan VA, West S, Davies 123
I-labeled serum amyloid P component scintigraphy in patients
FE, et al. Assessing myeloma bone disease with whole-body with amyloidosis. Am J Med. 2006;119:355.e15–24.
170. Bredella M, Steinbach L, Caputo G, Segall G, Hawkins R. Value 187. Dankerl A, Liebisch P, Glatting G, Friesen C, Blumstein
of FDG PET in the assessment of patients with multiple myeloma. NM, Kocot D, et al. Multiple myeloma: molecular imaging
AJR Am J Roentgenol. 2005;184:1199–204. with 11C-methionine PET/CT—initial experience. Radiology.
171. Durie BG, Waxman AD, D’Agnolo A, Williams C. Whole- 2007;242:498–508.
body 18F-FDG PET identifies high-risk myeloma. J Nucl Med. 188. Lapa C, Knop S, Schreder M, Rudelius M, Knott M, Jörg G,
2002;43:1457–63. et al. 11 C-Methionine-PET in multiple myeloma: correlation with
172. Breyer RJ, Mulligan ME, Smith SE, Line BR, Badros clinical parameters and bone marrow involvement. Theranostics.
AZ. Comparison of imaging with FDG PET/CT with other imag- 2016;6:254–61.
ing modalities in myeloma. Skelet Radiol. 2006;35:632–40. 189. Lee SM, Kim TS, Lee JW, Kwon HW, Kim YI, Kang SH,
173. Hillner BE, Siegel BA, Liu D, Shields AF, Gareen IF, Hanna L, et al. Incidental finding of an 11C-acetate PET-positive multiple
et al. Impact of positron emission tomography/computed tomog- myeloma. Ann Nucl Med. 2010;24:41–4.
raphy and positron emission tomography (PET) alone on expected 190. Huang SY, Chen BB, Lu HY, Lin HH, Wei SY, Hsu SC, et al.
management of patients with cancer: initial results from the Correlation among DCE-MRI measurements of bone marrow
National Oncologic PET Registry. J Clin Oncol. 2008;26:2155–61. angiogenesis, microvessel density, and extramedullary disease in
174. Hillner BE, Siegel BA, Shields AF, Liu D, Gareen IF, Hunt E, patients with multiple myeloma. Am J Hematol. 2012;87:837–9.
Coleman RE. Relationship between cancer type and impact of PET 191. Merz M, Ritsch J, Kunz C, Wagner B, Sauer S, Hose D, et al.
and PET/CT on intended management: findings of the national Dynamic contrast-enhanced magnetic resonance imaging for
oncologic PET registry. J Nucl Med. 2008;49(12):1928–35. assessment of antiangiogenic treatment effects in multiple
175. Chae M, Lee T, Park G, Yu J, Pai M, Kang H, et al. Comparing myeloma. Clin Cancer Res. 2015;21:106–12.
18
F-FDG-PET/CT with other imaging modalities for detecting 192. Dutoit JC, Claus E, Offner F, Noens L, Delanghe J, Verstraete
involving bone of multiple myeloma. J Nucl Med. 2007;48:351P. KL. Combined evaluation of conventional MRI, dynamic contrast-
176. Schirrmeister H, Buck AK, Bergmann L, Reske SN, Bommer enhanced MRI and diffusion weighted imaging for response
M. Positron emission tomography (PET) for staging of solitary evaluation of patients with multiple myeloma. Eur J Radiol.
plasmacytoma. Cancer Biother Radiopharm. 2003;18:841–5. 2016;85:373–82.
177. Nanni C, Rubello D, Zamagni E, Castellucci P, Ambrosini V, 193. Moulopoulos LA, Dimopoulos MA, Alexanian R, Leeds NE,
Montini G, et al. 18F-FDG PET/CT in myeloma with presumed Libshitz HI. Multiple myeloma: MR patterns of response to treat-
solitary plasmocytoma of bone. In Vivo. 2008;22:513–7. ment. Radiology. 1994;193:441–6.
178. Salaun PY, Gastinne T, Frampas E, Bodet-Milin C, Moreau P, 194. Hillengass J, Ayyaz S, Kilk K, Weber M-A, Hielscher T, Shah R,
Bodéré-Kraeber F. FDG-positron-emission tomography for stag- et al. Changes in magnetic resonance imaging before and after
ing and therapeutic assessment in patients with plasmacytoma. autologous stem cell transplantation correlate with response and
Haematologica. 2008;93:1269–71. survival in multiple myeloma. Haematologica. 2012;97:1757–60.
179. Nanni C, Zamagni E, Farsad M, Castellucci P, Tosi P, Cangini D, 195. Bannas P, Hentschel HB, Bley TA, Treszl A, Eulenburg C, Derlin
et al. Role of 18F-FDG PET/CT in the assessment of bone involve- T, et al. Diagnostic performance of whole-body MRI for the detec-
ment in newly diagnosed multiple myeloma: preliminary results. tion of persistent or relapsing disease in multiple myeloma after
Eur J Nucl Med Mol Imaging. 2006;33:525–31. stem cell transplantation. Eur Radiol. 2012;22:2007–12.
180. Hur J, Yoon C-S, Ryu YH, Yun MJ, Suh J-S. Comparative study of 196. Spinnato P, Bazzocchi A, Brioli A, Nanni C, Zamagni E, Albisinni
fluorodeoxyglucose positron emission tomography and magnetic U, et al. Contrast enhanced MRI and 18F-FDG PET-CT in the
resonance imaging for the detection of spinal bone marrow infil- assessment of multiple myeloma: a comparison of results in dif-
tration in untreated patients with multiple myeloma. Acta Radiol. ferent phases of the disease. Eur J Radiol. 2012;81:4013–8.
2008;49:427–35. 197. Fallahi D, Beiki D, Mousavi SA, Gholamrezanezhad A, Eftekhari
181. Bartel TB, Haessler J, Brown TLY, Shaughnessy JD, Van Rhee M, Fard-Esfahani A, Alimoghaddam K, et al. 99mTc-MIBI whole
F, Anaissie E, et al. F18-fluorodeoxyglucose positron emission body scintigraphy and P-glycoprotein for the prediction of mul-
tomography in the context of other imaging techniques and prog- tiple drug resistance in multiple myeloma patients. Hell J Nucl
nostic factors in multiple myeloma. Blood. 2009;114:2068–76. Med. 2009;12:255–9.
182. Gorospe L, Raman S, Echeveste J, Avril N, Herrero Y, Herna 198. Piwnica-Worms D, Chiu ML, Budding M, Kronauge JF, Kramer
Ndez S. Whole-body PET/CT: spectrum of physiological variants, RA, Croop JM. Functional imaging of multidrug-resistant
artifacts and interpretative pitfalls in cancer patients. Nucl Med P-glycoprotein with an organotechnetium complex. Cancer Res.
Commun. 2005;26:671–87. 1993;53:977–84.
183. Mesguich C, Fardanesh R, Tanenbaum L, Chari A, Jagannath S, 199. Patriarca F, Melli C, Damiani D, Michieli M, Michelutti A, Cavo
Kostakoglu L. State of the art imaging of multiple myeloma: com- M, et al. Plasma cell P170 expression and response to treatment in
parative review of FDG PET/CT imaging in various clinical set- multiple myeloma. Haematologica. 1996;81:232–7.
tings. Eur J Radiol. 2014;83:2203–23. 200. Mongkonsritragoon W, Kimlinger T, Ahmann G, Greipp P. Is
184. Nanni C, Zamagni E, Versari A, Chauvie S, Bianchi A, Rensi M, multidrug resistance (P-glycoprotein) an intrinsic characteristic of
et al. Image interpretation criteria for FDG PET/CT in multiple plasma cells in patients with monoclonal gammopathy of undeter-
myeloma: a new proposal from an Italian expert panel. IMPeTUs mined significance plasmacytoma, multiple myeloma and amyloi-
(Italian Myeloma criteria for PET USe). Eur J Nucl Med Mol dosis? Leuk Lymphoma. 1998;29:577–84.
Imaging. 2016;43:414–21. 201. Moreau P, Attal M, Caillot D, Macro M, Karlin L, Garderet L,
185. Nanni C, Zamagni E, Cavo M, Rubello D, Tacchetti P, Pettinato C, et al. Prospective evaluation of magnetic resonance imaging
et al. 11C-choline vs. 18F-FDG PET/CT in assessing bone involve- and [18F]fluorodeoxyglucose positron emission tomography-
ment in patients with multiple myeloma. World J Surg Oncol. computed tomography at diagnosis and before maintenance ther-
2007;5:68. https://doi.org/10.1186/1477-7819-5-68. apy in symptomatic patients with multiple myeloma included in
186. Cassou-Mounat T, Balogova S, Nataf V, Calzada M, Huchet V, the IFM/DFCI 2009 trial: results of the IMAJEM Study. J Clin
Kerrou K, et al. 18F-fluorocholine versus 18F-fluorodeoxyglucose Oncol. 2017;35(25):2911–8.
for PET/CT imaging in patients with suspected relapsing or pro- 202. Sachpekidis C, Mai EK, Goldschmidt H, Hillengass J, Hose D,
gressive multiple myeloma: a pilot study. Eur J Nucl Med Mol Pan L, et al. 18F-FDG dynamic PET/CT in patients with multiple
Imaging. 2016;43(11):1995–2004. myeloma. Clin Nucl Med. 2015;40:e300–7.
203. Tirumani SH, Sakellis C, Jacene H, Shinagare AB, Munshi NC, 212. Abruzzese E, Iuliano F, Trawinska MM, Di Maio M. 153Sm: its use
Ramaiya NH, et al. Role of FDG-PET/CT in extramedullary mul- in multiple myeloma and report of a clinical experience. Expert
tiple myeloma: correlation of FDG-PET/CT findings with clinical Opin Investig Drugs. 2008;17:1379–87.
outcome. Clin Nucl Med. 2016;41:e7–13. 213. Dispenzieri A, Wiseman GA, Lacy MQ, Hayman SR, Kumar SK,
204. Zamagni E, Nanni C, Patriarca F, Englaro E, Castellucci P, Geatti Buadi F, et al. A phase II study of 153Sm-EDTMP and high-dose
O, et al. A prospective comparison of 18F-fluorodeoxyglucose melphalan as a peripheral blood stem cell conditioning regimen
positron emission tomography-computed tomography, magnetic in patients with multiple myeloma. Am J Hematol. 2010;85:
resonance imaging and whole-body planar radiographs in the 409–13.
assessment of bone disease in newly diagnosed multiple myeloma. 214. Berenson JR, Yellin O, Patel R, Duvivier H, Nassir Y, Mapes R,
Haematologica. 2007;92:50–5. et al. A phase I study of samarium lexidronam/bortezomib combi-
205. Zamagni E, Nanni C, Gay F, Pezzi A, Patriarca F, Bellò M, et al. nation therapy for the treatment of relapsed or refractory multiple
18
F-FDG PET/CT focal, but not osteolytic, lesions predict the myeloma. Clin Cancer Res. 2009;15:1069–75.
progression of smoldering myeloma to active disease. Leukemia. 215. Rousseau C, Ferrer L, Supiot S, Bardiès M, Davodeau F, Faivre-
2016;30:417–22. Chauvet A, et al. Dosimetry results suggest feasibility of radio-
206. Siontis B, Kumar S, Dispenzieri A, Drake MT, Lacy MQ, Buadi immunotherapy using anti-CD138 (B-B4) antibody in multiple
F, et al. Positron emission tomography-computed tomography myeloma patients. Tumour Biol. 2012;33:679–88.
in the diagnostic evaluation of smoldering multiple myeloma: 216. Sauer S, Erba PA, Petrini M, Menrad A, Giovannoni L, Grana C,
identification of patients needing therapy. Blood Cancer J.
et al. Expression of the oncofetal ED-B-containing fibronectin iso-
2015;5:e364. https://doi.org/10.1038/bcj.2015.87. form in hematologic tumors enables ED-B-targeted 131I-L19SIP
207. Lonial S, Kaufman JL. Non-secretory myeloma: a clinician’s radioimmunotherapy in Hodgkin lymphoma patients. Blood.
guide. Oncology (Williston Park). 2013;27:924–30. 2009;113:2265–74.
208. Sachpekidis C, Hillengass J, Goldschmidt H, Mosebach J, Pan 217. Erba PA, Sollini M, Orciuolo E, Traino C, Petrini M, Paganelli
L, Pet F. Comparison of 18F-FDG PET/CT and PET/MRI in G, et al. Radioimmunotherapy with radretumab in patients with
patients with multiple myeloma. Am J Nucl Med Mol Imaging. relapsed hematologic malignancies. J Nucl Med. 2012;53:922–7.
2015;5:469–78. 218. Mesguich C, Zanotti-Fregonara P, Hindié E. New perspectives
209. Sollini M, Galimberti S, Boni R, Erba PA. Radionuclide therapy offered by nuclear medicine for the imaging and therapy of mul-
of leukemias and multiple myeloma. In: Strauss HW, Mariani G, tiple myeloma. Theranostics. 2016;6(2):287–90.
Volterrani D, Larson SM, editors. Nuclear oncology – from patho- 219. Herrmann K, Schottelius M, Lapa C, Osl T, Poschenrieder A,
physiology to clinical applications. New York: Springer; 2017. Hänscheid H, et al. First-in-human experience of CXCR4-directed
p. 1157–96. endoradiotherapy with 177Lu- and 90Y-labeled pentixather in
210. Orchard K, Cooper M, Lewington V, Tristan M, Zivanovic M, advanced-stage multiple myeloma with extensive intra- and extra-
Thom J, et al. Targeted radiotherapy in haematopoietic stem cell medullary disease. J Nucl Med. 2016;57(2):248–51.
transplantation: results of a phase I trial using an yttrium-90- 220. Bluemel C, Hahner S, Heinze B, Fassnacht M, Kroiss M, Bley
labelled anti-CD66 murine monoclonal antibody demonstrating TA, et al. Investigating the chemokine receptor 4 as potential ther-
consistent BM targeting. Bone Marrow Transpl. 2006;37:S45. anostic target in adrenocortical cancer patients. Clin Nucl Med.
211. Buchmann I, Meyer RG, Mier W, Haberkorn U. Myeloablative 2017;42(1):e29–34.
radioimmunotherapy in conditioning prior to haematological stem 221. Buck AK, Stolzenburg A, Hänscheid H, Schirbel A, Lückerath K,
cell transplantation: closing the gap between benefit and toxicity? Schottelius M, et al. Chemokine receptor - directed imaging and
Eur J Nucl Med Mol Imaging. 2009;36:484–98. therapy. Methods. 2017;130:63–71.
Therapy for Thyroid Disorders 27
Federica Guidoccio, Gayane Aghakhanyan,
and Mariano Grosso
Contents
27.1 Anatomy and Physiology of the Thyroid Gland 708
27.2 Benign Thyroid Disease 708
27.2.1 Nontoxic Multinodular Goiter 708
27.2.2 Toxic Adenoma and Multinodular Toxic Goiter 709
27.2.3 Autoimmune Thyroid Disease 710
27.2.4 Other Forms of Thyroiditis 711
27.3 Thyroid Ultrasonography 711
27.4 Radiopharmaceuticals for Thyroid Imaging 712
27.4.1 Radionuclide Diagnostic Applications in Benign Thyroid Disease 713
27.4.2 Radionuclide Therapy of Benign Thyroid Disease 715
27.5 Malignant Thyroid Disease 720
27.5.1 Background 720
27.5.2 Staging 720
27.5.3 Clinical Presentation 720
27.5.4 Ultrasonography 722
27.5.5 Fine Needle Aspiration Cytology 722
27.5.6 Radionuclide Imaging of DTC 723
27.5.7 Radionuclide Imaging of ATC 725
27.5.8 Radionuclide Imaging of MTC 726
27.6 Treatment of Thyroid Cancers 728
27.6.1 Overall Scenario 728
27.6.2 Primary Treatment of DTC 728
27.6.3 Side Effects of Radioiodide Therapy 732
27.6.4 Evaluation of Success of Thyroid Remnant Ablation 733
27.6.5 Levothyroxine Therapy 733
27.6.6 Follow-Up of DTC After Primary Treatment 735
27.6.7 Management of Locoregional Recurrences from DTC 737
27.6.8 Management of Distant Metastases 737
References 743
F. Guidoccio · G. Aghakhanyan
M. Grosso (*)
Regional Center of Nuclear Medicine, University Hospital of Pisa,
Pisa, Italy
e-mail: m.grosso@med.unipi.it

Learning Objectives disease or Hashimoto’s thyroiditis. The pyramidal lobe

• Understand the basic notion on anatomy and physiology (which at surgery is found in about 80% of patients) becomes
of the thyroid gland. thus an item of some importance diagnostically and in thy-
• Acquire the basic clinical notions on benign and malig- roid surgery.
nant thyroid diseases. The adult thyroid is composed of follicles or acini. The
• Learn the modalities of performing and interpreting follicle is filled with a colloid material that contains thyro-
radionuclide imaging in patients with benign or malig- globulin, the precursor macromolecule, and storage protein
nant thyroid diseases. for the thyroid hormones: thyroxine (T4) and triiodothyro-
• Understand the general principles for the treatment of nine (T3). In addition to the acinar cells, there are individual
benign and malignant thyroid diseases. cells or small groups of cells that are seen not to extend to the
• Learn the distinctive features of differentiated and non- follicular lumen and which may appear as clusters between
differentiated thyroid tumors. follicles. These parafollicular cells, called “C-cells,” are a dis-
• Understand the principles of image interpretation of tinct category probably derived from the neural crest via the
whole-body scintigraphy with 131I-iodide following pri- ultimobranchial body. The C-cells secrete calcitonin (or “thy-
mary treatment in patients with differentiated thyroid rocalcitonin”) in response to an increase in serum calcium.
cancer. This hormone is important in the regulation of bone resorp-
• Understand the basic modalities of implementing follow- tion and to a lesser extent influences the concentration of
up after radioiodine treatment for differentiated thyroid serum calcium. C-cells also express somatostatin receptors,
cancer. calcitonin gene-related peptide, gastrin-releasing peptide,
• Learn the main characteristics of iodine-refractory thy- katacalcin, and helodermin that could have either a stimula-
roid tumors and understand the rationale of new tory or inhibitory activity on thyroid hormone secretion. The
treatments. C-cells give rise to the “medullary” thyroid cancers. In adult
human they represent 1% of the cell population.
The main function of the thyroid gland is to provide ade-
quate amounts of thyroid hormone for the proper regulation
27.1 Anatomy and Physiology of a large number of bodily functions, e.g., energy expendi-
of the Thyroid Gland ture and metabolic rate. Thyroid hormone is an iodinated
hormone and thus requires the ability of the thyroid gland to
The thyroid gland is a butterfly-shaped organ located anteri- concentrate iodine from the circulation and organify it for
orly to the trachea at the level of the second and third tracheal incorporation into the thyroid hormone molecules. Thus, one
rings. It consists of two lobes connected by the isthmus in the of the most critical properties of a thyroid cell is its ability to
midline. Anteriorly, its surface is convex; posteriorly, it is trap iodine from the circulation, most often against a concen-
concave. The gland’s upper extremities are known as the tration gradient. Iodine, an essential component of the hor-
upper poles. Similarly, the lower extremities of the lateral mones produced by the thyroid gland, is widely but unevenly
lobes are known as the lower poles (or base of the lobe). The distributed in the earth’s environment. The native iodine con-
weight of the thyroid of the normal non-goitrous adult is tent of most foods and beverages is low, and most commonly
6–20 g depending on the body size and iodine supply. From consumed foods provide 3–80 μg per serving—against a
upper pole to base, the thyroid lobes usually measure 4 cm. minimum daily requirement of about 150 μg. Major dietary
Their width is 15–20 mm, and their thickness is 20–39 mm. sources of iodine are bread, milk, and to a lesser extent sea-
The isthmus is 12–15 mm high, lies across the trachea anteri- food. The processes of iodide uptake and thyroid hormone
orly just below the level of the cricoid cartilage, and connects synthesis are largely regulated by the concentration of the
the two lobes. The shape and attachments of the organ and its pituitary glycoprotein hormone, thyroid-stimulating hor-
relationship to the trachea are important in examination and mone (TSH). In the absence of pituitary or of thyrotroph
diagnosis and from the point of view of pressure symptoms. function, hypothyroidism ensues [1].
Occasionally, a pyramidal lobe is located in the midline,
superior to the isthmus. It represents a remnant of the thyro-
glossal duct, as the primitive thyroid gland descends from 27.2 Benign Thyroid Disease
the base of the tongue to its final location in the neck during
embryonic development. The importance of the pyramidal 27.2.1 Nontoxic Multinodular Goiter
lobe lies in its relation to developmental anomalies and also
in its propensity to undergo hypertrophy when the rest of the Thyroid nodules are common entities, frequently discovered
thyroid has been removed. Any pathological process that is in clinical practice either during physical examination, but
diffuse may involve the pyramidal lobe, for example, Graves’ also incidentally, during various imaging procedures. The
27 Hybrid Imaging and Radionuclide Therapy for Thyroid Disorders 709
normal thyroid gland is a fairly homogenous structure, but candidates for surgery, particularly if performed under ultra-
nodules often form within its parenchyma. A thyroid nodule sound guidance [3].
is a discrete lesion within the thyroid gland that is radiologi-
cally/ultrasonographically distinct from the surrounding thy-
roid parenchyma. These nodules may be only the growth and 27.2.2 Toxic Adenoma and Multinodular Toxic
fusion of localized colloid-filled follicles, or more or less Goiter
discrete adenomas, or cysts. Some palpable lesions may not
correspond to distinct radiologic abnormalities. Either subclinical or overt hyperthyroidism develops in a
Nonpalpable nodules detected on ultrasonography or large proportion of MNGs, frequently after exposure to
other anatomic imaging studies are termed incidentally dis- iodine excess. Multinodular toxic goiter (MTG) is the sec-
covered nodules or “incidentalomas.” Nonpalpable nodules ond most common cause of hyperthyroidism after Graves’
have the same risk of malignancy as do sonographically con- disease. A single nodule is called a toxic adenoma (Plummer
firmed palpable nodules of the same size. Generally, only disease) [4]. Toxic adenoma is due to a somatic activating
nodules >1 cm should be evaluated, since they have a greater mutation that results in an alteration of part of the TSH
potential to be clinically significant cancers. Occasionally, receptor, for which the receptor itself is always activated, in
there may be nodules <1 cm that require further evaluation the same manner as if it were always stimulated by
because of either clinical symptoms or certain ultrasound TSH. Since the mutation only occurs in one or more cellular
patterns suggesting a malignant nature or because of associ- clones (those that originate the adenoma), but not in the
ated lymphadenopathy. In very rare cases, some nodules majority of the remaining thyroid cells, the parenchyma not
<1 cm lack these sonographic and clinical warning signs, yet affected by the mutation retains a normal physiology of
they may nonetheless cause future morbidity and mortality. response to the receptor status for the TSH. Therefore, under
Given the cost/benefit considerations, attempts to diagnose conditions of suppressed TSH, the thyroid parenchyma sur-
and treat all such small thyroid cancers in an effort to prevent rounding the adenoma is in conditions of functional rest; this
exceedingly rare unfavorable outcomes are deemed to cause pathophysiological condition translates into a scintigraphic
more harm than good. In general, the guiding clinical strat- pattern of high uptake in the adenoma and low uptake in the
egy acknowledges that most thyroid nodules are low risk, extra-nodular parenchyma. This functional inhibition pro-
and many thyroid cancers pose minimal risk to human health tects the extra-nodular parenchyma from the cytotoxic action
and can be effectively treated [2]. of radioiodide administered for radiometabolic therapy,
The terms adenomatous goiter, nontoxic nodular goiter, which is instead concentrated upon the adenoma. The overall
and colloid nodular goiter are used interchangeably as result of radiometabolic therapy with radioiodide is therefore
descriptive terms when a multinodular goiter (MNG) is death of the adenomatous cells and hence reduction/disap-
found. The incidence of goiter, diffuse and nodular, is very pearance of the adenoma, while the extra-nodular paren-
much dependent on the status of iodine intake of the popula- chyma resumes its normal function. In fact, once the
tion. In areas of iodine deficiency, goiter prevalence may be inhibitory action on TSH secretion exerted by the increased
very high, and especially in goiters of longstanding, multi- levels of thyroid hormones secreted by the hyperfunctioning
nodularity develops frequently. MNG, most common of all adenoma (now disappeared) ceased, the normal regulation
the disorders of the thyroid gland, is the result of the genetic feedback between hypothalamus, hypophysis, and non-
heterogeneity of follicular cells and apparent acquisition of mutated thyroid parenchyma can take place.
new cellular qualities that become inheritable. MNG is most The prevalence of MTG increases with age and in the pres-
often detected simply as a mass in the neck, but sometimes ence of iodine deficiency; it may therefore be more common
an enlarging gland produces pressure symptoms. than Graves’ disease in older populations in regions of iodine
From 4% to 17% of MNGs fulfill the criteria of malignant deficiency. In turn, the prevalence of MNG evolving to MTG
change, whereas the majority of these lesions are not malig- is on average 5% in areas with sufficient iodine intake and
nant. Thus, the thyroid nodules are clinically important pri- 20% in areas of iodine deficiency. Iodine deficiency stimu-
marily due to their malignant potential. For this reason the lates thyroid cell proliferation, eventually resulting in devel-
initial evaluation should always include a history and physi- opment of somatic activating mutations of the TSH receptor.
cal examination focusing on features suggestive of malig- Genetic background and female gender, in addition to iodine
nancy. Serum TSH and thyroid ultrasonography are pivotal deficiency, are risk factors for its development.
in the evaluation of thyroid nodules, as they provide impor- A large prospective epidemiological study, recently con-
tant information regarding thyroid nodule function and the ducted in Denmark, in an area with mild to moderate iodine
presence of features suspicious for malignancy, respectively. deficiency, reported the following prevalence among the
Fine needle aspiration (FNA) biopsy is the most accurate and most frequent causes of hyperthyroidism: MTG 44%,
reliable tool for diagnosing thyroid malignancy and selecting Graves’ disease 37%, and toxic adenoma 6% [5].
27.2.3 Autoimmune Thyroid Disease cal presentation depends on the severity of thyrotoxicosis,
duration of disease, individual susceptibility to excess thy-
Autoimmune thyroid disease (AITD) is a prototypical organ- roid hormone, and patient’s age.
specific autoimmune disease. The etiology of AITD is multi- Thyrotoxicosis may cause unexplained weight loss
factorial, as interactions between genetic and environmental (despite an enhanced appetite, due to the increased metabolic
predisposing triggers lead to dysregulation of immune toler- rate), hyperactivity, nervousness, irritability, fatigue, insom-
ance. The prevalence of AITD is estimated to be 5%; how- nia, impaired concentration, fine tremor (best showed by
ever, the prevalence of antithyroid antibodies in the general having patients stretch out their fingers and feeling the fin-
population may be even higher [6]. Graves’ disease and gertips with the palm), sinus tachycardia (often associated
Hashimoto’s thyroiditis are the main clinical presentations of with palpitations, atrial fibrillation up to worsening of angina
AITD, characterized by lymphocytic infiltration of the thy- or heart failure in the elderly or those with pre-existing heart
roid parenchyma and clinically by thyrotoxicosis and hypo- disease), sweating, heat intolerance, pruritus, urticaria, alo-
thyroidism, respectively. Genetic susceptibility, female pecia, gastrointestinal disorder (often with diarrhea and
gender, and environmental triggers constitute the main fac- occasionally mild steatorrhea.), oligomenorrhea or amenor-
tors leading to the development of AITD. rhea in woman, impaired sexual function and gynecomastia
in men, and bone disorder (osteopenia in long-standing thy-
27.2.3.1 Graves’ Disease rotoxicosis with a small increase in fracture rate, mild hyper-
Graves’ disease (or toxic diffuse goiter) has been described calcemia, hypercalciuria).
for the first time in the nineteenth century by Dr. Robert Specific Graves’ disease manifestations include the so-
James Graves in Ireland. It is also known as Basedow-Flaiani called Graves’ ophthalmopathy, with a sensation of gritti-
disease, following the description of the typical clinical pre- ness, eye discomfort, and excess tearing. About one third of
sentation with ophthalmopathy and palpitations by Dr. Karl patients have proptosis, best detected by visualization of the
Adolph von Basedow (in Germany) and Dr. Giuseppe Flaiani sclera between the lower border of the iris and the lower eye-
(in Italy), in the same period. lid, with the eyes in the primary position. Symptoms may
Graves’ disease accounts for up to 60–80% of thyrotoxi- worsen up to diplopia, compression of the optic nerve, pap-
cosis (i.e., the state of thyroid hormone excess in the blood), illedema, peripheral field defects, and, if left untreated, per-
but its prevalence varies among populations depending manent loss of vision. The onset of Graves’ ophthalmopathy
mainly on iodine intake (high iodine intake is associated occurs within the year before or after the diagnosis of thyro-
with an increased prevalence of Graves’ disease). Graves’ toxicosis in 75% of patients but can sometimes precede or
disease occurs in up to 2% of women, while it is one tenth as follow thyrotoxicosis by several years, accounting for some
frequent in men. The disorder rarely begins before adoles- cases of euthyroid ophthalmopathy. Thyroid dermopathy
cence and typically occurs between 20 and 50 years of age, (5% of patients with Graves’ disease), also known with the
but it also occurs in the elderly. term “pretibial myxedema,” consists in edema and non-
The majority of investigators share the concept that inflamed, indurated plaque with a deep pink or purple color
Graves’ disease is a multifactorial disease caused by a and an “orange-skin” of the legs. Thyroid acropachy (i.e.,
complex interaction between genetic and environmental fac- soft tissue swelling of the hands and clubbing of the fingers)
tors. Several Graves’ disease susceptibility genes have been occurs in about 1% of patients with Graves’ disease and is
identified, which can be classified into immune-regulatory strongly associated with coincident skin and orbital
genes (HLA-DR, CTLA4, CD40, PTPN22) and thyroid-spe- involvement.
cific genes (thyroglobulin, TSHR). However, the penetrance The blood chemistry profile shows suppressed basal TSH
of the genetic determinants is presumably quite low, as also serum levels associated with increased fT4 and fT3 serum
suggested by the fact that Graves’ disease is not a hereditary levels. Associated biochemical abnormalities that may cause
disease, although it often recurs, along with autoimmune diagnostic confusion in thyrotoxicosis include elevation of
thyroiditis in general, within the same families. Among the bilirubin, liver enzymes, and ferritin. Microcytic anemia and
non-genetic factors involved in the etiology of Graves’ dis- thrombocytopenia may also occur.
ease, infections, iodine intake, smoking, and psychic stress Ultrasound often shows diffusely enlarged thyroid (usu-
have all been postulated to play a role, although no firm evi- ally up to two to three times its normal size). There may be a
dence is available [7, 8]. thrill or bruit due to the increased vascularity of the gland
The hyperthyroidism of Graves’ disease is related to the and the hyperdynamic circulation.
presence of thyroid-stimulating antibodies targeting the TSH
receptor (TRAb). Signs and symptoms include features that 27.2.3.2 Hashimoto Thyroiditis
are common to thyrotoxicosis from any cause, as well as Hashimoto thyroiditis is a chronic inflammation of the thy-
some features that are specific for Graves’ disease. The clini- roid gland initially described over a century ago but of still
incompletely defined etiopathogenesis. It is now considered tive tissue, which often extends into neighboring
as the most common autoimmune disease, the most common structures.
endocrine disorder, as well as the most common cause of • Miscellaneous varieties of thyroid inflammation or infil-
hypothyroidism. Based on the etiology, Hashimoto thyroid- tration including local manifestations of a generalized
itis can be classified into primary and secondary forms. disease processes, such as sarcoid and amyloid involve-
Primary Hashimoto thyroiditis is the most common form of ment of the thyroid. Radiation and direct trauma to the
thyroiditis and includes the cases that do not currently have thyroid gland may also cause thyroiditis [10].
identifiable causes. Primary Hashimoto thyroiditis encom-
passes a clinicopathological spectrum of six main entities:
classic form, fibrous variant, IgG4-related variant, juvenile
form, Hashitoxicosis, and painless (or silent) thyroiditis, the Key Learning Points
latter occurring either sporadically or in the postpartum. • Graves’ disease and Hashimoto’s thyroiditis are the
Clinically, the most common manifestation is an enlarge- main clinical presentations of autoimmune thyroid
ment of the thyroid gland (goiter), with or without hypothy- diseases, characterized by lymphocytic infiltration
roidism. Pathologically, the common denominator to all of the thyroid parenchyma and clinically by thyro-
variants is the marked lymphocytic infiltration of the thyroid. toxicosis and hypothyroidism, respectively.
Primary Hashimoto thyroiditis can occur in isolation or asso- • The terms adenomatous goiter, nontoxic nodular
ciated with other autoimmune diseases (e.g., type 1 diabetes goiter, and colloid nodular goiter are used inter-
mellitus, Sjögren syndrome, chronic atrophic gastritis) or changeably as descriptive terms when a multinodu-
other thyroid diseases [9]. lar goiter is found.
• Either subclinical or overt hyperthyroidism develops
in a large proportion of patients with multinodular
27.2.4 Other Forms of Thyroiditis goiter, frequently after exposure to iodine excess.
• Multinodular goiter is the second most common
The thyroid may be the site of an acute or chronic suppura- cause of hyperthyroidism after Graves’ disease.
tive or nonsuppurative inflammation. The diagnostic term • A single hyperfunctioning thyroid nodule causing
thyroiditis includes a group of inflammatory or inflammatory- hyperthyroidism is called a toxic adenoma
like conditions. The terminology that has been employed is (Plummer disease).
confusing, and no classification is ideal. The classification
presented here below takes into account etiology:
• Infectious thyroiditis (including either acute or chronic 27.3 Thyroid Ultrasonography

presentations), which may be acute suppurative thyroid-
itis (AST), nonsuppurative thyroiditis, or septic thyroid- The main reasons for the widespread use of thyroid ultraso-
itis. It includes all forms of infection, other than viral, and nography are its wide availability (several portable models
is caused by invasion of the thyroid by bacteria, mycobac- are now available at a relatively affordable price), low cost of
teria, fungi, protozoa, or flatworms. This disorder is very the procedure (if performed in the office or in the thyroid
rare. clinic), limited discomfort for the patient, and the nonioniz-
• De Quervain’s thyroiditis, commonly known as subacute ing nature of the method. Ultrasonography may detect non-
thyroiditis (SAT) but also termed subacute nonsuppura- palpable nodules and cysts, estimate nodule and goiter size
tive thyroiditis; granulomatous, pseudotuberculous, (volume), monitor the changes following therapy, and guide
pseudo-giant cell, or giant cell thyroiditis; and migratory the fine needle aspiration cytology (FNAC).
or creeping thyroiditis, and struma granulomatosa. This Microcalcifications, irregular or microlobulated margins,
condition, most likely of viral origin, lasts for a week to a hypoechogenicity, taller-than-wide shape, and increased
few months, with a tendency to recur. intranodular vascularity are considered as independent risk
• Autoimmune thyroiditis, commonly referred to as factors for malignancy. It is important to emphasize that
chronic, Hashimoto’s, or lymphocytic thyroiditis and also none of these features alone is sufficient to differentiate
known as lymphadenoid goiter and struma lymphomatosa benign from malignant tumors, but a combination of at least
(already described above). two of them better succeeds in pointing out a subset of
• Riedel’s thyroiditis, another disorder of unknown etiol- lesions at high risk for malignancy [5].
ogy. Synonyms include Riedel’s struma, ligneous thyroid- The possibility of measuring thyroid volume is another
itis, and invasive fibrous or chronic sclerosing thyroiditis. especially useful feature of ultrasonography, particularly
This condition is characterized by overgrowth of connec- after therapy with L-T4 or radioiodine ablation. Nevertheless,
there is an inter-observer coefficient of variation >10%.

Ultrasonography cannot evaluate a MNG that has partially Key Learning Points
migrated to the upper mediastinum [4]. • Thyroid ultrasonography allows targeting for
Besides identification of thyroid nodules, ultrasonogra- biopsy of nodules with suspicious appearance.
phy is able to characterize the echostructure of thyroid tissue • Besides identification of thyroid nodules, thyroid
in patients with AITD, where lymphocytic infiltration and ultrasonography can characterize the structure of
disruption of tissue architecture cause a reduction in thyroid thyroid parenchyma in patients with either subacute
echogenicity. Hashimoto thyroiditis and Graves’ disease thyroiditis (generally of viral etiology) or chronic
appear similar on gray-scale ultrasound, but power Doppler autoimmune thyroiditis.
will demonstrate increased blood flow in Graves’ disease • Ultrasonography of the neck provides important
(“thyroid inferno”), with Hashimoto thyroiditis showing the information on lymph node status in patients with
full spectrum from absent to normal to increased blood flow. thyroid cancer.
In granulomatous or de Quervain’s thyroiditis, the hypoecho-
genicity may be localized to one lobe or even a portion of a
lobe that is involved. Hypoechogenicity is a major character- 27.4 Radiopharmaceuticals for Thyroid
istic finding on ultrasound of patients having Hashimoto thy- Imaging
roiditis. However, other ultrasonographic findings in addition
to hypoechogenicity occur in Hashimoto’s thyroiditis. The most commonly used radiopharmaceuticals for first-
Heterogeneity in the ultrasound appearance of the thyroid choice diagnostic thyroid imaging include123I-iodide,
caused by destruction of the normal homogeneous “ground- 131
I-iodide, and 99mTc-pertechnetate. They can be adminis-
glass” architectural pattern of thyroid tissue results in the tered intravenously (the standard administration route for
formation of pseudonodules that may be numerous and 99m
Tc-pertechnetate) or orally (the standard administration
resemble a “bag of marbles.” These pseudonodules are not route for radioiodine). Each of these imaging agents has
well outlined or defined. Their appearance may also change some advantage and disadvantages. The advantages of 99mTc
over time, sometimes even over a few weeks. Pseudonodules include the wide availability in nuclear medicine depart-
are thought to be “diffuse lakes of lymphocytes.” Another ments, the low energy of the gamma photons (140 keV), and
frequent finding is the presence of tiny cystic lesions, which the relatively short half-life (T1/2 6 h). These characteristics
may be described as a “swiss cheese” appearance of the make it possible to use much higher activity than 131I at a
gland. These diffuse cystic lesions are usually 2–3 mm in lower cost, to acquire more accurate images in a shorter
size. Yet another characteristic feature of Hashimoto thyroid- acquisition time and with an absorbed dose much lower than
itis is echogenic strands or septa. These strands have been with the diagnostic 131I activity [12]. 99mTc-pertechnetate is
described as thin echogenic septa traversing the thyroid tis- an iodomimetic compound, which means that it is trans-
sue, sometimes giving the thyroid a lobulated appearance. ported into thyroid cells through the sodium-iodide (Na+/I−)
They are thought to be due to fibrosis within the gland. symporter system (NIS) in a similar way as the isotopes of
Fibrosis can also develop in pseudonodules, changing them iodine. The maximal accumulation of 99mTc-pertechnetate in
from hypoechoic to hyperechoic [11]. the thyroid gland is observed 15–20 min after intravenous
A recent advance in the diagnosis of thyroid nodules has administration, after which the curve of thyroidal 99mTc-
been brought by the use of thyroid elastography. This is a pertechnetate activity reaches a plateau (around 30 min
dynamic technique that assesses the hardness of the tissue as postinjection) and then declines.
an indicator of malignancy. This technique appears to be 123
I is a pure gamma emitter with relatively low energy
highly specific (96–100%) and sensitive (82–97%) in the diag- (159 keV) and a short half-life (T1/2 13 h); its physical prop-
nostic evaluation of thyroid nodules, irrespective of nodule erties are largely similar to those of 99mTc. The advantage of
size, or location within the thyroid gland. The diagnostic yield 123
I-iodide versus 99mTc-pertechnetate is its higher specificity
of elastography is impaired in nodules with a calcified shell, and persistence of accumulation in the thyroid due to the
cystic lesions, and MGN with coalescent nodules, because the organification pathway that allows later acquisitions and
margins need to be well demarcated for proper interpretation. careful assessment of iodine kinetics in the thyroid tissue.
It is not suitable for diagnosis of follicular carcinoma, and its Compared with 99mTc, 123I provides higher resolution images
use is restricted to high-end US devices. It constitutes a prom- with a clear demarcation of the thyroid tissue from surround-
ising tool in selecting nodules for FNAC [4]. ing tissues, therefore allowing better visualization of
The role of neck ultrasonography in the management of retrosternal thyroid tissue and of the thyroid gland when thy-
patients with thyroid cancer, either before or after primary roid uptake is low.
treatment (particularly for evaluating the lymph node status), 131
I emits beta and gamma radiations, with a half-life of
is detailed in one of the following sections of this chapter. 8.1 days. Its gamma rays, used for diagnostic purposes, have
higher energy (364 keV) than 99mTc and 123I. As a result, 27.4.1.2 Thyroid Scintigraphy
131
I-iodide scintigraphy should be limited to perform a radio- Thyroid scintigraphy allows the assessment of thyroid
iodine thyroid uptake test, monitoring of treatment in patients regional function and detection of areas of autonomously
with differentiated thyroid cancer, imaging of ectopic thy- functioning thyroid nodules. Thyroid scintigraphy visualizes
roid (i.e., retrosternal or lingual thyroid and struma ovarii), the distribution of active thyroid tissue and displays the dif-
and dosimetric pretreatment evaluations [12]. ferential accumulation of radionuclides in the investigated
cells, thus providing a functional map. This technique is a
fundamental tool in the clinical and surgical management of
27.4.1 Radionuclide Diagnostic Applications thyroid diseases. Thyroid scintigraphy is recommended in all
in Benign Thyroid Disease cases of thyroid nodule or MNG if the TSH level is below the
lower limit of the reference range or if ectopic thyroid tissue
27.4.1.1 Radioiodine Uptake Test or a retrosternal goiter is suspected.
The radioactive iodine uptake test (RAIU) is a useful diag- Practical procedure guidelines on patients’ preparation
nostic tool for assessing disorders of thyroid function. It is and how to perform a thyroid scintigraphy with either 99mTc-
performed using a gamma camera or a gamma probe posi- pertechnetate, 123I-iodide, or 131I-iodide are detailed in Chap.
tioned at a fixed distance from the patient’s neck. 46 of this book.
Measurements are performed at various intervals after oral Planar images are obtained at standard times post-
administration of a tracer amount of 131I-iodide (1–3.7 MBq administration (15–20 min for 99mTc-pertechnetate, 3–4 h
or 30–100 μCi). Measurements in the first 24 h are crucial for for 123I-iodide, 24 h for 131I-iodide); different examples of
the assessment of the maximal iodine accumulation in the scintigraphic patterns in different thyroid conditions are
thyroid. Therefore, for practical reasons, measurements are shown in Figs. 27.1 and 27.2. Anatomical landmarks are
often performed at 4 and 24 h after administration of the diag- set by placing a radioactive marker over the sternal notch
nostic amount of 131I-iodide. The test is mainly performed in and recording its position in the scan image. The acquisi-
patients planned for radioiodine treatment due to hyperthy- tion of SPECT/CT may provide additional diagnostic
roidism. For dosimetric purposes, it is important to determine information showing the exact topography and size of the
the residence time of 131I in the thyroid tissue; therefore, mea- nodules, particularly useful with retrosternal goiter and
surements can be extended over several consecutive days. complex thyroid nodularity cases (see Figs. 27.3 and
Radioiodine uptake is a parameter with a relatively high 27.4).
variability. On average, in euthyroid patients the uptake is Failure of thyroid scintigraphy to visualize the specific
25–50% 24 h after the administration. It may be higher in iodine-trapping function of the thyroid gland in an individual
hyperthyroidism and lower in hypothyroidism. Apart from whose thyroid gland is normally in place (as demonstrated
the thyroid functional status, many other (mostly iatrogenic) by clinical palpation and/or ultrasound imaging) may be due
factors may influence the level of thyroid iodine uptake [12]. to the following conditions:
a b
Fig. 27.1 Planar images obtained in two different patients with hyper- responding to a thyroid nodule, with completely suppressed uptake in the
thyroidism 15 min after the i.v. injection of 150 MBq 99mTc-Pertechnetate, extra-nodular parenchyma, as typically seen in patients with Plummer
using a gamma camera equipped with a LEHR collimator. (a) thyrotoxicosis (from Volterrani D, Erba PA, Mariani G. Fondamenti di
Homogeneously and diffusely increased tracer uptake in an enlarged thy- Medicina Nucleare—Tecniche e Applicazioni. Milan: Springer, 2010,
roid gland, as typically seen in patients with Graves’s disease. (b) Single with permission of Springer Science + Business Media)
focus of increased tracer uptake at the base of the right thyroid lobe cor-
a b
Fig. 27.2 (a) Planar gamma camera image obtained with a pinhole col- expected to deliver an absorbed dose of 300 Gy to the nodule. (b) 99mTc-
limator 24 h after administration of 131I-iodide (1.85 MBq) in a patient Pertechnetate thyroid scintigraphy (gamma camera equipped with a
with Plummer thyrotoxicosis scheduled for radioiodine therapy: single LEHR collimator) obtained 6 months after radiometabolic therapy,
focus of markedly increased tracer uptake at the apex of the right thyroid showing persistent mildly increased tracer uptake in the nodule at the
lobe corresponding to a thyroid nodule, with completely suppressed apex of the right thyroid lobe, but with some recovery of function of the
uptake in the extra-nodular parenchyma. Based on uptake values (43.7% extra-nodular parenchyma (from Volterrani D, Erba PA, Mariani
at 4 h, 73% at 24 h) and on ultrasonographic volume of the thyroid nod- G. Fondamenti di Medicina Nucleare—Tecniche e Applicazioni. Milan:
ule (9 mL or 9 g), 185 MBq of 131I-iodide was administered for therapy, Springer, 2010, with permission of Springer Science + Business Media)
• Exceedingly enlarged iodine pool in the body (caused by, • Single thyroid nodules with a suppressed TSH level.
e.g., medication with iodine-containing drugs, recent • MNGs, even without suppressed TSH (although this indi-
administration of an iodinated X-ray contrast medium, etc.) cation is not univocal and should also take into account
• Acute/subacute thyroiditis. the iodine supply status).
• Therapy with TSH-suppressive doses of thyroid hor- • To identify cold or indeterminate areas for FNAC and hot
mones in patients with euthyroid (multi)nodular goiter. areas that do not need cytologic evaluation.
• Factitious hyperthyroidism due to excessive consumption • Large MNGs, especially with substernal extension.
of thyroid hormones (either T3 or T4), often as undue • Diagnosis of ectopic thyroid tissue.
self-medication. • Subclinical hyperthyroidism, to identify occult hyper-
functioning tissue.
On the basis of the pattern of radionuclide uptake, nodules • Follicular lesions, to identify a functioning cellular ade-
may be classified as hyperfunctioning (“hot”), hypofunction- noma that could be benign.
ing (“cold”), or indeterminate. Hot nodules almost never rep- • To distinguish low-uptake from high-uptake
resent clinically significant malignant lesions, whereas cold thyrotoxicosis.
or indeterminate nodules have a reported malignancy risk of • To determine eligibility for radioiodine therapy.
3–15%. As most thyroid lesions are cold or indeterminate, but
only a minority of them are malignant, the predictive value of
hypofunctioning or indeterminate nodules for the presence of
malignancy is low. Diagnostic specificity is further decreased Key Learning Points
in case of small lesions (<1 cm) that are below the resolution • Thyroid nodules constitute the most common disor-
threshold of scintigraphy. der of the endocrine system.
Thyroid scintigraphy allows the early identification of thy- • The radioactive iodine uptake test is a useful diag-
roid autonomy and discrimination of cold and indeterminate nostic tool for assessing disorders of thyroid func-
nodules in MNGs for FNAC. It should be performed in all tion. Radioiodine uptake may be higher in
patients with nodular goiter scheduled for treatment with hyperthyroidism and lower in hypothyroidism.
radioiodine, since it allows the identification of the anatomical • Thyroid scintigraphy allows the early diagnosis of
distribution of active thyroid tissue, which is important in the thyroid autonomy and prioritization for fine needle
selection of the therapeutic 131I activity to be administered [8]. aspiration cytology of cold and indeterminate nod-
In summary, thyroid scintigraphy is indicated in the fol- ules in multinodular goiters.
lowing settings:
a b
Fig. 27.3 99mTc-Pertechnetate thyroid scintigraphy in a patient with terization. (b) Selected fused SPECT/CT images in the axial (left),
bulky retrosternal goiter. (a) Planar anterior view, showing a multinodu- coronal (middle), and sagittal sections (right). (c) Fused SPECT/TC
lar goiter with areas with in homogeneously increased tracer uptake; images in axial planes encompassing the entire gland from the neck to
following planar imaging, SPECT/CT was acquired for better charac- the upper chest (left to right and top to bottom)
27.4.2 Radionuclide Therapy of Benign destruction of the follicles; these combined effects on thy-
Thyroid Disease roid parenchyma ultimately result in fibrosis, after variable
periods from treatment. The reduction of functioning thy-
The use of radiometabolic therapy with radioiodine for thyroid tissue consequent to this form of therapy is therefore
roid disorders dates back to 1940, when patients with comparable to a surgical removal. Radiometabolic therapy
hyperthyroidism were treated for the first time with with 131I-iodide is indicated in the following benign
131
I-iodide. Although this radionuclide emits both β− and γ diseases:
radiations, its therapeutic activity is largely due (94%) to
the particulate β− radiation which has a mean range in tis- • Graves’ disease (toxic diffuse goiter)
sue of 0.4 mm. • Toxic adenoma and multinodular toxic goiter
The early radiobiologic effect of this radiation causes • Nontoxic nodular goiter
cellular necrosis with subsequent inflammatory reaction and • Nontoxic diffuse nodular goiter
Fig. 27.4 Thyroid scintigraphy with

123
I-iodide in a patient with recurrent a
multinodular toxic goiter after
unsuccessful trans-axillary robotic
surgery. (a) Anterior/posterior
whole-body views (left panel) and
anterior/posterior spot views, showing
multiple areas of functioning thyroid
parenchyma in the neck, with a major
focus of markedly increased uptake
extending from the neck to below the
sternal notch; following planar
imaging SPECT/CT was acquired for
better characterization. (b) Fused
SPECT/CT 3D volume rendering
showing the topographic correlates of
the iodide-avid tissue relative to bone
structure (upper left), and fused
SPECT/TC images in coronal planes
encompassing the entire gland from
an anterior to a posterior plane (right
panel, left to right and top to bottom),
showing three small areas of
radioiodide uptake in the medial
anterior cervical region consistent
with residual thyroid tissue and
thyroglossal duct; another larger
focus of uptake (about 3 cm in size)
located more inferiorly corresponds
to the largest remnant of thyroid
tissue
b
27.4.2.1 Hyperthyroidism implement such low-iodine regimen, a preparation that is use-

The most common causes of hyperthyroidism are Graves’ ful also in patients with thyroid cancer before radioiodide scin-
disease, toxic adenoma, and toxic MNG. The goal of therapy tigraphy or before radioiodide therapy (see further below).
for hyperthyroidism is to achieve a non-hyperthyroid status— Pretreatment of selected patients with thionamides
either a euthyroid state or iatrogenic hypothyroidism to be (methimazole or propylthiouracil) to deplete thyroid hor-
compensated to the euthyroid state with oral levothyroxine. mone stores may be helpful, considering that administration
As described in detail in Chap. 46 of this book, there are of therapeutic 131I can cause radiation-induced thyroiditis
certain well-established protocols for preparation of the patients with a release of stored thyroid hormones into the circula-
for radiometabolic therapy with 131I-iodide. In particular, in tion. This effect occasionally results in transient worsening
order to optimize uptake of radioiodide in the thyroid tissue, the of hyperthyroidism and, rarely, precipitation of a thyroid
iodide pool in the body should be depleted. This is achieved by storm; this event is more likely to occur following adminis-
discontinuing use of iodide-containing medications and by fol- tration of greater radioiodine activities to patients with a
lowing a low-iodine diet for at least 2 weeks before radioiodide large, iodine-avid thyroid gland. Accordingly, elderly
administration; care should also be taken in avoiding excessive patients and patients with significant pre-existing heart dis-
iodine intake through other sources, e.g., drugs, iodinated con- ease, severe systemic illness, or debilitation may benefit
trast agents, etc. Table 27.1 provides indication on how to from pretreatment with thionamides, which should be dis-
Table 27.1 High-iodine content foods/products to be limited/avoided The current trend is to achieve hypothyroidism after a single
before scintigraphy or radioiodide therapy in patients with DTC
treatment with 131I-iodide, rather than a euthyroidism, in order to
Amount of time it avoid frequent office visits and laboratory testing to detect late-
should be avoided
onset (even several years later) hypothyroidism and to decrease
Food/product before therapy
Iodized salt, sea salt 2 weeks the risk of untreated, persistent, or recurrent hyperthyroidism.
Milk or other dairy products, including ice 2 weeks
cream, cheese, yogurt, etc. Graves’ Disease
Eggs 2 weeks The hyperthyroidism of Graves’ disease is treated using anti-
Seafood, including fish, shellfish, kelp, or 2 weeks thyroid drugs (thionamides, such as propylthiouracil and
seaweed
methimazole) or by reducing the amount of thyroid tissue
Foods that contain the additives 2 weeks
carrageenan, agar-agar, algin alginates
either with radioiodine therapy or with thyroidectomy.
Cured and corned foods (ham, lox, corned 2 weeks Although all such treatments are effective, opinions vary
beef, sauerkraut) regarding the first-choice treatment and indications. Antithyroid
Breads made with iodinated dough 2 weeks drugs are the predominant therapy in many centers in Europe
conditioners and Japan, whereas radioiodine is more often the first line of
Foods and medications containing red food 2 weeks treatment in North America. These differences reflect the fact
dyes
Chocolate 2 weeks
that no single approach is optimal and that patients may require
Molasses 2 weeks multiple treatments to achieve remission.
Soy products (soy sauce, soy milk) 2 weeks Besides the possible occurrence of severe side effects (such
Additional drugs interfering with adequate radioiodide uptake and as bone marrow toxicity and liver toxicity), patients treated
suggested length of withdrawal: with antithyroid drugs often tend to relapse, especially in case
Lugol’s solution 3 weeks of severe hyperthyroidism and large goiters. Surgical thyroid-
Potassium iodide solution 3 weeks ectomy and radioiodine therapy constitute therefore a valid
Topical iodine (e.g., surgical skin 3 weeks
alternative option for permanent treatment of Graves’ disease.
preparation)
Iodinated radiographic contrast agents:
Thyroidectomy is preferred when the goiter is very large;
Intravenous (water soluble) 4 weeks (with normal major complications include bleeding, laryngeal edema,
renal function) hypoparathyroidism, and damage of the recurrent laryngeal
Lipophilic (rarely used) >4 weeks nerves. Nevertheless, these complications are unusual in the
Amiodarone >3 months hands of highly experienced surgeons.
In addition, restaurant food should be 3 weeks Therapy with radioiodine, which has been used for almost
avoided, because there is no way to know
whether they use iodized salt
80 years, has largely replaced surgery because it is easy to
perform and has proved to be effective. It causes progressive
Reproduced with permission from: Strauss HW, Mariani G, Volterrani
D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology to destruction of thyroid cells due to the β radiations emitted
−
Clinical Applications, 2nd Edition. New York: Springer; 2017 by I and can be used either as initial treatment or for recur-
131
ring disease after therapy with antithyroid drugs. It is highly

effective, with a cure rate approaching 100% after one or
continued for 3–5 days before radioiodine therapy and can more cycles. Radiation safety precautions are necessary over
be resumed 2–3 days afterward. Treatment with beta- the first few days after treatment, but the exact guidelines
blockers can be helpful for symptomatic control and should vary depending on local protocols and regulations, as
not be discontinued before treatment with 131I-iodide. described in Chaps. 12 and 13 of this book. In general,
Before radioiodine therapy, the treating physician must patients need to avoid close, prolonged contact with children
explain to the patient the procedure, treatment, possible com- and pregnant women for several days.
plications and side effects, therapeutic alternatives, and Sometimes there may be mild pain due to radiation thy-
expected outcome. The fetal thyroid gland concentrates roiditis in the anterior cervical region, about 1–2 weeks after
iodine by weeks 10–13; therefore, pregnancy is an absolute treatment. As mentioned above, there is also a small risk of
contraindication to radiometabolic therapy with 131I-iodide. thyrotoxic crisis due to cytolysis of thyroid cells, which can
Breastfeeding women must be advised to stop breastfeeding, be minimized by antithyroid drugs. Although rarely, hyper-
and therapy must be delayed until lactation ceases, in order to thyroidism can persist for several months before the thera-
minimize the radiation dose both to the child and to the breast. peutic effect of radioiodine takes full effect (up to
After treatment, thyrotoxic patient must be closely fol- 6–8 months). In case of persistent hyperthyroidism, the
lowed, as 131I-induced hypothyroidism may occur within patient can be treated with a second dose of radioiodine, usu-
2–3 months after therapy. Levothyroxine replacement ther- ally not earlier than 8–12 months after the first dose.
apy should be started as soon as TSH elevation is detected Pregnancy and breastfeeding are absolute contraindica-
and should have as its goal a euthyroid, symptom-free state. tions to radioiodine treatment, but patients can conceive
safely 6 months after treatment. The presence of severe oph- radioiodine uptake are experimentally measured. Effectiveness
thalmopathy requires caution, and some authorities advocate of this approach based on the “volume algorithm” is compa-
the use of corticosteroids to prevent exacerbation of ophthal- rable to that of the fixed 400 Gy thyroid- absorbed dose
mopathy. The overall risk of radiation-induced malignancies method [18]. However, the cure of Graves’ disease with the
after radioiodine treatment in adults is not increased with fixed 400 Gy dose is achieved at the expense of a significantly
respect to the general population. greater administered activity than in the thyroid mass reduc-
Efforts to calculate an optimal dose of radioiodine that tion method. Therefore, the “volume algorithm” approach
achieves euthyroidism without a high incidence of relapse or allows the administration of a lower 131I-iodide activity, thus
progression to hypothyroidism have not been successful. Some implying a lower radiation burden to the thyroid itself
patients inevitably relapse after a single treatment because the (296 Gy) and, more notably, to the remainder of the body.
biologic effects of radiation vary between individuals. However, This feature of the “volume algorithm” approach (that
it is now well known that the development of hypothyroidism is does not imply any relevant additional burden to the health-
progressive, with an annual incidence of 2–3% even many years care staff in terms of time dedicated to each patient) results
after therapy [13]. Moreover, low-dose radioiodine treatment in associated advantages to the patient (e.g., by reducing the
has a higher treatment failure rate; in these cases further cycles radiation burden to the whole body, as well as the prescrip-
of antithyroid drug therapy and additional 131I-iodide doses are tion of time off work following treatment with radioiodine),
needed [14]. The currently accepted goal of radioiodine therapy to the environment (by reducing the risk of contamination for
to control hyperthyroidism by rendering the patient hypothyroid family members and/or caretakers in general), and to the
requires greater doses, resulting in higher cure rates [15]. hospital staff (by reducing professional exposure linked to
At the present there is still no univocal consensus regard- daily handling of radioactive materials).
ing the most appropriate 131I-iodide dosing regimen. A prac-
tical strategy is to administer a fixed activity of generally Toxic Adenoma and Multinodular Toxic Goiter
185, 370, or 555 MBq (5, 10, or 15 mCi) based on clinical Together with surgery, therapy with 131I-iodide is one of the
features, such as severity of thyrotoxicosis, size of the goiter treatments of choice for patients with toxic adenoma or toxic
(leading to increase the dose), and the radioiodine uptake MNG, considering that antithyroid drugs are generally
value (leading to reduce the dose). Another strategy is to unable to achieve cure of hyperthyroidism in these patients.
administer an activity individually calculated based on a In patients with toxic adenoma, treatment with 131I should be
fixed value of target-absorbed dose (ranging between 100 performed in the phase in which the hyperfunctioning ade-
and 400 Gy according to different guidelines) based on noma inhibits radioiodide uptake by the parenchyma sur-
MIRD formulations and dependent on the thyroid gland size rounding the nodule, as visualized at scintigraphy. This
and the 24-h radioiodine uptake [16] or on the effective occurs in the conditions of overt hyperthyroidism (TSH sup-
radioiodine half-life in the thyroid gland [17]. pressed and elevated free thyroid hormones) or subclinical
Although it is reasonable to assume that effectiveness of hyperthyroidism (TSH suppressed with free thyroid hor-
such therapy depends on administration of an appropriate mones in the normal range). In the latter condition, the ben-
radiation dose to the thyroid, still there is no agreement on the efit of treatment is still the subject of discussion, although it
optimal radiation dose to deliver to the thyroid gland. Over should be emphasized that subclinical hyperthyroidism is an
several years of clinical experience with patients undergoing important risk factor for cardiovascular events (mostly car-
radiometabolic therapy for treatment of Graves’ disease, it diac arrhythmia) in individuals over 60 years of age.
was observed that there is a threshold value of thyroid gland Two different approaches are used to choose the 131I-iodide
volume after treatment that guarantees success of therapy activity to be administered: a fixed “maximal” activity (typi-
(i.e., achievement of euthyroidism or hypothyroidism). This cally 600 MBq—or 16 mCi, the maximum activity that in
final post-therapy thyroid volume is not a fixed value across most countries can be administered on an outpatient basis) or
patients but is instead related in each individual patient to an activity derived from a dosimetric calculation, that is, the
baseline thyroid mass and to the value of radioiodine maxi- lowest activity able to deliver an acceptable radiation dose to
mum uptake (usually 24 h after therapy administration). the target tissue. Although these methods are both effective,
Based on this observation, a new algorithm (the so-called the “As Low As Reasonably Achievable” (ALARA) principle
volume algorithm) has been developed aimed at personaliz- should be always pursued, avoiding unjustified radioactivity
ing the amount of 131I-iodide activity to be administered in exposure to the patient, to people living/working near the
order to cure patients from hyperthyroidism; according to this patient, and to the environment.
approach, the activity of 131I-iodide to be administered is cal- Dosimetric calculations imply measurement of 131I-iodide
culated using a relatively simple equation in which the only uptake after the administration of a tracer activity and evalua-
unknown variable is “activity” as a function of the desirable tion of functioning target tissue volume. Also for patients with
(optimal) target post-therapy thyroid volume that guarantees toxic MGN, the multiparameter dosimetric approach clinically
success of treatment, while the baseline thyroid mass and validated for Graves’ disease (see above) has been validated.
Nontoxic Multinodular Goiter patient with a thyroid mass of 70 g, the initial estimate of
The ideal treatment for nontoxic multinodular goiter (NTNG) the activity of 131I-iodide to be administered for therapy
is controversial, in part due to the variation in the natural his- would be 70 × 3.7 = 259 MBq (or 7 mCi); assuming the
tory of these goiters. Some patients present an enlargement 24-h uptake value of this patient is 30%, the activity to be
in goiter size with time, along with the development of nod- administered activity would be 259/0.3 = 863 MBq.
ules, symptoms of compression, and cosmetic-related prob- Instead, assuming a 50% 24-h uptake value, the activity to
lems, whereas in some patients the enlarged goiter may be administered would be 259/0.5 = 518 MBq.
remain stable or even reduce in size spontaneously with Besides the general side effects possibly caused by thera-
time, in about 20% of the women and 5% of men. peutic administration of 131I-iodide, in the short-term this
Current alternatives for NTNG treatment include: therapy can cause radiation-induced inflammation with
edema (possibly causing transitory worsening of symptoms
• Clinical observation for asymptomatic patients. due to mechanical obstruction) and sudden release of thy-
• TSH-suppressive therapy with exogenous levothyroxine roid hormones in circulating blood (possibly causing transi-
• Radioiodine therapy, alone or preceded by administration tory hyperthyroidism). Both these effects can be prevented
of recombinant human TSH (rhTSH) by adequate medication, such as, e.g., steroid drugs and
• Surgery beta-blockers. Although rarely, autoimmune thyroid disease
may develop 2–6 months after therapy, due to exposure of
Choice among these options is based on patient-specific the immune system to thyroid-derived antigens released
considerations taking into account for each patient the risks, during the radiation-induced cytolysis of thyroid cells.
benefits, and availability of the various techniques, experience The efficacy of this therapy is demonstrated by an average
of the treating physician, and patient’s personal preference. reduction in thyroid volume equal to about 30–40% within
Especially in elderly patients (often with comorbidities that 1 year posttreatment, with further reductions up to 50–60%
contraindicate surgery), therapy with 131I-iodide with the pur- of the initial volume in the following year. Within 2–3 years
pose of reducing size of the goiter is a valid alternative. Before after radioiodine therapy, most of the patients develop hypo-
actually administering radioiodide, patients must be screened thyroidism, which can be easily treated with replacement
for feasibility of this therapeutic approach, as follows: hormone therapy (l-Thyroxine).
• Malignancy must be excluded, by performing FNAC in

all nodule >1 cm in size.
• The total thyroid volume/mass must be evaluated using Key Learning Points
the most adequate imaging modality (usually ultrasound • Radiometabolic therapy with 131I-iodide is indicated
but also including CT and/or MRI for bulky goiters for the most common causes of hyperthyroidism—
extending beyond the sternal notch—thus evaluating also Graves’ disease, toxic adenoma, and multinodular
tracheal diameter). toxic goiter.
• The “active” thyroid volume must be (semi)quantitatively • Although 131I-iodide emits both β− and γ radiations, its
evaluated by thyroid scintigraphy, considering that areas therapeutic activity is largely due (94%) to the β− par-
with low uptake (such as, e.g., cystic lesions) do not con- ticles causing cellular necrosis with subsequent inflam-
centrate radioiodide and are therefore not expected to matory reaction and destruction of thyroid follicles.
respond to radiometabolic therapy. • At the present there is still no univocal consensus
• A radioiodine uptake test must be performed, considering regarding the most appropriate 131I-iodide dosing
that the lower is the 24-h uptake value, the greater must be regimen for the treatment of Graves’ disease.
the 131I-iodide activity to be administered. • A practical strategy is to administer a fixed activity
• In case the 24 h uptake of a trace amount of radioiodine is based on empiric estimates, generally 185, 370, or
<10–15%, low-dose stimulation with recombinant human 555 MBq (5, 10, or 15 mCi).
TSH should be considered (usually 0.3 mg versus the • Another strategy is to administer an activity individu-
0.9 mg administered in patients with DTC). ally calculated based on a fixed value of target-absorbed
dose (ranging between 100 and 400 Gy according to
The activity of 131I-iodide to be administered is usually different guidelines) based on MIRD formulation.
calculated according to a fixed activity per gram approach • A new algorithm (the so-called volume algorithm),
corrected for the 24-h uptake value. An activity of 3.7 MBq based on baseline thyroid mass, radioiodine uptake,
(or 100 μCi) per gram of functioning thyroid tissue is gen- and desired post-therapy thyroid mass, has been
erally considered sufficient for volume-reduction purposes developed to optimize the definition of the radioac-
of NTNG; this activity is divided by the fractional 24-h tivity to be administered to patients.
radioiodine uptake value, as exemplified hereinafter. In a
27.5 Malignant Thyroid Disease Table 27.2 Classification of thyroid cancer according to the AJCC/
UICC TNM system, 8th edition (2017)
27.5.1 Background TX Primary tumor cannot be assessed
T0 No evidence of primary tumor
T1a Tumor size maximum 2 cm, limited to the thyroid
Thyroid cancer (TC) is the most frequent endocrine malig-
T1b Tumor size maximum 1 cm, limited to the thyroid
nancy. According to the National Institutes of Health
T2 Tumor size >1 cm up to a maximum of 2 cm, limited to the
Surveillance, Epidemiology, and End Results (SEER) data- thyroid
base, in the past 20 years, TC incidence has increased signifi- T3 Tumor size >2 cm up to 4 cm, limited to the thyroid
cantly in the United States, Canada, Australia, and Western T3a Tumor size >4 cm, limited to the thyroid, or any tumor with
Europe and some Asian countries [19]. Since the mortality is macroscopic extrathyroidal extension (musculus
low (0.5%), the prevalence is very high; incidence is higher sternohyoideus, musculus sternothyroideus, musculus
omohyoideus)
in women than in men (3:1). TCs are distinguished into well- T3b Tumor size >4 cm, limited to the thyroid any tumor with
differentiated thyroid cancer (DTC), poorly differentiated macroscopic extrathyroidal extension (musculus
thyroid cancer (PDTC), anaplastic thyroid cancer (ATC), sternohyoideus, musculus sternothyroideus, musculus
and medullary thyroid cancer (MTC). DTC types, which rep- omohyoideus)
resent the majority of all thyroid carcinomas with favorable T4a Any tumor size with extrathyroidal extension beyond the
thyroid capsule and invasion of the subcutaneous soft tissue,
prognosis, include papillary thyroid cancer (PTC) and fol- larynx, trachea, esophagus, and/or recurrent laryngeal nerve
licular thyroid cancer (FTC). ATC is a rare tumor (<3% of all T4b Any tumor size with invasion of prevertebral fascia,
TCs) derived from follicular cells with a very poor prognosis mediastinal vessels or carotid artery
and almost invariably lethal, with >90% mortality rate at Nx Regional lymph nodes cannot be assessed
1 year. PDTC is similar to ATC; it is less common and has N0 No regional lymph node metastases
poor prognosis. MTC, which constitutes about 5% of all TCs N1 Regional lymph node metastases
N1a Lymph node metastases unilateral in level VI or upper
and can present as sporadic or familial forms, is a well-
mediastinum
differentiated cancer arising from the parafollicular, N1b Metastases in other unilateral, bilateral, or contralateral
calcitonin-producing C-cells; its 5-year survival rate is cervical lymph nodes (level I, II, III, IV, and V) or
approximately 86% for MTC patients. Other very rare TCs retropharyngeal
account for non-epithelial tumors, lymphoma, and carcino- M0 No distant metastases
mas characterized by the presence of mucin-producing cells M1 Distant metastases
and keratin.
the American Thyroid Association (ATA) is based on the risk

27.5.2 Staging of recurrence after initial treatment [21] and distinguishes
low, intermediate, and high risk to develop recurrence during
Although many staging systems have been proposed in the follow-up. This classification is useful for planning the thera-
past, the most widely used system for TC is the tumor–node– peutic strategies, and the risk level can be modified after
metastasis (TNM) classification system defined jointly by therapy and during follow-up; this so-called ongoing risk re-
the International Union Against Cancer (Union Internationale stratification has been demonstrated to have useful clinical
Contre le Cancer, UICC) and by the American Joint impact [22].
Committee Against Cancer (AJCC). It is based on the histo-
pathologic description of the local condition (tumor size,
node metastases) and of distant metastases (M). Age (<55 or 27.5.3 Clinical Presentation
>55 years at diagnosis) is also important in this classifica-
tion. The latest eighth edition [20] TNM classification issued Thyroid tumors usually present as a thyroid nodule, either
in 2017 is detailed in Table 27.2. The corresponding prog- isolated or within a MNG. Although thyroid nodules are com-
nostic implications in terms of observed 5-year survival for mon, especially in subjects older than 50 years, TC represents
patients with PTC and FTC are shown in Fig. 27.5. only 4–5% of all thyroid nodules. Diagnostic work-up of thy-
ATCs are always classified as stage IV according to the roid nodules for the differential diagnosis between benign
TNM classification, subdivided into IVA (intrathyroidal and malignant lesions is therefore necessary. Complete medi-
tumors), IVB (extrathyroidal extension but without distant cal history of the patient and physical examination of the neck
metastases), and IVC (with distant metastases). are important steps in this process. In fact, history of familial
Although the TNM classification in not very useful in pre- thyroid cancer and/or of exposure to radiation during child-
dicting the risk of recurrence, although it is relevant to the hood increases the risk of malignancy, while accurate palpa-
survival probabilities, while the classification proposed by tion yields information on location and size of the nodules,
Fig. 27.5 Prevalence of 70

different histotypes of thyroid
cancers in a consecutive series
of over 8000 patients 60
followed at the Department of
Endocrinology of the
Fraction of cancer patients, %

University of Pisa Medical 50
School (Pisa, Italy) in the
period 1969–2005
(reproduced with permission 40
from: Strauss HW, Mariani G,
Nuclear Oncology – From 30
Applications, 2nd Edition.
New York: Springer; 2017) 20
10
0
Papillary Follicular Medullary Unknown Anaplastic Lymphoma
their consistency and motility during swallowing, and on any It has been demonstrated that 47% of patients diagnosed
extra-nodular thyroid tissue. Neck examination can also clar- with ATC had previous or concurrent DTC, thus indicating
ify the presence and extent of lymph node involvement, which that a high fraction of ATCs develop through dedifferentia-
is frequently found in PTC [23]. tion of DTCs. It is therefore not surprising that genomic
Nodules that are firm in consistency and irregular or fixed alterations observed in DTCs are also present in ATCs and
to local tissues, palpable neck lymph nodes, rapid increase in may be used as potential targets. Since approximately
size of the nodules over weeks or months, and dysphagia or 20–40% of ATCs display the BRAF V600E mutation, these
hoarseness should undergo further diagnostic work-up, patients could benefit from BRAF inhibitors therapy.
although these signs are not definitely specific for a However, RAS mutations are also prevalent in ATCs, and
malignancy. their coexistence with the BRAF mutation induces resistance
The clinical presentation of an ATC is usually more to vemurafenib.
impressive than that of a DTC: a big and firm thyroid nodule, MTC is a well-differentiated thyroid tumor maintaining
recently developed, is the most common presentation often the biochemical and pathological features of the parafollicu-
associated with dysphonia or other signs/symptoms of local lar, calcitonin-producing C-cells from which it derives.
compression/invasion, such as dysphagia or dyspnea. Most There are at least four features that make the C-cells a sepa-
ATCs are detected in stage IVB, with unresectable extrathy- rate entity from follicular cells: (1) the peculiar distribution
roidal mass. Despite many attempts with different drug regi- in the thyroid gland, which is prevalent at the junction of the
mens, there are still no hopes for a better future. The upper third and the lower two-thirds and along the central
outcomes of chemotherapy with single or multiple agents, vertical axis of each thyroid lobe; (2) the total growth and
radiotherapy, and combined treatment are disappointing, and functional independence from TSH; (3) the inability to con-
such treatments do not improve survival. The common clini- centrate and retain iodide; and (4) the production and secre-
cal history reported by patients is the rapid growth of a neck tion of calcitonin, a biogenic amine which is almost
lesion either in the thyroid bed or in the lateral cervical exclusively produced by normal and malignant C-cells.
region. The average size of the mass at the time of initial MTC occurs either sporadically or as an inherited dis-
diagnosis is quite large (8 cm in maximum diameter, ranging ease. The majority of MTC cases, nearly 75%, are sporadic.
from 3 to 15 cm). Regional lymph node metastases and vocal Familiar MTC is inherited in an autosomal dominant pattern
cord paralysis are present in 30–40% of the cases [24]. and constitutes a part of the “multiple endocrine neoplasia”
Systemic hematogenous metastases occur in up to 75% of type 2 (MEN2) syndrome. There are two clinically distinct
the patients, the lung being the most common site (80%), types of MEN2 syndrome: MEN2A and MEN2B. MEN2A
followed by the bone, skin, and brain [25]. Cardiac and intra- is defined by the presence of MTC, primary hyperparathy-
abdominal metastases have also been described [25]. Death roidism (PHP), and pheochromocytoma (PHEO). MTC is
is usually related to extensive local disease with ultimate air- diagnosed in nearly 100% of patients, whereas the risk of
way obstruction rather than to distant metastases. PHEO and PHP is substantially lower (50% and 20–30%,
respectively). Clinical manifestation of the MEN2B syn- minimal aggressiveness and indolent course of most of the
drome involves MTC, PHEO, and typical clinical features thyroid cancers.
with marfanoid habitus and multiple mucosal neuromas Some ultrasound features are more frequently observed in
among others. Hereditary MTC is caused by function muta- malignant thyroid nodules [28]. In particular, hypoecho-
tions in the RET proto-oncogene. According to the current genicity, the absence of a peripheral halo, irregular borders,
ATA guidelines on MTC, RET mutations are stratified into microcalcifications, and hypervascularization are strongly sug-
three MTC risk levels based on of genotype–phenotype cor- gestive of malignancy, especially if combined. No differences
relations, the penetrance of RET mutations, and aggressive- have been observed in the neck ultrasound features for the pap-
ness: moderate, high, and highest. illary, follicular, or medullary TCs. Neck ultrasound is so rele-
Positive immunohistochemistry for calcitonin is diagnos- vant in the evaluation of thyroid nodules that it plays a leading
tic of MTC, although positivity for chromogranin A and CEA role in the choice of the nodules to be submitted to FNAC [2].
may also be useful. Calcitonin is a sensitive and specific Current ATA guidelines recommend diagnostic thyroid
marker used for the diagnosis and monitoring of MTC ultrasonography to be performed only in patients with
patients, since only <1% of MTCs are low-differentiated can- known or suspected thyroid nodules or in the presence of
cers not producing calcitonin. The baseline calcitonin level risk factors [2].
may be slightly elevated in patients with renal failure, autoim-
mune diseases, hyperparathyroidism, FTC, some neuroendo-
crine pancreatic cancers, prostate cancer, or lung cancer. 27.5.5 Fine Needle Aspiration Cytology
However, in cancers other than MTC, calcitonin concentra-
tion does not increase in stimulation tests (using calcium or Fine needle aspiration cytology (FNAC) is currently the
pentagastrin). Calcitonin level is related to the stage of the method of choice for presurgical diagnosis of TC, as it can
disease. It is used to assess MTC progression and prognosis, reliably screen the nature of a thyroid nodule. In particular,
as well as in the follow-up of a hereditary (associated with the it can identify with a remarkable degree of certainty those
germline mutation of the RET proto-oncogene) and sporadic cases requiring thyroid surgery, either because of a definite
MTC. Both baseline and stimulated calcitonin concentrations diagnosis of PTC or because of findings requiring histologic
influence the timing of prophylactic thyroidectomy in appar- elucidation (as is the case, e.g., cytology with a microfol-
ently healthy carriers of the RET mutation [26]. licular pattern). FNAC is best performed under US guid-
It is important to note that the serum levels of TSH, FT3, ance, which improves its diagnostic accuracy by decreasing
and FT4, that are routinely assayed to assess thyroid func- the number of inadequate specimens and false-negative
tion, are not helpful for the differential diagnosis between results; in fact, the acellular or nondiagnostic aspirate is
benign and malignant nodules. Also the serum level of thyro- reduced from 14% to 8% in US-guided FNAC, especially if
globulin, which is usually elevated in case of nodular goiter, the needle is passed more than once in the nodule under
is not informative as to the nature of the nodule. On the other evaluation.
hand, elevated serum calcitonin levels can preoperatively The ATA guidelines [29] recommend FNAs in specific
suggest the possibility of an MTC, and further evaluation to circumstances, as follows:
confirm this suspicion is required.
• Nodules ≥1 cm in greatest dimension with a high suspi-
cion sonographic pattern
27.5.4 Ultrasonography • Nodules ≥1 cm in greatest dimension with an intermedi-
ate suspicion sonographic pattern
Ultrasound can image thyroid structures, internal/surround- • Nodules ≥1.5 cm in greatest dimension with low suspi-
ing blood flow, and adjacent tissues. It is being widely cion sonographic pattern
employed for nodule detection, characterization, risk stratifi- • Nodules ≥2 cm in greatest dimension with very low sus-
cation, treatment monitoring, and post-thyroidectomy cancer picion sonographic pattern (e.g., spongiform)
surveillance. Compared with other imaging modalities,
ultrasound has high spatial resolution, uses no ionizing radi- Guidelines issued by the Society of Radiologists in
ation, and has low cost. For these reasons, ultrasound is the Ultrasound (SRU) [29] are also widely used and are some-
first-line imaging tool for evaluation of the thyroid. Despite what simpler than the ATA guidelines. According to the SRU
a lack of evidence for screening efficacy and the significant criteria, FNAC is recommended in case of:
harm potential associated with overdiagnosis, ultrasound is
in common use in some countries as a screening tool for thy- • Nodules 1.0 cm in largest diameter if microcalcifications
roid cancer [27]. Nevertheless, the screening for thyroid nod- are present
ules by ultrasound, or by any other types of imaging studies, • Nodules 1.5 cm in largest diameter with any of the fol-
is not recommended in the general population because of the lowing signs:
–– Solid or almost entirely solid nodules that have an indeterminate cytology and to optimize
–– Coarse calcifications within the nodule surgical planning. The major concern is the cost of these
• Nodules 2.0 cm in largest diameter and mixed solid and methods and the fact that they are available only in specific
cystic or almost entirely cystic with a solid mural centers and cannot be routinely used [32].
component
• Nodules that have shown substantial growth since prior
ultrasound examination
Key Learning Points
• Thyroid cancers (TCs) are classified into well-
According to the National Cancer Institute Thyroid FNAC
differentiated thyroid cancer (DTC), poorly differ-
State of the Science Conference, the cytological categories
entiated thyroid cancer (PDTC), anaplastic thyroid
differ in their implied risk of malignancy [30]. The following
cancer (ATC), and medullary thyroid cancer (MTC).
five categories are denominated “Thy1” to “Thy5” (nondiag-
• DTC types, which represent the majority of all thy-
nostic, benign, indeterminate, suspicious-for-malignancy,
roid carcinomas with favorable prognosis, include
and malignancy), as follows:
papillary thyroid cancer (PTC) and follicular thy-
roid cancer (FTC).
• Thy1: nondiagnostic aspirate due to inadequate sample or
• Although the TNM classification is relevant as to
to the presence of technical artifacts; aspirates that con-
the probability of survival, it is not very useful for
tain only cyst fluid, histiocytes, and erythrocytes are con-
predicting the risk of recurrence.
sidered inadequate; a minimum of six clusters of at least
• The classification proposed by the American
ten follicular cells assures the adequacy of cytological
Thyroid Association (ATA) is based on the risk of
sampling.
recurrence after initial treatment and distinguishes
• Thy2: benign, nonneoplastic lesion consistent with nodu-
low, intermediate, and high risk to develop recur-
lar goiter or thyroiditis.
rence during follow-up.
• Thy3: cytology is not diagnostic (it represents 10–15% of
• FNAC is currently the method of choice for presur-
all FNACs); this category includes all (micro)follicular
gical diagnosis of TC, as it can reliably screen the
lesions, for which the differential diagnosis between
nature of a thyroid nodule.
benign follicular lesion and FTC or follicular variant of
• The following five cytological categories have been
PTC should be considered; subclassification of this cate-
defined: Thy1 (nondiagnostic), Thy2 (benign),
gory into Thy3A and Thy3B identifies groups with a
Thy3A (indeterminate low risk), Thy3B (indetermi-
lower (10–15%) and higher (40%) risk of malignancy,
nate high risk), Thy4 (suspicious for malignancy),
respectively.
and Thy5 (malignancy).
• Thy4: lesion suspicious for malignant carcinoma.
• Thy5: diagnosis of malignancy.
Nondiagnostic and benign cytologies are characterized by 27.5.6 Radionuclide Imaging of DTC
a low risk of malignancy (1–4% and 0–3%, respectively),
whereas the risks of malignancy for suspicious-for- While thyroid scintigraphy with either radioiodide or 99mTc-
malignancy and malignant cytologies are high (60–75% and pertechnetate can distinguish nodules as “hot” (i.e., function-
97–99%, respectively). The indeterminate categories are ing and usually benign) or “cold” (i.e., nonfunctioning), it
characterized by an intermediate risk of malignancy [31]. cannot distinguish the benign or malignant nature of cold
Immunohistochemical analysis (e.g., HBME-1, galectin- nodules. In fact, expression of the Na+/I− symporter (NIS)
3, and CXCR4) and molecular analysis for genetic biomark- gene in DTC is markedly reduced (10- to 1200-fold) with
ers involved in DTC (such as RET/PTC, PAX-8-PPARγ, and respect to normal thyroid follicular cells [33], thus resulting
BRAF mutations) could be useful for evaluating indetermi- in a “cold” appearance on conventional thyroid scintigraphy
nate category cytologies (mostly Thy3) and also for prognos- with either radioiodide or 99mTc-pertechnetate. However,
tic purposes in case of malignancy. Methods for molecular equally “cold” appears on thyroid scintigraphy of any benign
analysis follow two different approaches: those that “rule nonfunctioning adenomas, where expression of the NIS gene
out” malignancy (to reduce the overtreatment of benign nod- is decreased by 2- to 700-fold. Therefore, because of its low
ules) and those that “rule in” malignancy (to optimize surgi- specificity, the thyroid scan is employed as an adjunctive
cal planning). The “rule-out” malignancy approaches aim to rather than as a first-line diagnostic test. In particular, a thy-
identify benign nodules that have an indeterminate cytology roid scan is helpful when a hyperfunctioning nodule is con-
and to reduce the overtreatment of benign nodules, while sidered in the differential diagnosis because of a suppressed
“rule-in” malignancy approaches aim to identify malignant TSH level [33]. Furthermore, thyroid scintigraphy may be
considered as a possible additional diagnostic aid in those sensitivity and 100% negative predictive value for [18F]FDG
cases with undetermined FNAC (class Thy3) associated with PET/CT, although with lower specificity (69%), overall accu-
serum TSH levels seemingly within normal limits or border- racy (79%), and positive predictive value (62%) [34].
line with suppressed levels. More recently, another study prospectively investigated
On the other hand, all the scintigraphic approaches with the diagnostic performance of [18F]FDG PET/CT, 99mTc-
tumor-seeking radiopharmaceuticals (such as, e.g., 99mTc- Sestamibi scintigraphy, and multiparametric neck US in a
Sestamibi) that have been proposed for diagnosing DTC in similar group of 87 patients with Thy3 FNAC. [18F]FDG
patients with nodular goiter have failed because of their very PET/CT had significantly greater sensitivity (94%), diagnos-
low specificity. In fact, these agents accumulate both in TC tic accuracy (66%), and negative predictive value (98%) than
and in benign adenomas; considering that DTC only consti- either 99mTc-Sestamibi scintigraphy or neck ultrasound.
tutes <5% of all thyroid nodules (very low pretest probability), Association of PET/CT with ultrasound findings signifi-
this feature translates into an alarmingly low specificity that cantly increased specificity but decreased sensitivity [35].
only triggers further diagnostic testing without appreciably On the other hand, increased uptake of [18F]FDG in the
restricting the population of patients with suspected lesions. thyroid gland is detected in a certain fraction (up to 4%) of
Although the above consideration holds true also for the patients undergoing an [18F]FDG PET/CT scan for thyroid-
nonspecific tumor-imaging agent [18F]FDG, the very high neg- unrelated reasons, an occurrence defined as “incidentaloma”
ative predictive value of [18F]FDG PET should nevertheless be that can predict thyroid malignancy especially when tracer
noticed. In particular, Giovanella et al. evaluated with [18F] uptake is focal rather than diffuse (Figs. 27.6 and 27.7).
FDG PET/CT a group of 88 euthyroid patients with a solitary A meta-analysis of 31 articles reporting the results of 197,296
thyroid nodule and nondiagnostic FNAC (Thy3 category); in PET studies and 3659 focal thyroid incidentalomas concluded
their series, none of the 44 patients with a negative [18F]FDG that the pooled proportion of malignancy in patients with
scan had malignancy, while all 21 patients with DTC had a incidentalomas is 19.8% (95% CI, 15.3–24.7%) [36]. Based
positive [18F]FDG scan (as also had 18 out of the 67 patients on these results, further diagnostic investigation (e.g., with
with a benign thyroid nodule). These findings result in 100% image-guided FNAC) is to be recommended in all patients
Fig. 27.6 Examples of thyroid

incidentalomas found on a
whole-body [18F]FDG PET/CT
performed for non-thyroid-
related indications; although with
similar patterns of increased
tracer uptake in the thyroid bed,
final diagnosis ascertained by
FNAC and/or histology can be
quite different. For all cases:
axial CT image in left panel, with
corresponding axial [18F]FDG
PET image in right panel. (a)
Patient with cervical cancer: b
mildly increased [18F]FDG
uptake in the thyroid with a
diffuse pattern, more evident in
left lobe. Final diagnosis: chronic
thyroiditis. (b) Patient with
gastric cancer: focal increased
[18F]FDG uptake in right thyroid
lobe. Final diagnosis: adenoma-
tous goiter. (c) Patient with
cervical cancer: focal increased
[18F]FDG uptake in left thyroid c
lobe. Final diagnosis: papillary
thyroid cancer (modified and
reproduced with permission
from: Liu Y. Clinical significance
of thyroid uptake on F18-
fluorodeoxyglucose positron
emission tomography. Ann Nucl
Med. 2009;23:17–23)
Fig. 27.7 Incidental detection of papillary thyroid cancer (PTC) in a ter). Radioiodide remnant ablation with 131I-iodide was successful, and
patient undergoing [18F]FDG PET/CT during follow-up for malignant subsequent long-term follow-up was favorable (from Volterrani D, Erba
melanoma. FNAC from the [18F]FDG-avid lesion in the left thyroid PA, Mariani G. Fondamenti di Medicina Nucleare—Tecniche e
lobe (SUVmax > 14) showed a Thy4 pattern. Histology after thyroidec- Applicazioni. Milan: Springer, 2010, with permission of Springer
tomy revealed a well-differentiated PTC (>2 cm in maximum diame- Science + Business Media)
with thyroid incidentaloma detected in the [18F]FDG PET/CT the use of radioiodide as in the case of DTC. Instead, simi-
scan performed for any reason, because of the relatively high larly to what happens in other types of biologically aggres-
proportion of thyroid malignancy in these patients—as also sive cancers, ATC cells express high levels of glucose
confirmed by more recent large-cohort studies [37]. transporters, in particular GLUT1 and GLUT3. Therefore,
PET/CT with [18F]FDG constitutes the main radionuclide
imaging technique potentially useful in ATC patients, both
27.5.7 Radionuclide Imaging of ATC for initial staging and during follow-up for the detection of
residual or metastatic disease after the initial treatment [38]
Since the ATC cells typically lack the prominent features of (Fig. 27.8). On the other hand, morphologic imaging tech-
the original thyroid follicular cells, in particular their niques such as ultrasound, CT, or MRI also play an impor-
response to stimulation by TSH and the ability to concen- tant role for defining the local extent of disease in patients
trate iodine, the radionuclide techniques employed for stag- with ATC, especially in order to assess infiltration of the
ing and monitoring patients with ATC cannot be based on anatomic structures surrounding the thyroid.
Fig. 27.8 [18F]FDG PET/CT performed for primary staging in a patient cates and restricts the trachea. A lung metastasis in the left parahilar
with anaplastic thyroid cancer affecting predominantly the left lobe. region is also detected (modified from Volterrani D, Erba PA, Mariani
Coronal, transaxial, and sagittal sections (CT, PET, and fused PET/CT) G. Fondamenti di Medicina Nucleare—Tecniche e Applicazioni. Milan:
define well the local extent of disease, showing a large lesion that dislo- Springer, 2010, with permission of Springer Science + Business Media)
27.5.8 Radionuclide Imaging of MTC targets upon which radionuclide-based imaging and therapy
are based are therefore different from those utilized in the diag-
Although MTC develops in the same gland, it constitutes a dis- nostic and therapeutic approach to DTC as defined above. In
tinct neoplastic disease that differs from DTC with regard to particular, MTC shares with other n euroendocrine neoplasms
cancer origin, molecular pathogenesis, tumor aggressiveness, some features regarding uptake and intracellular processing of
treatment modalities, and long-term outcomes. The molecular some amines and expression of cell membrane receptors for
regulatory peptides. Besides the production of the hormone show persistently elevated serum calcitonin level after sur-
calcitonin, another somewhat peculiar feature of MTC in the gery, regardless of radical surgical approach. MTC relapse
scenario of neuroendocrine neoplasms is that MTC cells is diagnosed in more than 50% of them during a mean
express sizable amounts of the tumor-associated CEA. 10-year follow-up. It should be emphasized that not only the
Despite high diagnostic sensitivity especially in the postop- absolute calcitonin value but also its dynamics (i.e., change
erative work-up, at present the use of DMSA labeled with pen- in serum levels over time) plays an important role in MTC
tavalent technetium-99m (99mTc(V)-DMSA) as an MTC- follow-up.
imaging agent is interesting only from the historical point of If biochemical remission is not obtained, the source of
view, although this radiopharmaceutical is occasionally still elevated calcitonin level should be localized. However, imag-
employed as a reference in multiple agent comparative studies. ing studies are useless if calcitonin level is below 150 pg/mL,
Based on the production of CEA by MTC cells, consider- because of a very low probability of the detection and local-
able efforts have been devoted to develop novel approaches ization of MTC recurrence. The ATA recommends assess-
to imaging, and possibly therapy, with the use of anti-CEA ment of calcitonin and CEA levels every 3–6 months to
monoclonal preparations labeled with either indium-111, determine their doubling time [40]. If an increase in calcito-
radioiodine, technetium-99m, or gallium-68 for PET imag- nin level exceeds 150 pg/mL, imaging studies to detect MTC
ing [39]. However, all such efforts have not translated into recurrences are recommended [40].
imaging/therapy approaches that can be used in the current Neck ultrasound should be performed first, as lymph
clinical practice of nuclear medicine. node recurrence is the most common location of relapse,
Nevertheless, the most intensely investigated nuclear medi- particularly when only a slight increase in serum calcitonin
cine applications for the management of patients with MTC, concentration is noticed. The early detection and adequate
from both the imaging and the therapeutic viewpoints, are treatment of MTC relapse has an important impact on dis-
based on two biological features expressed by several types of ease prognosis and patient survival. An analysis of the
neuroendocrine neoplasms, namely, the ability to take up and results of reoperation due to lymph node recurrence revealed
concentrate biological amines and the enhanced expression of that if calcitonin level before reoperation was >1000 pg/mL,
membrane-associated receptors for a variety of biological pep- a final biochemical remission was obtained only in 1 of 76
tides, most notably somatostatin. The first feature is the basis patients. When calcitonin level is higher than 500 pg/mL,
for tumor targeting with radioiodinated MIBG for either imag- additional imaging studies including chest and abdominal
ing purposes or therapeutic purposes and for PET imaging CT or MRI and PET ([18F]FDG PET/CT or 18F-DOPA PET/
with 18F-DOPA. The second feature (expression of receptors CT) should be performed to confirm metastatic dissemina-
for regulatory peptides) constitutes the basis for either single- tion [40].
photon or PET imaging as well as for possible therapy using
properly designed and radiolabeled somatostatin analogs.
As in the case of other neuroendocrine neoplasms with a
somewhat indolent pattern of tumor growth, the diagnostic
performance of the most widely used PET tracer in oncol- Key Learning Points
ogy, [18F]FDG, is suboptimal in patients with MTC, unless • Patients with multinodular thyroid have the same
the disease progresses toward a more aggressive pattern of risk of having a malignancy as patients with a single
tumor growth. Therefore, the first-line approach to radionu- thyroid nodule.
clide imaging for these tumors is represented by the use of • Thyroid fine needle aspiration cytology (FNAC)
radiopharmaceuticals addressing the specific neuroendo- has the inherent limitation of yielding in determined
crine features mentioned above rather than by the use of [18F] results in 20–30% of the cases.
FDG. Conversely, as neuroendocrine neoplasms take on a • Thyroid surgery is required for a definitive diagno-
biologically aggressive pattern of growth, they tend to lose sis in these cases.
some of the biological features specific of the tumor of ori- • FNAC, in conjunction with ultrasonography, con-
gin; in these cases, PET with [18F]FDG performs better than stitutes the cornerstone of thyroid nodule
imaging with the more neuroendocrine-specific tracers. evaluation.
Postoperative assessment of the calcitonin and CEA • Based on adequate sampling, FNAC is 95% accu-
serum levels is recommended in all MTC patients. rate for diagnosing thyroid cancer.
Normalization of serum calcitonin after surgery indicates an • If biochemical remission is not obtained in MTC
excellent treatment outcome. The 10-year overall survival in patients after surgery and the calcitonin level is
patients with normal calcitonin level after total thyroidec- >500 pg/mL, additional imaging studies including
tomy with an adequate range of cervical lymphadenectomy [18F]FDG PET/CT or 18F-DOPA PET/CT should be
is nearly 100%. However, 80% of patients with palpable performed to confirm metastatic disease.
MTC and 50% of those with nonpalpable macroscopic MTC
27.6 Treatment of Thyroid Cancers Radionuclide therapy has no role in the treatment of ana-
plastic or poorly differentiated TCs, because these tumors
27.6.1 Overall Scenario cannot concentrate radioiodide; in these patients therapy is
based on surgery (whenever possible—also with simple deb-
According to the most recent guidelines [2], thyroid remnant ulking/palliative purposes), chemotherapy, and
ablation with 131I-iodide is indicated in DTC patients with a EBRT. Prognosis of these patients is very poor.
moderate to high risk of recurrence. Patients are prepared Surgery is the first-choice option for the treatment of
with a low-iodine diet (for at least 2 weeks) and with either patients with medullary TCs (MTC). Follow-up is based on
withdrawal of replacement therapy with l-thyroxine (L-T4) monitoring of biochemical tumor recurrence (rising serum
for a period sufficient to achieve serum TSH levels >30 μIU/ calcitonin and/or CEA levels) and on diagnostic imaging
mL (usually within 4–6 weeks) or administration of exoge- with radiopharmaceuticals that are commonly used for the
nous human recombinant TSH. Administered activities of whole spectrum of neuroendocrine tumors: 123I-MIBG and
131
I-iodide vary considerably between 1.11 and 3.7 GBq (30– radiolabeled somatostatin analogs (111In-DTPA-
100 mCi), according to the estimated risk of recurrence. A pentetreotide, Ga-DOTA-NOC/TATE) in addition to the
68
whole-body scintigraphy (131I-WBS, preferably completed overall tumor-imaging agent [18F]FDG. Radionuclide ther-
with SPECT/CT acquisitions at selected sites) is performed apy of recurrent/metastatic MTC includes 131I-MIBG and
4–7 days after 131I-iodide administration, as it can detect 90
Y- or 177Lu-DOTA-TOC/NOC/TATE. As in the case of
lymph node involvement or unexpected metastases. Therapy radioiodine-refractory DTC, treatment of advanced recur-
with 131I-iodide causes in some instances side effects that are rent/metastatic MTC includes novel targeted therapies such
usually mild, transient, and well manageable. More impor- as TKIs.
tant side effects (e.g., a radiation-induced second malig-
nancy) can be expected only after repeat administration of
radioiodide for therapy of recurrent metastatic disease and 27.6.2 Primary Treatment of DTC
for very high cumulative doses.
The major diagnostic modalities employed to follow 27.6.2.1 Surgery
patients with DTC treated with remnant ablation is measure- Unlike follicular thyroid cancer (FTC, which usually pres-
ment of serum thyroglobulin (Tg), 131I-WBS, and neck ultra- ents as a solitary nodule), papillary thyroid cancer (PTC) fre-
sound. Additional diagnostic imaging (e.g., CT, MRI) is quently exhibits multifocal growth. Until recently, total
applicable in selected circumstances; in fact, although CT thyroidectomy was always indicated as initial treatment
and MRI can in principle localize very small lesions in the when the diagnosis of PTC was known before surgery, e.g.,
neck, chest, and bones, the features of such lesions are rarely when the ultrasound findings had been typical for cancer, or
specific for recurrent/metastatic DTC. fine FNAC had shown a Thy4 or Thy5 pattern. In fact, given
If lymph node metastasis is suspected on the basis of the high rate of multifocality and bilaterality of PTC, total
imaging findings, a FNAC should be performed, also thyroidectomy results in lower recurrence rates than lobec-
assaying Tg in the needle washing. Serum Tg levels that tomy or subtotal thyroidectomy [41]. However, with the
become detectable or show an increasing trend during wide diffusion of neck ultrasound and the possibility to
follow-up indicate the need for further evaluation, possi- detect early possible recurrence in the residual lobe, simple
bly with additional 131I-iodide therapy in case of small lobectomy is nowadays indicated when the PTC is in a single
metastatic lesions (<1 cm in size) with a relatively high nodule and of small size [2]. The main reason for this choice
radioiodine uptake. In case of bulkier lymph node metas- is the lower prevalence of complications after lobectomy
tasis, surgery is preferable—possibly followed by adju- than after total thyroidectomy, either vocal cord palsy or per-
vant radioiodine therapy. External beam radiation therapy manent hypoparathyroidism.
(EBRT) has little role in the treatment of recurrent/meta- When locoregional disease is not detected preoperatively
static DTC, except in case of tumors exhibiting a trend to (e.g., by ultrasound examination of the neck [42] or preop-
dedifferentiation. erative FNAC), the prevailing consensus is that there is no
Patients with recurrent DTC may develop poorly differen- need to routinely perform lateral lymph node dissection in
tiated lesions that do not concentrate radioiodide. In these patients with DTC. Conversely, some authors suggest central
patients [18F]FDG PET/CT has an important prognostic role lymph node dissection for primary surgery in lymph node-
and is useful to determine the sites and extent of these metas- positive PTC, in order to abrogate the serum levels of Tg—
tases. When DTC becomes refractory to 131I-iodide therapy, which would then become a reliable tumor-associated marker
metastatic and progressive lesions can be treated with [43]. Nevertheless, if neck lymph node metastases, either in
tyrosine- kinase inhibitors (TKI) such as sorafenib and the central or lateral compartment, are diagnosed by ultra-
lenvatinib. sound examination before thyroidectomy, surgery must
include lateral neck dissection [44]. Selective excision of

confirmed lymph node metastases (the so-called “cherry
picking” or “node plucking” technique) is justified only in
those patients with thyroid cancer who have already had sys-
tematic dissections of the respective neck area(s) during
prior operations [45].
27.6.2.2 A djuvant Ablation of Postsurgical

Remnants with 131I-Iodide
The rationale of performing radioiodide therapy of DTC
after surgery in order to ablate postsurgical thyroid remnants
is based on the selective ability of thyroid cells to concen-
trate iodine, which is then utilized to synthesize Tg and thy-
roid hormones. This capacity (which is maintained by DTC
tumor cells to a much lesser extent than in normal thyroid
cells) is significantly enhanced by TSH stimulation, a condi-
tion that can be achieved either by discontinuing replace-
ment/suppressive hormonal therapy with L-thyroxine or by
administering exogenous human recombinant TSH (see fur- Fig. 27.9 Spot planar imaging of the upper chest and neck region
ther below); full benefits of the diagnostic and therapeutic (anterior view) obtained 7 days after administration of 3.7 GBq
uses of 131I-iodide can be achieved on the basis of this (100 mCi) 131I-iodide for thyroid remnant ablation in a patient treated
property. for PTC. In addition to at least two obvious thyroid tissue remnants with
intense radioiodide uptake (solid black arrows), there are at least two
Ablation of postsurgical thyroid remnants with 131I-iodide areas with lower intensity uptake (indicated by open arrows), indicating
is indicated in DTC to eliminate any residual, macroscopi- lymph node lesions that were previously unknown; FNAC with assay of
cally normal thyroid tissue left after total or near-total thy- Tg in the needle washing confirmed the presence of metastasis (from
roidectomy [46–48]. This therapeutic strategy ensues four Volterrani D, Erba PA, Mariani G. Fondamenti di Medicina Nucleare—
Tecniche e Applicazioni. Milan: Springer, 2010, with permission of
major benefits: Springer Science + Business Media)
1. Since 50% of PTCs are multifocal, 131I-iodide ablation

may decrease the recurrence and mortality rates, by most recent ATA guidelines do not recommend as mandatory
destroying postoperative microscopic residual tumor foci. the use of 131I-iodidethyroid remnant ablation in patients
2. 131I-iodide ablation of normal appearing residual thyroid with lymph node-negative, unifocal intrathyroidal microcar-
tissue facilitates the early detection of recurrences based cinomas (T1aN0M0) [2]. Also in patients with larger tumors
on sequential measurements of serum Tg. In this regard, (>1.5 cm), with multifocal tumors, or with extension beyond
it should be noted that thyroid remnants synthesize and the thyroid capsule or with lymph node metastasis, the so-
secrete Tg that is indistinguishable from Tg possibly pro- called “intermediate-risk” cases, the benefits of ablation with
duced by residual thyroid cancer cells. 131
I-iodide continue to be debated [46, 48–52]. Therefore, in
3. 131I-iodide ablation makes it possible to perform a highly these patients the use of radioiodine for the purpose of ablat-
sensitive diagnostic whole-body scan (131I-WBS) ing normal remnant tissue should be “selective” and debated
4–7 days after administration of the ablative 131I activity, case by case, although the prevailing consensus is that these
which may reveal previously undetected metastatic tumor conditions should indeed be treated [53]. In patients with
lesions, as identified by foci of uptake outside the thyroid high-risk DTC, there is instead a clear need for 131I-iodide
bed (Fig. 27.9). treatment to prolong disease-free and overall survival (see
4. By reducing/ablating the mass of normal thyroid tissue, Table 27.3)
any subsequent therapeutic use of 131I-iodide in case of On the other hand, even if a real clinical benefit of
recurrent/metastatic DTC will be more effective. 131
I-iodide remnant ablation is not fully recognized for lower-
risk tumors, its role in making easier the follow-up of DTC
However, since the indication for ablation with 131I-iodide patients in any risk category has been proposed for many
is based on prognostic factors and no benefits from radioio- years as one of the rationales to perform it. In fact, since
dine ablation for patients with intrathyroid tumors <1.5 cm either normal or neoplastic thyroid tissue may be responsible
have been demonstrated [46, 48–50], the prevailing consen- for Tg production in patients with thyroid remnants, serum
sus is that these “low risk”thyroid cancer patients should not Tg measurement will be more useful for detecting persistent/
undergo postsurgical radioiodine ablation. In this regard, the recurrent disease after all sources of Tg secretion alternative
Table 27.3 Indications for radioiodide treatment in DTC patients administration of any given activity of 131I-iodide, the
Indication to amount of radiation delivered to the tumor is small com-
Risk category TNM (6th ed.) Benefits I-iodide
l3l
pared to that delivered to normal thyroid tissue. A quantita-
Low T1aN0M0 None No tive dosimetric approach determines the therapeutic activity
Intermediate T1b-T2N0M0 Could Yes/noa required to deliver a sufficient dose of radiation to treat and
reduce
relapses can also predict the outcome of 131I-iodide therapy. The rec-
High T1-2N1M0-1, Reduces Yes ommended radiation dose for ablation of thyroid remnants
T1-2N0M1, T3-4N0- relapse and is 300 Gy [47].
1, and/or M0-1 death It should also be considered that the dosimetric burden
European Consensus: Pacini F, Schlumberger M, Dralle H, Elisei R, must not exceed the maximum permissible irradiation to the
Smit JW, Wiersinga W; European Thyroid Cancer Taskforce. European
whole body and particularly to the bone marrow (2 Gy); to
consensus for the management of patients with differentiated thyroid
carcinoma of the follicular epithelium. Eur J Endocrinol. 2006;154:787– this purpose, an empiric estimate has been suggested, i.e., to
803. Erratum in: Eur J Endocrinol. 2006;155:385 administer 131I-iodide activities such that whole-body reten-
Reproduced with permission from: Strauss HW, Mariani G, Volterrani tion at 48 h post-administration should not exceed 4.44 GBq
D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology to
(or 120 mCi). Furthermore, particular attention should be
Clinical Applications, 2nd Edition. New York: Springer; 2017
a
“Yes”: European Consensus; “no”: ATA guidelines paid to dosimetric estimates in patients with diffuse lung
metastases, in whom radioiodine might cause pulmonary
fibrosis over certain values of accumulation/retention; in
to tumor tissue (i.e., normal thyroid remnants) have been these patients the estimated whole-body retention of 131I at
eliminated. Any detectable serum Tg level after total thyroid- 48 h should not exceed 2.96 GBq (80 mCi) [57, 58].
ectomy and successful remnant ablation should therefore Nevertheless, at the same time, therapeutic activity to be
indicate the presence of residual tumor. Nevertheless, over administered should be calculated so as to ensure the deliv-
the year it has been recognized that the trend of increase of ery of an effective dose to the tumor lesions; if this condition
serum Tg is more important than the absolute value to fol- cannot be achieved because of low radioiodine uptake, then
low-up these patients [54], especially when combined with a it would be appropriate to consider therapies alternative to
neck ultrasound scan performed by expert endocrinologists radioiodine therapy, such as surgery (whenever possible),
or radiologists [55]. EBRT, and/or chemotherapy [47, 58].
Furthermore, also the diagnostic 131I-WBS is more sensi- Administration of 131I-iodide in activities prefixed accord-
tive and specific after radioiodine ablation of normal thyroid ing to certain thresholds irrespective of pretreatment dosi-
remnants; moreover, imaging with radioiodide WBS and metric estimates is still the most widely employed approach.
ultrasound evaluation of the neck are the only reliable diag- Its main advantage is simplicity, although probably not be in
nostic modalities in patients in whom the assay of serum Tg the optimal solution for all cases; in fact, any prefixed activ-
is unreliable because of the presence of circulating anti-Tg ity may be excessive in some cases and insufficient in others.
autoantibodies. For ablation purposes, a single administration of either 1.11
Although 131I-iodide can be very successful in achieving or 3.7 GBq (30 or 100 mCi) of 131I-iodide ensues complete
all of the above objectives, many factors may affect the destruction of the remnants in 90% of the cases when uptake
degree of success of 131I-iodide thyroid remnant ablation. of a tracer amount of radioiodide in the residue is <2%
These include, but are not limited to, the extent of surgery, (which translates into a residual thyroid tissue mass < 2 g),
percent radioiodide uptake in normal residual thyroid tissue, while ablation is successful in only two-thirds of the cases
volume of normal residual thyroid tissue, prescribed activity when uptake is >2% (i.e., a residual thyroid mass >2 g) [47].
of 131I-iodide, geometrical shape of normal residual thyroid Accurate individualized dosimetric estimates require a set
tissue, effective half-life of 131I in the normal residual thyroid of information on the target tissue. Considering that the radi-
tissue, patient’s compliance with low-iodine diet, serum ation dose is basically delivered energy per mass, such infor-
level of TSH, prior use of diagnostic 131I-iodide with possible mation includes the volume (or mass) of the target tissue and
stunning, and definition of successful radioiodine ablation. the retention kinetics of 131I-iodide in that target. While
The kinetics of radioiodine accumulation in DTC cells imaging techniques can provide volumetric estimates of the
differs from that in normal thyroid tissue. In the latter, target tissue, the kinetics of uptake and retention of radioio-
uptake of radioiodide is approximately 1% of the adminis- dide can be derived from serial measurements following
tered activity per gram of tissue, with a biological half-life administration of a tracer activity of 131I-iodide (such as, e.g.,
of about 8 days; whereas, in the tumor tissue, uptake of 3.7 MBq); since such measurements are usually limited to
131
I-iodide varies from 0.5 to 0.001% of administered activ- only 24 h (or 48 h at the latest) post-administration, certain
ity per gram, with half-lives ranging from a few days to a assumptions derived from standard pharmacokinetic models
few hours [56]. This difference explains why, following are adopted for describing the kinetics of radioiodide reten-
tion from the last time-point onward. Although preferable in u seful to limit the hypothyroidism symptoms, because of the
principle to more simplistic empiric approaches, from a much faster metabolic clearance of T3 versus that of T4 (bio-
practical standpoint, the approach of administering 131I-iodide logic half-life about 10 h versus 15 days, respectively).
activities based on pretreatment dosimetric estimates has However, T3 requires a three times per day administration,
four major drawbacks, which apply not only to the remnant and many patients refer tachycardia soon after its
ablation case but also to the case of radioiodide therapy of consumption.
recurrent/metastatic lesions from DTC, as follows: The use of rhTSH (Thyrogen®) is approved in the United
States and Europe for pre-therapeutic stimulation before
• The estimated volume in which radioiodide is concen- administering radioiodide for ablation of thyroid tissue rem-
trated is in general only an approximation derived from, nants after near-total or total thyroidectomy for DTC, in
e.g., ultrasound evaluation; moreover, this volume cannot patients without evidence of distant metastases. Very recently
be measured in patients with metastases that are too small two independent multicentric randomized studies, one per-
for viewing by imaging techniques, as is the case for dif- formed in France [61] and one in the United Kingdom [62],
fuse lung metastases that are not discernible as separate have clearly demonstrated that low-risk and intermediate-
lesions by chest X-ray or CT. risk DTC patients can successfully be ablated either when
• The biological half-life of 131I-iodide in the target tissues treated with 1.11 or 3.7 GBq 131I-iodide (30 or 100 mCi) and
differs from one patient to another and even between dif- regardless of the type of TSH stimulation (i.e., obtained after
ferent metastatic lesions in any given patient and in differ- withdrawal of L-T4 therapy or after rhTSH administration).
ent periods; thus, dosimetric estimates based on standard After these studies, the use of rhTSH for remnant ablation
models may have a wide range of approximation. has been approved in combination with activities ranging
• The intestinal absorption of 131I-iodide can vary even over from 1.11 to 3.7 GBq (30–100 mCi) of 131I-iodide.
short intervals of time, such as those occurring between Irrespective of the modality of TSH stimulation (i.e.,
administration of the tracer activity for the dosimetric through increased endogenous secretion following L-T4
study and administration of the actual therapeutic withdrawal or through administration of exogenous rhTSH),
activity. an important component of patient’s preparation for radioio-
• Dose distribution in a certain tumor target tissue can be dide therapy is to ensure that the pool of native iodide in the
heterogeneous, as shown by the typical patchy distribu- body fluids is reduced as much as possible. In fact, reducing
tion of radioiodide visualized on tissue autoradiographs. competition from unlabeled iodide for uptake of radioiodide
In this case, voxel dosimetry could be useful for predict- in the thyroid cells increases the chances for the radioactive
ing the possibility of either whole or partial ablation of iodide atoms to be accumulated in the thyroid cells or in the
tumor tissue. In this regard, 3D dosimetry makes it pos- DTC cells. Depletion of the iodide pool in the body is
sible to obtain the so-called dose-volume histogram, a achieved by following a low-iodine diet for at least 2 weeks
map of the different absorbed dose values in different por- before radioiodide administration and by avoiding excessive
tions of the target tissue [59]. iodine intake through other sources, such as, e.g., drugs,
iodinated contrast agents, etc. (see Table 27.1) [63, 64] (see
Thyroid remnant ablation with 131I-iodide must be per- also practical procedural guidelines listed in Chap. 46 of this
formed under adequate TSH stimulation, which is necessary book).
in order to enhance the uptake of radioiodine by follicular The absence of iodine contamination leading to increased
thyroid cells. Ablation of thyroid remnants can be performed iodine pool in the body should always be verified by measur-
by administering 131I-iodide after two i.m. injections of ing urinary iodine content, preferably normalized as iodine/
0.9 mg human recombinant TSH (rhTSH) performed 48 and creatinine ratio (expressed as μg/g); in case this value is
24 h before the treatment [2, 60]. If rhTSH administration is greater than 250 μg/g [64], the opportunity should be consid-
not available, a period of L-T4 withdrawal of at least 4 weeks ered of administering 131I-iodide only after an adequate fur-
is required; in this case, serum TSH should be >30 μIU/mL ther period on a low-iodine diet, in order to reduce the urinary
before radioiodide is given, otherwise the withdrawal period iodine/creatinine ratio below such level [65]. Assessing uri-
is prolonged for another 1–2 weeks. The main drawback of nary iodine might be relevant especially in those patients
this traditional modality of patient’s preparation for radioio- who have a negative post-ablation/therapy scan despite the
dide therapy is the development of hypothyroidism symp- evidence of residual/recurrent disease as assessed by serum
toms after about 2–3 weeks from hormone withdrawal and Tg. In women of child-bearing age, pregnancy must be
continuing for other 2–3 weeks after the reintroduction of the excluded before administering radioiodine for therapy.
L-T4 therapy. Replacing L-T4 with T3 for 2 weeks before Although the issue of thyroid stunning (characterized by a
complete hormonal withdrawal and taking T3 with L-T4 reduction of therapeutic radioiodide uptake consequent to
during the 2–3 weeks after the radioiodine treatment can be prior administration of a diagnostic activity of 131I-iodide) was
recognized more than 50 years ago [66, 67], it still is a matter 27.6.3 Side Effects of Radioiodide Therapy
of debate, and its clinical significance and even its existence
have been questioned [68]. The frequency of thyroid stunning, Short-term side effects of therapy with 131I-iodide for thyroid
a condition that potentially compromises the therapeutic effi- remnant ablation are usually minimal and transient [92, 93].
cacy and outcome of radioiodide treatment, appears to increase Nausea and gastric pain occur frequently after 131I-iodide
with increasing activities of radioiodide administered in the treatment and last a few days; these symptoms can be pre-
diagnostic phase [69, 70]. Indeed, the exposure of follicular vented/treated by administering proton-pump inhibitors or
thyroid cells (or of follicular-derived tumor cells) to sublethal similar drugs. Sialoadenitis is also frequent, with pain and
doses of radiation from 131I-iodide (as it might be caused by enlargement of the salivary glands but rarely leading to per-
radioiodide administered for a diagnostic 131I-WBS, usually in manent xerostomia; for prophylaxis, patients are advised to
the order of 185 MBq) may decrease the efficiency of subse- drink large quantities of fluids and to stimulate salivary
quent treatment with 131I-iodide [71, 72]. This effect is associ- secretion with lemon juice. Loss or change of taste is reported
ated with non-ablative radiation doses (up to 10 Gy) that by the majority of patients receiving therapy with 131I-iodide,
reduce by as much as 50% the ability of the tissue to concen- but usually lasts only a few days. When large thyroid rem-
trate and store radioiodide [67]. nants are present and completion thyroidectomy is not fea-
Reduced NIS expression [73] seems to be the main deter- sible, steroids should be given to prevent swelling in the neck
minant of thyroid stunning, as it is involved in the response due to radiation-induced inflammation.
to irradiation-induced DNA damage [74, 75]. Because of the As to the long-term side effects, the overall relative risk
fear of this effect, a diagnostic 131I-WBS is usually not rec- of a second primary tumor is increased only in patients
ommended before thyroid remnant ablation with 131I-iodide. treated with high cumulative activities of 131I-iodide, an
If a diagnostic 131I-iodide scan prior to ablation therapy is occurrence that is encountered in patients with recurrent/
deemed useful (e.g., in patients with suspected metastases or metastatic DTC. In particular, a significantly greater risk of
to evaluate the size of the remnants), then it is recommended leukemia or of other second primary tumors has been
to administer no more than 74 MBq of 131I-iodide and to per- reported for patients treated with cumulative activities
form ablation within 48–72 h. Alternatively, 123I-iodide can exceeding 22 GBq (or 600 mCi) of 131I-iodide, especially if
be used for scintigraphic pre-ablation assessments [76], as radioiodide therapy has been associated with external radio-
the low-energy Auger electrons emitted by this radionuclide therapy [94]. It should also be emphasized that some reports
are devoid of any stunning effect on thyroid follicular cells. support the notion that, although the incidence of second
For radiodosimetric purposes using simplified models, primary malignancies is indeed increased in patients with
uptake of radioiodide in the thyroid remnant can be mea- DTC, this increase is not linked to long-term effects of
sured (and scintigraphy of the neck can be performed) by radioiodide therapy but rather to some not yet identified
administering a very low activity of 131I-iodide, such as 1.85 genetic factors [95, 96].
or 3.7 MBq (50 or 100 μCi). Considering that many patients with DTC are women in
A post-therapy 131I-WBS performed 4–7 days after admin- their reproductive age, one of the concerns for radioiodide
istration of 131I-iodide for thyroid remnant ablation is therapy is the possibility of long-term adverse effects on
extremely useful to evaluate the distribution and the sites of fertility and on the outcomes of pregnancies posttreatment.
radioiodide uptake. Scintigraphic detection of focal areas of In this regard, a recent study based on 2673 pregnancies in
radioiodide uptake outside the thyroid bed demonstrates the DTC patients who had been treated with radioiodide con-
presence of metastatic disease, so that additional treatments cluded that there is no evidence that exposure to radioio-
must be considered, either further radiometabolic therapy (at dine affects the outcome of subsequent pregnancies and
an adequate interval, e.g., at least 6 months after ablation) or offspring [97].
surgery. Similarly as with the post-ablation scan, the presence Radioiodine treatment may result in transient impair-
of additional metastatic foci has been reported in 10–26% of ment of male gonadal function, consisting in a reduction of
patients in the post-therapy scan compared to those detected motility and reduced sperm count in a variable number of
in the diagnostic scan, leading to change the management patients [98]. Changes in the germinal epithelium and
strategy in 9–15% of the patients [77, 78]. Post-ablation Leydig cell function (with consequent increase in serum
SPECT/CT fusion imaging provides superior diagnostic FSH levels) are usually transient and may have some clini-
information, concerning especially lymph node involvement, cal significance in subjects with a pre-existing impairment
than planar 131I-WBS or even stand-alone SPECT (Figs. 27.10 of fertility [99]. Radiation fibrosis may develop in patients
and 27.11) [79–91]. SPECT/CT plays therefore an important with diffuse lung metastases who have received repeated
role in the evaluation of disease after ablation therapy, as it excessive activities (over 5.55 MBq, or 150 mCi) of radio-
may alter management of these patients, either by upstaging iodide at short intervals, especially if less than 6 months
or by downstaging the disease in about 25% of the cases [82]. [100, 101].
Fig. 27.10 131I-WBS (left

panel), spot planar images (upper
right panels), and SPECT/CT
acquisitions of the cervical region
obtained 8 days after administra-
tion of 3.7 GBq (100 mCi)
131
I-iodide for thyroid remnant
ablation in a 21-year-old man
submitted to thyroidectomy
because of a PTC occupying
virtually the entire right thyroid
lobe (4.5 cm in maximum
diameter). The planar whole-
body scan reveals two foci of
uptake in the right paramedian
aspect of the neck; spot
acquisitions in the anterior view
and right oblique view confirm
the presence the two foci of
uptake, one more intense, almost
central just above the sternal
notch (indicated by the open
circle), and one less intense,
more laterally and superiorly; the
inferior focus of intense uptake
was interpreted as representing
postsurgical thyroid remnant,
while the focus of uptake located
more laterally and superiorly was
interpreted as possibly represent-
ing lymph node metastasis.
SPECT/CT (CT component in
upper left box, SPECT compo-
nent in upper right box, fused
SPECT/CT in lower left box,
MIP in lower right box) confirms
that the focus of less intense
uptake corresponds to an
enlarged lymph node. FNAC
with assay of Tg in the needle
washing confirmed the presence
of metastasis (reproduced with
permission from: Strauss HW,
Mariani G, Volterrani D, Larson
SM, Eds. Nuclear Oncology –
From Pathophysiology to Clinical
New York: Springer; 2017)
27.6.4 Evaluation of Success of Thyroid • RhTSH-stimulated serum Tg levels <1 ng/mL in the
Remnant Ablation absence of detectable anti-Tg antibodies (although a 2 ng/
mL threshold is frequently adopted in the clinical routine,
The success of thyroid remnant ablation is usually evaluated and the FDA recommends the 2.5 ng/mL threshold).
6–12 months after radioiodine therapy based on the follow- • Absence of suspicious findings on neck ultrasound.
ing criteria [60, 102]:
• Negative radioiodine uptake in the thyroid bed (“no vis- 27.6.5 Levothyroxine Therapy
ible uptake in the thyroid bed” or, if visible, uptake
<0.1%, an arbitrarily set threshold) on diagnostic The two main goals for L-T4 treatment in PTC and FTC
131
I-WBS. patients after thyroidectomy are (1) to correct postoperative
Fig. 27.11 Left panel: post-ablation 131I-WBS in a 33-year-old man nent at the top, CT component in the middle, fused SPECT/CT at the
submitted to thyroidectomy and left cervical lymphadenectomy because bottom), demonstrating that the focal area of uptake was actually bone
of PTC of the left lobe (follicular variant) with presurgical diagnosis of metastasis located in the posterior arc of a left rib. Right panel: coronal
lymph node metastasis; serum Tg under rhTSH stimulation was 94 ng/ SPECT/CT section of the pelvis (SPECT component at the top, CT
mL. Besides intense radioiodine uptake in the thyroid remnant, there component in the middle, fused SPECT/CT at the bottom), demonstrat-
was an area of focal uptake in the left cervical region (interpreted as a ing that the focal area of uptake lateral to the bladder was actually
lymph node metastasis in the 131I-WBS), while activity accumulated in metastasis in the left iliac bone (from Mariani G, Bruselli L, Kuwert T,
the bladder hampered correct interpretation of the pelvic region. Center et al. A review on the clinical uses of SPECT/CT. Eur J Nucl Med Mol
panel: axial SPECT/CT section of the cervical region (SPECT compo- Imaging. 2010;37:1959–1985, with permission)
hypothyroidism and (2) to suppress the endogenous secre- also important to emphasize that L-T4 suppressive therapy
tion of TSH (which is a potent growth factor for follicular in children has no adverse effect on bone maturation, final
thyroid cells as well as for follicular-derived tumor cells), at height, and pubertal development. Since a 25% increase in
least until the evidence that the patient has been cured is the FT4 level is usually observed when blood is drawn
achieved. 3–4 h after ingestion of L-T4, patients should be instructed
The optimal daily L-T4 dose for suppressing TSH not to take their medication the morning before blood
secretion ranges between 1.6 and 2.8 μg/kg body weight. testing.
However, this dose must be tailored on an individual basis, An important issue for DTC patients is for how long
since it is only roughly correlated with body weight and L-T4 therapy must be continued on a TSH-suppressive
age. Younger patients, especially children, and patients dose. Patients with the evidence of persistent/recurrent dis-
with a high lean body mass usually require higher doses ease or at a high risk of recurrence should be kept on sup-
per kg of body weight [103]. By contrast, older patients pressive therapy in order to decrease the risk of recurrence
require lower L-T4 daily doses, and further reductions are [107], whereas in patients with favorable prognostic factors
needed for subjects with concurrent heart disease [104]. In and evidence of complete cure (i.e., a negative 131I-WBS
addition to monitoring serum TSH (which should be sup- and undetectable TSH-stimulated serum Tg in the absence
pressed, i.e., less than 0.1 μIU/mL), the serum levels of of anti-Tg autoantibodies), the dose of L-T4 may be
FT3 and FT4 must also be measured; even when the TSH decreased with the aim of maintaining serum TSH levels
suppression is required, both FT3 and FT4 levels should be between 0.1 and 0.5 μU/mL.
in the normal range in order to avoid possible iatrogenic In contrast, both ATC and MTC, which are constituted by
thyrotoxicosis [105]. When patients are monitored as TSH-independent cells, do not benefit from TSH suppres-
described above, L-T4 suppressive therapy is safe and sion and only require L-T4 replacement therapy to correct
devoid of any long-term adverse effects [106, 107]. It is postoperative hypothyroidism.
Undetectable Tg on L-T4
Key Learning Points
• Thyroid remnant ablation with 131I-iodide is indi- AbTg and neck
US: negative
cated in DTC patients with a moderate to high risk
of recurrence, whereas it is not recommended in
patients with a low risk of recurrence. rhTSH-Tg
Stimulation test
• Administered activities of 131I-iodide for ablation
vary considerably between 1.11 and 3.7 GBq (30–
100 mCi), according to the estimated risk of Tg- Tg+
recurrence. follow-up
• A post-therapy 131I-WBS performed 4–7 days after
administration of 131I-iodide for thyroid remnant
ablation is extremely useful to evaluate the distribu- 131I-iodide
therapy
tion and the sites of radioiodide uptake.
• Different protocols have been developed for
patient’s preparation before radioiodide therapy, Fig. 27.12 Suggested algorithm of follow-up in DTC patients treated
concerning both a suitable stimulation of radioio- with total thyroidectomy and postsurgical remnant ablation with
131
I-iodide. When serum Tg on L-T4 suppressive therapy is undetect-
dide uptake achieved through serum TSH levels able, an rhTSH stimulation test for Tg should be performed to confirm
>30 μU/mL and adequate reduction of the iodine the absence of recurrent/metastatic disease. A negative rhTSH stimula-
pool in the body. tion test (left arm of the diagram) does not require further therapy, and
the patient is included in a regular follow-up program. A positive rhTSH
stimulation test (right arm of the diagram) indicates instead the pres-
ence of recurrent/metastatic disease and should be treated with further
radioiodine therapy; localization of the sites of recurrence and/or
27.6.6 Follow-Up of DTC After Primary metastasis is usually possible in the post-therapy 131I-WBS (preferably
integrated with SPECT/CT) (reproduced with permission from: Strauss
Treatment HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology –
From Pathophysiology to Clinical Applications, 2nd Edition. New York:
The two aims of follow-up for patients treated with surgery Springer; 2017)
and with radioiodide remnant ablation for DTC are (1) to
maintain adequate L-T4 therapy and (2) to discover the per-
sistence or recurrence of thyroid cancer early on (Fig. 27.12). According to the method used for assaying serum Tg,
Follow-up is virtually lifelong because, especially in DTC TgAb may cause false-positive results (i.e., when employ-
patients, recurrences can occur even 10–20 years after appar- ing a radioimmunoassay method) or false-negative results
ently successful initial treatment [50, 108]. The major diag- (i.e., when employing an immunoradiometric assay
nostic modalities employed for the follow-up of DTC method) [50]. It should also be noted that even noncom-
patients are measurement of serum Tg, 131I-WBS, and neck petitive immunometric assays based on detection tech-
ultrasound examination. Evaluation of these parameters at niques other than radioactivity counting may be affected
6–12 month intervals enables the clinicians to assess the by circulating heterophilic autoantibodies [112]. In
clinical status of the patients and to identify early those patients with circulating TgAb, who account for about
patients who need additional treatment(s) [109, 110]. 20–25% of all patients with DTC [113], serum Tg mea-
surement alone cannot be relied upon if not correlated to
27.6.6.1 Monitoring of Serum Tg serum TgAb.
Since Tg is produced exclusively by normal or neoplastic 2. Since TSH is a growth factor for both normal follicular
thyroid follicular cells, it constitutes an excellent tumor- cells and follicular-derived cancer cells, the production of
associated marker after removal of the thyroid gland. In fact, Tg (therefore its level in circulating blood) is kept to a
serum Tg should be undetectable after total thyroidectomy minimum under L-T4 suppressive therapy. Therefore,
and successful thyroid remnant ablation. Whereas, when the undetectable serum Tg levels under L-T4 suppressive
serum Tg levels of these patients remain detectable or rise, therapy (which would indicate cure of the disease) must
they indicate the presence of persistent or recurrent disease be confirmed by a TSH stimulation test.
[111]. There are two major limitations with regard to the
measurement of serum Tg as a tumor marker for DTC: Tg stimulation is currently performed by administra-
tion of exogenous rhTSH rather than by discontinuing
1. The presence of circulating anti-Tg autoantibodies administration of suppressive L-T4 so as to induce hypo-

(TgAb) may interfere with the measurement of serum Tg. thyroidism; in fact, the latter is an uncomfortable condi-
rhTSH rhTSH With the introduction and diffusion of the ultrasensitive Tg

0.9 mg i.m. 0.9 mg i.m.
assays [116], the rhTSH stimulation test can be avoided when
the basal serum Tg is <0.1 ng/mL and reserved to those cases
with a serum Tg level on the gray zone (i.e., from 0.1 to 1 ng/
mL). In fact, in this latter group, the stimulation test can
clearly distinguish the true “positive” cases (about 25%) that
require further investigation from negative cases for which
Days 0 1 2 3 4 5 the clinical remission of the disease can be defined [117].
TH, TSH
TSH Tg Tg
Tg, AbTg 27.6.6.2 Whole-Body Scintigraphy
with 131I-Iodide
Fig. 27.13 Schedule of rhTSH administration for the Tg stimulation
test: on day 1 blood is taken before the administration of rhTSH for A I-WBS performed 6–12 months after thyroid remnant
131
assaying thyroid hormones, TSH, Tg, and Tg-Antibodies (TgAb). ablation is useful in the follow-up of patients with high or
RhTSH is injected i.m. at day 1 and 2; at day 3 serum TSH is assayed intermediate risk for persistent disease, although its diagnos-
again to verify the correct stimulus, while serum Tg levels are measured
tic sensitivity is lower than that of the serum Tg stimulation
again at days 4 (optional) and 5 (reproduced with permission from:
Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear test [118]. Similarly as for the Tg stimulation test, radioio-
Oncology – From Pathophysiology to Clinical Applications, 2nd dide scintigraphy is performed under TSH stimulation (fol-
Edition. New York: Springer; 2017) lowing either L-T4 withdrawal or administration of rhTSH);
images are usually acquired 72 h after administration of
tion that, though only temporary, may profoundly impact 185 MBq (5 mCi) of 131I-iodide if patients have been pre-
the quality of life of many patients [114]. The protocol for pared with L-T4 withdrawal or 48 h in case of preparation
performing rhTSH administration to stimulate serum Tg with rhTSH. When there is high risk of persistent/recurring
(i.e., the rhTSH stimulation Tg test) is illustrated in disease (such as, e.g., detectable serum Tg levels with sup-
Fig. 27.13. pressed TSH levels), 123I-Iodide can be used instead of
An rhTSH-stimulated serum Tg measurement is usually 131I-iodide in order not to jeopardize with a possible stunning
obtained 6–12 months after total or near-total thyroidectomy effect the efficacy of subsequent radioiodide therapy. In this
and 131I-iodide ablation in DTC patients with undetectable instance, images are usually acquired 48 or 24 h after admin-
serum Tg (<0.1 ng/mL when using the newer-generation, istering i.v. 185 MBq of 123I-iodide depending on preparation
ultrasensitive assays) on L-T4 suppression therapy. The of the patient, i.e., with L-T4 withdrawal or with rhTSH
resulting change in serum Tg levels at the peak of rhTSH stimulation, respectively [119, 120].
stimulation (48–72 h after the last of two 0.9 mg doses in the Diagnostic radioiodide WBS is not necessary in low-risk
euthyroid state) versus baseline is recorded. According to the patients who are clinically free of disease, with undetectable
cutoff of serum Tg chosen (which may vary in different insti- (<0.1 ng/mL) serum Tg while on L-T4 suppressive therapy,
tutions but is usually set at 2 ng/mL although the FDA rec- and whose neck ultrasound examination is negative [118],
ommends 2.5 ng/mL), the following situations may be provided they have not circulating TgAb possibly interfering
encountered [115]: with the serum Tg assay (see above).
Serum Tg levels that become detectable upon TSH stimu-
• Serum Tg remains undetectable (<0.1 ng/mL); up to 65% lation indicate the need for further treatment with 131I-iodide
of the low-risk patients have no rise in serum Tg concen- [121, 122]. Clearly, a positive serum Tg stimulation test does
trations after rhTSH. not provide any information on topographic location of per-
• Serum Tg rises between 1 and 2 ng/mL; this occurs in sistent/recurring disease nor on the ability of tumor lesions to
about 20% of low-risk patients and necessitates further concentrate radioiodide (therefore on their susceptibility to
periodic rhTSH stimulation tests, perhaps as often as once be treated with further 131I-iodide therapy). Such information
yearly. Neck ultrasound plays an important diagnostic would instead be provided by a positive radioiodide WBS. On
role in this group. In some patients, the high TSH- the other hand, some patients with DTC may have persistent/
stimulated serum Tg level will spontaneously decline recurring disease (as demonstrated by their abnormal serum
over 1 or more years [109], while in others it will steadily Tg levels), yet their tumor lesions may have lost the ability to
rise, and metastases will become apparent [110]. concentrate and retain radioiodide, either as a biologic fea-
• Serum Tg rises above 2 ng/mL; this occurs in about 20% ture of some DTC histologic types [123, 124] or at some
of low-risk patients and usually indicates a substantial point in the evolution of the disease after having been treated
thyroid remnant or residual cancer. Patients with this pat- with 131I-iodide therapy [125].
tern usually require further imaging evaluation to localize Whereas, if stimulated serum Tg remains undetectable (in
the source of circulating Tg. the absence of serum TgAb), the patient can be considered as
cured [125]. The diagnostic 131I-WBS plays an important dose absorbed by the tumor tissue is at least 80–100 Gy, a
role when it is not possible to rely neither on baseline nor on target that is, however, difficult to achieve when the tumor
stimulated serum Tg because of a positive serum TgAb titer. lesion is large (i.e., greater than 1 cm in maximum diameter).
In this regard, some role as a surrogate tumor marker can be Therefore, in case of bulkier recurrences, but still limited to
played also by the TgAb titer itself, which should decline lymph nodes (as shown by morphologic imaging, such as
and disappear when the disease is cured (as a consequence of ultrasound, CT, and/or MRI), the first therapeutic option
reduced/absent antigenic stimulation on TgAb production), should be surgery, whenever possible [21]. In this clinical
while it usually persists elevated or increases even further in setting, 131I-iodide can play a role as adjuvant therapy,
case of persistent or recurring disease [126]. whereas the initial therapeutic approach can be based on
administration of radioiodide in patients with small lymph
27.6.6.3 Neck Ultrasound node metastasis showing high radioiodide uptake [21]. The
Neck ultrasound examination is an integral component of 131
I activity to be administered can be based on dosimetric
follow-up evaluation in all DTC patients, especially those at estimates aiming at releasing the maximal safe dose to the
high risk for recurrent disease and those with suspicious clin- patient, yet with a radiation dose >85 Gy to the lymph
ical findings. Metastatic lymph nodes can be detected with node(s), as well as to metastatic lesions at other sites
neck ultrasound in the lateral and, even though to a lesser (Fig. 27.14). Alternatively, an empiric activity based on aver-
degree, in the central neck compartments. The ultrasound age estimates and prior experience can be administered, usu-
features of metastatic lymph nodes differ in a number of ally 3.7–5.5 GBq of 131I-iodide.
respects from those of possibly enlarged but benign lymph Although imaging modalities, such as ultrasound, CT,
nodes. In particular, benign lymph nodes are usually smaller, and MRI, generally have high sensitivities, their interpreta-
thin, or oval in shape and have a conspicuous hilus, while tion may be difficult in previously dissected areas, because
lymph nodes with metastasis are roundish in shape, may dis- of anatomic distortion and scar tissue. In contrast, 131I-WBS
play microcalcification, are hypoechoic, and do not have a may demonstrate neoplastic foci more specifically.
visible hilus [42]. However, if a lymph node metastasis is Furthermore, a 131I-WBS performed after administration of a
suspected, an FNA should be performed and the aspirate be large (i.e., therapeutic) amount of 131I-iodide may detect new
sent both for cytology and for Tg measurement [125]. previously unrecognized lesions in up to 60–80% of the
cases and remains the most sensitive diagnostic/localizing
imaging modality inpatients with functioning metastases
from DTC [127]. Small and not growing cervical lymph
Key Learning Points nodes should not necessarily be treated. A “wait-and-see”
• The major diagnostic modalities employed to fol- strategy is nowadays suggested in these cases since the neck
low patients with DTC treated with remnant abla- ultrasound can easily allow to monitor these lesions to ensure
tion is measurement of serum thyroglobulin (Tg), timely surgical treatment or any other appropriate therapeu-
131
I-WBS, and neck ultrasound. tic strategy [128].
• Additional diagnostic imaging is applicable in
selected circumstances.
• Although CT and MRI can in principle localize 27.6.8 Management of Distant Metastases
very small lesions in the neck, chest, and bones, the
features of such lesions are rarely specific for recur- Distant metastases occur in a subgroup of DTC patients and
rent/metastatic DTC. can be discovered from 2 to 3 up to more than 10 years after
initial treatment. Most distant metastases from DTC are
located in the lung (50–60%) or bones (20–30%). Liver,
brain, and skin metastases, which are much more rare, are
27.6.7 Management of Locoregional found in about 3% of patients at the time of the identification
Recurrences from DTC of a diffuse metastatic disease. The site of metastatic disease
depends on a number of factors. Metastases in the lungs are
Although recurrences are usually identified during the first more frequently observed in patients with PTC and in
year postprimary treatment, they may develop later on in the younger patients; indeed, they are virtually the only sites of
course of follow-up [49, 50]. Detection of a local or regional distant metastases in children [129, 130]. In contrast, bone
recurrence dictates complete work-up to evaluate its extent metastases are more frequently observed in patients with
and to search for distant metastases. Therapy with radioio- FTC and in older patients [127].
dide ensues complete response in 75–80% of the patients DTC patients with metastatic disease should undergo
with regional lymph node metastasis, provided the radiation complete diagnostic work-up. A 131I-WBS performed after
a b c
Fig. 27.14 Histology of lesion removed from the lumbar spine responding to an average absorbed dose of 250 Gy per lesion based on
revealed metastasis from well-differentiated thyroid cancer in a 75-year- administration of 3.7 GBq (100 mCi) of 131I-iodide. The two bulky
old woman submitted 3 years earlier to thyroidectomy because of mul- lesions in the skull were removed surgically before radioiodine therapy.
tinodular goiter; histology had been reported as negative for malignancy The post-therapeutic 131I-WBS scan confirmed high uptake in the meta-
and was not available for review. A dosimetric study until 72 h post- static bone lesions, with only some residual uptake in the skull (b). At
administration of 123I-iodide was performed under hrTSH stimulation the 1-year follow-up, serum Tg was undetectable with suppressed TSH,
and converted into 131I-equivalent data. Besides obvious uptake in the peaking at 0.6 ng/mL following rhTSH stimulation; the diagnostic131I-
postsurgical thyroid remnant, scintigraphy revealed numerous bone WBS was completely negative (c). Over 6 years later, a follow-up contin-
metastases scattered throughout virtually all the skeleton, including two ued to be negative (from Volterrani D, Erba PA, Mariani G. Fondamenti
bulky lesions in the skull (a); peak serum Tg was 3810 ng/mL. Estimates di Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer, 2010,
of the absorbed radiation dose were very favorable, i.e., 63.5 mGy cor- with permission of Springer Science + Business Media)
the administration of a therapeutic activity of 131I-iodide 150 mCi) of 131I-iodide in adults and approximately 37 MBq/
(3.7 GBq or 100 mCi) may clarify the extent and location of kg body weight (or 1 mCi/kg) in young children [138–141].
metastatic disease. In particular, a 131I-WBS so performed Higher activities, i.e., 7.4 GBq (200 mCi), or more have
often reveals lung metastases not visible on plain chest instead been recommended inpatients with bone metastases
X-rays nor on CT. In this regard, only about half the patients [141]. In patients with diffuse, small lung metastases that are
with normal chest X-rays, who have diffuse uptake of not visible on X-rays, a dosimetric approach is not feasible,
131
I-iodide in the lungs, have peripheral micronodular lung as the mass of neoplastic tissue cannot be estimated. In these
metastases detectable in the CT scan [131, 132]. MRI is par- cases, some centers adopt the so-called Benua–Leeper
ticularly useful in the work-up of metastases of the spine and approach, which is based on the maximum tolerable absorbed
base of the skull. Since most bone metastases from DTC are dose to the organ at risk [142]. For therapy with 131I-iodide,
hypervascularized, arterial embolization may help to facili- the organ at risk is the red bone marrow; 2 Gy is the maxi-
tate subsequent surgical removal [133–137]. mum limit that the red marrow can absorb. The maximum
Conventional therapeutic strategies such as surgery, tolerable administered activity (i.e., the activity that would
radioiodide, EBRT, and systemic therapy with the new “tar- imply a 2 Gy dose to the red marrow) can be calculated using
geted” therapies can all be used to treat metastatic DTC. A simple parameters, such as the blood kinetics following
radiation dose >80 Gy should be delivered to achieve cure; administration of a tracer activity of 131I-iodide [143].
instead, the chances of success are remote in case of radia- The outcome of 131I-iodide therapy in patients with pul-
tion doses <35 Gy [9]. Although based one empiric estimates monary metastases from DTC is variable from patient to
that do not take into account uptake measurements for each patient (Figs. 27.15 and 27.16), and not obviously is related
lesion in each individual patient, recommended therapeutic to the initial total uptake of radioiodide in the lungs.
activities that should satisfy such desired radiation dose Neoplastic masses are probably smaller in patients with
threshold are generally in the range of 3.7–5.5 GBq (100– lower uptake. Subsequent treatments with 131I-iodide are
a b
Fig. 27.15 Post-therapeutic 131I-WBS images obtained in two separate ment following repeated 131I-iodide therapies totaling 28.5 GBq
occasions after administration of 131I-iodide in a patient with pulmonary (770 mCi) (serum Tg 8 ng/mL under endogenous TSH stimulation);
metastases from DTC. (a) October 2005 (at the age of 25 years): diffuse thymic uptake (frequently seen in younger patients) is now more evi-
uptake in the pulmonary parenchyma, indicating military-type meta- dent in the post-therapy scan (from Volterrani D, Erba PA, Mariani
static involvement (serum Tg 21 ng/mL following TSH stimulation G. Fondamenti di Medicina Nucleare—Tecniche e Applicazioni. Milan:
achieved with L-T4 withdrawal). (b) June 2009: progressive improve- Springer, 2010, with permission of SpringerScience + Business Media)
given 6 months later for 1–2 years, and then annually until lesions [144]. For retreatment with radioiodide, the serum
there is no more evidence of residual radioiodide uptake on TSH level must be >30 μIU/mL. As mentioned above, such
post-therapeutic 131I-WBS. Of course, between two treat- high TSH levels can be obtained through withdrawal of L-T4
ment courses, L-T4 TSH-suppressive therapy is continued therapy, with resulting transient hypothyroidism; unfortu-
with the aim of reducing serum TSH to levels not stimulating nately, the use of recombinant human TSH has not been
tumor growth. approved yet for the preparation of patients with metastatic
Although there is virtually no limit to the cumulative DTC before therapy with 131I-iodide (except in case of failure
131
I-iodide activity that can be given to patients with distant of the pituitary to produce TSH), although there are evi-
metastases, the risk of second primary malignancies (espe- dences that it can allow to obtain the same results than the
cially leukemia) rises slightly for cumulative administered withdrawal of L-T4 [145–147].
activities higher than 18.5–22.2 GBq (500–600 mCi) [93]. Pre-therapy evaluation with a diagnostic 131I-WBS (per-
Furthermore, it should be considered that higher cumulative formed by administering 185 MBq 131I-iodide) is useless in
activities are presumably of little benefit because of the pro- patients with known metastases, since the results of this
gressive loss of iodine-trapping property of metastatic evaluation will not alter the therapeutic strategy and, because
a b
Fig. 27.16 Post-therapeutic 131I-WBS images obtained in two separate lar lesion, but development of macronodular lesions in the right lung
sequential occasions after administration of 131I-iodide in a patient with (serum Tg 1918 ng/mL under rhTSH stimulation) (from Volterrani D,
pulmonary metastases from DTC. (a) September 2006: pattern of dif- Erba PA, Mariani G. Fondamenti di Medicina Nucleare – Tecniche e
fuse metastatic involvement associated with a macronodular pattern, Applicazioni. Milan: Springer, 2010, with permission of Springer
more obvious in the left parahilar region (serum Tg 251 ng/mL under Science + Business Media)
rhTSH stimulation). (b) July 2007: partial regression of the left parahi-
of the possible “stunning” effect, may actually interfere Furthermore, one of the main advantages perceived with
with uptake of the subsequent therapeutic activity of the use of 124I-iodide PET/CT—whenever this procedure is
131
I-iodide. Instead, a 131I-WBS performed 4–7 days after available—in patients with DTC candidate to therapy with
administration of a therapeutic activity of 131I-iodide may 131
I-iodide is the possibility of accurate quantification of
better characterize both the distribution and metabolic activ- radioiodide uptake as a basis for more accurate dosimetric
ity of metastases, including those that had gone undetected estimates than with the single-photon emitting radionuclides
prior to 131I-iodide therapy (Fig. 27.17). Alternatively, the [148, 149]. In fact, lesion dosimetry based on planar gamma
use of 123I-iodide can be envisaged in the pre-therapy phase camera or SPECT images may result in inaccurate estimates
(either for diagnostic/localization purposes or for dosimet- of the absorbed dose due to the difficulty of measuring the
ric estimates), since this radionuclide does not induce a target lesion volumes (especially for small lesions) and of
stunning effect. reliably measuring radioiodine uptake using 131I-iodide as
a b
Fig. 27.17 Post-therapeutic 131I-WBS images obtained in a 53-year- was unrecognized before therapy and was undetectable on diagnostic
old patient recently submitted to thyroidectomy because of PTC (fol- CT (b, transaxial sections at different levels). SPECT/CT (c) better
licular variant) infiltrating the thyroid capsule. Ultrasound examination characterized the pulmonary areas that, despite not being associated
prior to remnant ablation with 131I-iodide revealed metastasis in cervical with abnormalities on the CT component, were the site of lung metas-
lymph nodes; 4.8 GBq (130 mCi) of radioiodide was administered upon tasis from DTC (reproduced with permission from: Strauss HW,
rhTSH stimulation, with ensuing 219 ng/mL peak Tg level. (a) Mariani G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
Radioiodide uptake in the metastatic cervical lymph nodes, associated Pathophysiology to Clinical Applications, 2nd Edition. New York:
with a pattern of diffuse uptake in the lungs indicating metastasis that Springer; 2017)
the tracer (due to low spatial resolution and imaging sensitiv- bining imaging scans at 4, 24, 48, 72, and 96 h post-
ity). The feasibility of calculating the mean absorbed dose in administration (more rarely until 144 h) [148, 149]. The
patients using 124I-iodide PET images has initially been 5-point protocol is certainly the most accurate approach,
described in 1999 [150], after reports on preliminary quanti- although even a simplified 2-point protocol (i.e., 24 and
tative phantom studies. 96 h) can be employed for determining the lesion absorbed
Accurate lesion quantification for dosimetry estimates is dose per administered 131I-iodide activity (LDpA) within a
possible using the approach of recovery coefficients (RCs), a 20% uncertainty compared to the more complex protocol
method that allows to correct the activity concentration in the [152].
lesions imaged for partial-volume effect and for the prompt As a quantitative tomographic imaging modality, PET/
gamma coincidence effect due to the complex decay scheme CT with 124I-iodide is especially useful to calculate the
of 124I [151]. Several protocols for dosimetry estimates have mean dose for normal organs using a three-dimensional
used serial PET measurements from 3 to 5 time points, com- internal dosimetry software, the results of which are com-
parable with those obtained with OLINDA dosimetry 27.6.8.1 A dvanced DTC Refractory
[153]. Widely variable values have been reported for the to Radioiodide Treatment
mean absorbed dose estimated with 124I-iodide PET, rang- About 15% of DTCs become radioiodide refractory after
ing 0.4–49 Gy/Bq (absorbed dose distribution from 0.3 to a several courses of 131I-iodide treatments because of tumor
maximum of 4000 Gy). This is probably the result of the dedifferentiation. In these patients, other therapeutic strate-
wide range in the radioiodine effective half-life values gies must be employed. Among these, the approaches most
observed in recurrent/metastatic lesions from DTC, that is, commonly adopted in the past attempted at inducing “redif-
between 6 and 80 h in different reports [146, 152, 153]. As ferentiation” of the “dedifferentiated” TC cells.
per today’s knowledge, 124I-iodide PET dosimetry certainly When DTCs are metastatic, non-iodine-avid, and rapidly
constitutes an important tool in pre-therapy evaluation of growing, the effectiveness of conventional therapies such as
patients with DTC [154], to be possibly used in the clinical EBRT and chemotherapy is very low. The only chemotherapy
routine whenever available [155]. However, evidence- agent that is somewhat effective in patients with metastatic
based long-term data on the clinical impact of such accu- DTC is doxorubicin [157]. Although a 33% response rate has
rate dosimetric calculations have not yet been published. been reported with doxorubicin as a single agent, this observa-
When radioiodide uptake is no longer demonstrable on tion has not been confirmed by subsequent studies, in which
post-therapeutic WBS, treatment with 131I-iodide should be response rates ranged from 0% to 22% (all responses being
discontinued. EBRT and local treatments (thermoablation, partial and lasting only few months, without any survival ben-
laser-therapy, etc.) may be considered as the first approach efit) [158, 159]. In two studies where doxorubicin was com-
when radioiodide is no more effective and surgical removal of bined with cisplatin, the response rates were similar as those
metastases is not feasible. However, since the growth rate of observed with doxorubicin alone but at the expense of major
DTC metastases is usually slow, a “wait-and-see” approach toxicity [160, 161]. Although an increased response rate has
with 6-monthly controls seems warranted in this setting [21, been observed in patients treated with cisplatin plus epirubicin
53]. Instead, if the lesions are rapidly growing and fulfilling, after TSH stimulation relative to non- stimulated patients
the RECIST criteria, a targeted therapy, can be initiated [156]. treated with the same drugs [162], these promising results
should be confirmed in larger series. With the introduction of
newer targeted therapies based on tyrosine-kinase inhibitors
Key Learning Points (TKIs), the use of conventional chemotherapeutic agents is
• Detection of local or regional recurrence dictates strictly limited to those few cases that cannot be treated with
complete work-up to evaluate its extent and to TKIs [2].
search for distant metastases. The possibility to revert the ability to take up iodine in
• The 131I-iodide activity to be administered can be radioiodine-refractory DTC has been recurrently
based on dosimetric estimates aiming at releasing explored. Treatment with retinoic acid has proven effec-
the maximal safe dose to the patient, yet with a radi- tive in redifferentiating neoplastic cell clones in patients
ation dose >85 Gy to the lymph node(s), as well as with acute promyelocytic leukemia, thus making them
to metastatic lesions at other sites. susceptible to chemotherapy. Several studies have been
• Alternatively, an empiric activity based on average performed which demonstrated the ability of retinoic
estimates and prior experience can be administered, acid to reinduce iodine uptake ability in DTC. Despite all
usually 3.7–5.5 GBq of 131I-iodide. these efforts, the final result is that retinoic acid can work
• Most distant metastases from DTC are located in but only on a relatively low percentage of dedifferenti-
the lung (50–60%) or bones (20–30%). ated DTCs and determining just a small control of tumor
• Conventional therapeutic strategies such as surgery, growth.
radioiodide, external beam radiation therapy, and Recently, selumetinib, a new drug able to block the activ-
systemic therapy with the new “targeted” therapies ity of the mitogen-activated protein kinase (MEK), has been
can all be used to treat metastatic DTC. demonstrated to be able to both reinduce the ability of taking
• In patients with distant metastases, pre-therapy up iodine in dedifferentiated DTC and determine an objec-
evaluation with a diagnostic 123I-WBS or PET/CT tive tumor size reduction as documented by CT imaging.
with 124I-iodide can be useful. Clinical trials on the effect of selumetinib on metastatic
radioiodine-refractory DTC are ongoing.
Thyroid Association and American Association of Clinical

Key Learning Points Endocrinologists. Thyroid. 2011;21:593–646.
16. Farrar JJ, Toft AD. Iodine-131 treatment of hyperthyroidism: cur-
• About 15% of DTCs become radioiodide refractory rent issues. Clin Endocrinol (Oxf). 1991;35:207–12.
after several courses of 131I-iodide treatments—due 17. Howarth D, Epstein M, Lan L, Tan P, Booker J. Determination of
to tumor dedifferentiation. the optimal minimum radioiodine dose in patients with Graves’ dis-
ease: a clinical outcome study. Eur J Nucl Med. 2001;28:1489–95.
• With the introduction of newer targeted therapies based 18. Orsini F, et al. Personalization of radioiodine treatment for
on tyrosine-kinase inhibitors (TKIs), the use of conven- Graves’ disease: a prospective, randomized study with a novel
tional chemotherapeutic agents is strictly limited to method for calculating the optimal 131I-iodide activity based on
those few cases that cannot be treated with TKIs. target reduction of thyroid mass. Q J Nucl Med Mol Imaging.
2012;56:496–502.
19. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA
Cancer J Clin. 2016;66:7–30.
20. Brierley JD, Gospodarowicz MK, Wittekind C. TNM classifica-
References tion of malignant tumours. 8th ed. Weinheim, Germany: Wiley;
2017. p. 69–71.
21. Cooper DS, Doherty GM, Haugen BR, et al. Revised American
1. Maenhaut C, Christophe D, Vassart G, et al. Ontogeny, anat-
Thyroid Association management guidelines for patients with
omy, metabolism and physiology of the thyroid. [Updated 2015
thyroid nodules and differentiated thyroid cancer. Thyroid.
Jul 15]. In: De Groot LJ, Chrousos G, Dungan K, et al, eds.
2009;19:1167–214.
Endotext [Internet]. South Dartmouth (MA): MDText.com, Inc.;
22. Tuttle RM, Tala H, Shah J, et al. Estimating risk of recurrence in
2000. Available from: https://www.ncbi.nlm.nih.gov/books/
differentiated thyroid cancer after total thyroidectomy and radio-
NBK285554/
active iodine remnant ablation: using response to therapy variables
2. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel
to modify the initial risk estimates predicted by the new American
SJ, Nikiforov YE, et al. 2015 American Thyroid Association
Thyroid Association staging system. Thyroid. 2010;20:1341–9.
Management Guidelines for adult patients with thyroid
23. Grimm D Current knowledge in thyroid cancer—from bench to
nodules and differentiated thyroid cancer: The American
bedside. Int J Mol Sci. 2017;18(7). pii: E1529. doi: https://doi.
Thyroid Association Guidelines Task Force on Thyroid
org/10.3390/ijms18071529
Nodules and Differentiated Thyroid Cancer. Thyroid.
24. Neff RL, Farrar WB, Kloos RT, Burman KD. Anaplastic thyroid
2016;26:1–133.
cancer. Endocrinol Metab Clin N Am. 2008;37:525–38.
3. Popoveniuc G, Jonklaas J. Thyroid nodules. Med Clin North Am.
25. Samaan NA, Ordonez NG. Uncommon types of thyroid cancer.
2012;96:329–49.
Endocrinol Metab Clin N Am. 1990;19:637–48.
4. Kravets I. Hyperthyroidism: diagnosis and treatment. Am Fam
26. Rusinek D, Chmielik E, Krajewska J, Jarzab M, Oczko-
Physician. 2016;93:363–70.
Wojciechowska M, Czarniecka A, Jarzab B. Current advances in
5. Carlè A, Bulow Pedersen I, Knudsen N, et al. Epidemiology of
thyroid cancer management. Are we ready for the epidemic rise of
subtypes of hyperthyroidism in Denmark. A population-based
diagnoses? Int J Mol Sci. 2017;18(8). pii: E1817. doi: https://doi.
study. Eur J Endocrinol. 2011;164:801–9.
org/10.3390/ijms18081817
6. Ferrari SM, Fallahi P, Antonelli A, Benvenga S. Environmental
27. Li Q, Lin X, Shao Y, Xiang F, Samir AE. Imaging and screening of
issues in thyroid diseases. Front Endocrinol. 2017;8:50. https://
thyroid cancer. Radiol Clin North Am. 2017;55:1261–71.
doi.org/10.3389/fendo.2017.00050. eCollection 2017.
28. Rago T, Vitti P, Chiovato L, et al. Role of conventional ultraso-
7. Menconi F, Marcocci C, Marinò M. Diagnosis and classification
nography and color flow-Doppler sonography in predicting malig-
of Graves’ disease. Autoimmun Rev. 2014;13:398–402.
nancy in “cold” thyroid nodules. Eur J Endocrinol. 1998;138:41–6.
8. Antonelli A, Ferrari SM, Corrado A, Di Domenicantonio A,
29. Frates MCM, Benson CBC, Charboneau JWJ, et al. Management
Fallahi P. Autoimmune thyroid disorders. Autoimmun Rev.
of thyroid nodules detected at US Society of Radiologists
2015;14:174–80.
in Ultrasound consensus conference statement. Radiology.
9. Caturegli P, De Remigis A, Rose NR. Hashimoto thyroiditis: clini-
2005;237:794–800.
cal and diagnostic criteria. Autoimmun Rev. 2014;13:391–7.
30. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cyto-
10. Shrestha RT, Hennessey J. Acute and subacute, and Riedel’s thy-
pathology. Am J Clin Pathol. 2009;132:658–65.
roiditis. [Updated 2015 Dec 8]. In: De Groot LJ, Chrousos G,
31. Eszlinger M, Lau L, Ghaznavi S, Symonds C, Chandarana
Dungan K, et al., editors. Endotext [Internet]. South Dartmouth
SP, Khalil M, Paschke R. Molecular profiling of thyroid nod-
(MA): MDText.com, Inc.; 2000. Available from: https://www.
ule fine- needle aspiration cytology. Nat Rev Endocrinol.
ncbi.nlm.nih.gov/books/NBK285553/
2017;13:415–24.
11. Dean DS. Thyroiditis. In: Baskin HJ, Duick DS, Levine RA, edi-
32. Lazar V, Bidart JM, Caillou B, Mahé C, Lacroix L, Filetti S,
tors. Thyroid ultrasound and ultrasound-guided FNA. 2nd ed.
Schlumberger M. Expression of the Na+/I− symporter gene in
Boston, MA: Springer US; 2008. p. 63–75.
human thyroid tumors: a comparison study with other thyroid-
12. Garberoglio S, Testori O. Role of nuclear medicine in the diagno-
specific genes. J Clin Endocrinol Metab. 1999;84:3228–34.
sis of benign thyroid diseases. Front Horm Res. 2016;45:24–36.
33. Giovanella S, Suriano S, Maffioli M, Ceriani M. 18FDG-positron
13. Metso S, Jaatinen P, Huhtala H, Luukkaala T, Oksala H, Salmi
emission tomography/computed tomography (PET/CT) scanning
J. Long-term follow-up study of radioiodine treatment of hyper-
in thyroid nodules with nondiagnostic cytology. Clin Endocrinol.
thyroidism. Clin Endocrinol (Oxf). 2004;61:641–8.
2011;74:644–8.
14. Sridama V, Mccormick M, Kaplan EL, Fauchet R, Degroot
34. Piccardo A, Puntoni M, Treglia G, Foppiani L, Bertagna F, Paparo
LJ. Long-term follow-up study of compensated low-dose 131I ther-
F, et al. Thyroid nodules with indeterminate cytology: prospec-
apy for Graves’ disease. N Engl J Med. 1984;311:426–32.
tive comparison between 18F-FDG-PET/CT, multiparametric neck
15. Bahn RS, Burch HB, Cooper DS, et al. Hyperthyroidism and other
ultrasonography, 99mTc-MIBI scintigraphy and histology. Eur J
causes of thyrotoxicosis: management guidelines of the American
Endocrinol. 2016;174:693–703.
35. Nayan S, Ramakrishna J, Gupta MK. The proportion of malig- improved survival for patients with intermediate-risk papillary
nancy in incidental thyroid lesions on 18F-FDG PET study: a sys- thyroid cancer. J Clin Endocrinol Metab. 2015;100:1529–36.
tematic review and meta-analysis. Otolaryngol Head Neck Surg. 54. Elisei R, Agate L, Viola D, Matrone A, Biagini A, Molinaro
2014;151:190–200. E. How to manage patients with differentiated thyroid cancer and
36. Agrawal K, Weaver J, Ul-Hassan F, Jeannon JP, Simo R, Carroll P, a rising serum thyroglobulin level. Endocrinol Metab Clin North
et al. Incidence and significance of incidental focal thyroid uptake Am. 2014;43:331–44.
on 18F-FDG PET study in a large patient cohort: retrospective 55. Wang LY, Roman BR, Palmer FL, et al. Effectiveness of routine
single-centre experience in the United Kingdom. Eur Thyroid J. ultrasonographic surveillance of patients with low-risk papillary
2015;4:115–22. carcinoma of the thyroid. Surgery. 2016;159:1390–5.
37. Palaniswamy SS, Subramanyam P. Diagnostic utility of PETCT in 56. Remy H, Borget I, Leboulleux S, et al. 131I effective half-
thyroid malignancies: an update. Ann Nucl Med. 2013;27:681–93. life and dosimetry in thyroid cancer patients. J Nucl Med.
38. Chiacchio S, Lorenzoni A, Boni G, Rubello D, Elisei R, Mariani 2008;49:1445–50.
G. Anaplastic thyroid cancer: prevalence, diagnosis and treatment. 57. Mazzaferri EL, Kloos RT. Clinical review 128: current approaches
Minerva Endocrinol. 2008;33:341–57. to primary therapy for papillary and follicular thyroid cancer. J
39. Bodet-Milin C, Faivre-Chauvet A, Carlier T, Rauscher A, Clin Endocrinol Metab. 2001;86:1447–63.
Bourgeois M, Cerato E, et al. Immuno-PET using anticarcinoem- 58. Sgouros G, Song H, Ladenson PW, Wahl RL. Lung toxicity in
bryonic antigen bispecific antibody and 68Ga-labeled peptide in radioiodine therapy of thyroid carcinoma: development of a dose-
metastatic medullary thyroid carcinoma: clinical optimization of rate method and dosimetric implications of the 80-mCi rule. J
the pretargeting parameters in a first-in-human trial. J Nucl Med. Nucl Med. 2006;47:1977–84.
2016;57:1505–11. 59. Bolch WE, Bouchet LG, Robertson JS, et al. MIRD pamphlet
40. Wells SA, Asa SL, Dralle H, Elisei R, Evans DB, Gagel RF, et al. No. 17: the dosimetry of nonuniform activity distributions—
Revised American Thyroid Association guidelines for the manage- radionuclide S values at the voxel level. J Nucl Med.
ment of medullary thyroid carcinoma. Thyroid. 2015;25:567–610. 1999;40(Suppl):11S–36S.
41. Bilimoria KY, Bentrem DJ, Ko CY, Stewart AK, Winchester DP, 60. Pacini F, Ladenson PW, Schlumberger M, et al. Radioiodine
Talamonti MS, Sturgeon C. Extent of surgery affects survival for ablation of thyroid remnants after preparation with recombinant
papillary thyroid cancer. Ann Surg. 2007;246:375–81. human thyrotropin in differentiated thyroid carcinoma: results of
42. Leboulleux S, Girard E, Rose M, et al. Ultrasound criteria of an international, randomized, controlled study. J Clin Endocrinol
malignancy for cervical lymph nodes in patients followed up Metab. 2006;91:926–32.
for differentiated thyroid cancer. J Clin Endocrinol Metab. 61. Schlumberger M, Catargi B, Borget I, et al. Strategies of radio-
2007;92:3590–4. iodine ablation in patients with low-risk thyroid cancer. N Engl J
43. Sywak M, Cornford L, Roach P, Stalberg P, Sidhu S, Delbridge Med. 2012;366:1663–73.
L. Routine ipsilateral level VI lymphadenectomy reduces postop- 62. Mallick U, Harmer C, Yap B, et al. Ablation with low-dose radio-
erative thyroglobulin levels in papillary thyroid cancer. Surgery. iodine and thyrotropin alfa in thyroid cancer. N Engl J Med.
2006;140:1000–5. 2012;366:1674–85.
44. Palazzo FF, Gosnell J, Savio R, Reeve TS, Sidhu SB, Sywak MS, 63. Silberstein EB, Alavi A, Balon HR, et al. Procedure guideline
et al. Lymphadenectomy for papillary cancer: changes in practice for therapy of thyroid disease with iodine-131. J Nucl Med.
over four decades. Eur J Surg Oncol. 2006;32:340–4. 2002;43:856–61.
45. Carty SE, Cooper DS, Doherty GM, Duh QY, Kloos RT, Mandel 64. Park JT II, Hennessey JV. Two-week low iodine diet is necessary
SJ, et al. on behalf of ATA, AAES, AAOHNS, AHNS. Consensus for adequate outpatient preparation for radioiodine rhTSH scan-
statement on the terminology and classification of central neck ning in patients taking levothyroxine. Thyroid. 2004;14:57–63.
dissection for thyroid cancer. Thyroid 2009;19:1153–1158 65. Regalbuto C, Gullo D, Vigneri R, Pezzino V. Measurement of
46. DeGroot LJ, Kaplan EL, McCormick M, Straus FH. Natural his- iodine before 131I in thyroid cancer. Lancet. 1994;344:1501–2.
tory, treatment, and course of papillary thyroid carcinoma. J Clin 66. Rawson RW, Rall JE, Peacock W. Limitations in the treatment
Endocrinol Metab. 1990;71:414–24. of cancer of the thyroid with radioactive iodine. Trans Assoc Am
47. Maxon HR III, Englaro EE, et al. Radioiodine-131 therapy for Phys. 1951;64:179–98.
well-differentiated thyroid cancer – a quantitative radiation dosi- 67. Bajén MT, Mañé S, Muñoz A, Ramòn GJ. Effect of diagnostic
metric approach: outcome and validation in 85 patients. J Nucl dose of 185 MBq 131I on postsurgical thyroid remnants. J Nucl
Med. 1992;33:1132–6. Med. 2000;41:2038–42.
48. Mazzaferri EL, Jhiang SM. Long-term impact of initial surgical 68. McDougall IR, Iagaru A. Thyroid stunning: fact or fiction? Semin
and medical therapy on papillary and follicular thyroid cancer. Am Nucl Med. 2011;41:105–12.
J Med. 1994;97:418–28. 69. Park HM. Stunned thyroid after high dose 131I imaging. Clin Nucl
49. Tubiana M, Schlumberger M, Rougier P, et al. Long-term results Med. 1992;17:501–2.
and prognostic factors in patients with differentiated thyroid car- 70. Sisson JC, Avram AM, Lawson SA, Gauger PG, Doherty
cinoma. Cancer. 1985;55:794–804. GM. The so-called stunning of thyroid tissue. J Nucl Med.
50. Baudin E, Travagli JP, Ropers J, et al. Microcarcinoma of the 2006;47:1406–12.
thyroid gland: the Gustave-Roussy Institute experience. Cancer. 71. Park HM, Perkins OW, Edmondson JW, Schnute RB, Manatunga
1998;83:553–9. A. Influence of diagnostic radioiodines on the uptake of ablative
51. Simpson WJ, Panzarella T, Carruthers JS, Gospodarowicz MK, dose of iodine-131. Thyroid. 1994;4:49–54.
Sutcliffe SB. Papillary and follicular thyroid cancer: impact 72. Muratet JP, Giraud P, Daver A, Minier JF, Gamelin E, Larra
of treatment in 1578 patients. Int J Radiat Oncol Biol Phys. F. Predicting the efficacy of first iodine-131 treatment in differen-
1988;14:1063–75. tiated thyroid carcinoma. J Nucl Med. 1997;38:1362–8.
52. Grebe SK, Hay ID. Follicular cell-derived thyroid carcinomas. 73. Nordén MM, Larsson F, Tedelind S, Carlsson T, Lundh C,
Cancer Treat Res. 1997;89:91–140. Forssell-Aronsson E, Nilsson M. Down-regulation of the sodium/
53. Ruel E, Thomas S, Dinan M, Perkins JM, Roman SA, Sosa iodide symporter explains131I-induced thyroid stunning. Cancer
JA. Adjuvant radioactive iodine therapy is associated with Res. 2007;6:7512–7.
74. Medvedec M. Thyroid stunning in vivo and in vitro. Nucl Med 90. Avram AM. Radioiodine scintigraphy with SPECT/CT: an impor-
Commun. 2005;26:731–5. tant diagnostic tool for thyroid cancer staging and risk stratifica-
75. Caudill CM, Zhu Z, Ciampi R, Stringer JR, Nikiforov YE. Dose- tion. J Nucl Med Technol. 2014;42:170–80.
dependent generation of RET/PTC in human thyroid cells after 91. Al Balooshi B, Vinjamuri S. Should all patients with differentiated
in vitro exposure to gamma-radiation: a model of carcinogenic thyroid carcinoma undergo 131I SPECT-CT scanning rather than
chromosomal rearrangement induced by ionizing radiation. J Clin 131
I whole-body scanning? Nucl Med Commun. 2015;36:540–52.
Endocrinol Metab. 2005;90:2364–9. 92. Allweiss P, Braunstein GD, Katz A, Waxman A. Sialadenitis fol-
76. de Geus-Oei LF, Oei HY, Hennemann G, Krenning EP. Sensitivity lowing I-131 therapy for thyroid carcinoma: concise communica-
of 123I whole-body scan and thyroglobulin in the detection of tion. J Nucl Med. 1984;25:755–8.
metastases or recurrent differentiated thyroid cancer. Eur J Nucl 93. Alexander C, Bader JB, Schaefer A, Finke C, Kirsch
Med Mol Imaging. 2002;29:768–74. CM. Intermediate and long-term side effects of high-dose radioio-
77. Sherman SI, Tielens ET, Sostre S, Wharam MD Jr, Ladenson dinetherapy for thyroid carcinoma. J Nucl Med. 1998;39:1551–4.
PW. Clinical utility of posttreatment radioiodine scans in the man- 94. de Vathaire F, Schlumberger M, Delisle MJ, Francese C, Challeton
agement of patients with thyroid carcinoma. J Clin Endocrinol C, de la Genardière E, et al. Leukaemias and cancers follow-
Metab. 1994;78:629–34. ing iodine-131 administration for thyroid cancer. Br J Cancer.
78. Fatourechi V, Hay ID, Mullan BP, Wiseman GA, Eghbali- 1997;75:734–9.
Fatourechi GZ, Thorson LM, Gorman CA. Are posttherapy 95. Verkooijen RB, Smit JW, Romijn JA, Stokkel M. The incidence
radioiodine scans informative and do they influence subsequent of second primary tumors in thyroid cancer patients is increased,
therapy of patients with differentiated thyroid cancer? Thyroid. but not related to treatment of thyroid cancer. Eur J Endocrinol.
2000;10:573–7. 2006;155:801–6.
79. Chen L, Luo Q, Shen Y, Yu Y, Yuan Z, Lu H, Zhu R. Incremental 96. Subramanian S, Goldstein DP, et al. Second primary malignancy
value of 131I SPECT/CT in the management of patients with dif- risk in thyroid cancer survivors: a systematic review and meta-
ferentiated thyroid carcinoma. J Nucl Med. 2008;49:1952–7. analysis. Thyroid. 2007;17:1277–88.
80. Wong KK, Zarzhevsky N, Cahill JM, Frey KA, Avram 97. Garsi J-P, Schlumberger M, Rubino C, et al. Therapeutic admin-
AM. Incremental value of diagnostic 131I SPECT/CT fusion imag- istration of 131I for differentiated thyroid cancer: radiation dose to
ing in the evaluation of differentiated thyroid carcinoma. AJR Am ovaries and outcome of pregnancies. J Nucl Med. 2008;49:845–52.
J Roentgenol. 2008;191:1785–94. 98. Wichers M, Benz E, Palmedo H, Biersack HJ, Grünwald F,
81. Aide N, Heutte N, Rame JP, Rousseau E, Loiseau C, Henry- Klingmüller D. Testicular function after radioiodine therapy for
Amar M, Bardet S. Clinical relevance of single-photon emission thyroid carcinoma. Eur J Nucl Med. 2000;27:503–7.
computed tomography/computed tomography of the neck and 99. Rosário PW, Barroso AL, Rezende LL, Padro EL, Borges
thorax in postablation 131I scintigraphy for thyroid cancer. J Clin MA, Guimares VC, Purisch S. Testicular function after
Endocrinol Metab. 2009;94:2075–84. radioiodine therapy in patients with thyroid cancer. Thyroid.
82. Schmidt D, Szikszai A, Linke R, Bautz W, Kuwert T. Impact 2006;16:667–70.
of 131I SPECT/spiral CT on nodal staging of differentiated 100. Hebestreit H, Biko J, Drozd V, Demidchik Y, Burkhardt A, Trusen
thyroid carcinoma at the first radioablation. J Nucl Med. A, et al. Pulmonary fibrosis in youth treated with radioiodine for
2009;50:18–23. juvenile thyroid cancer and lung metastases after Chernobyl. Eur
83. Spanu A, Solinas ME, Chessa F, Sanna D, Nuvoli S, Madeddu J Nucl Med Mol Imaging. 2011;38:1683–90.
G. 131I SPECT/CT in the follow-up of differentiated thyroid car- 101. Reiners C, Biko J, Haenscheid H, Hebestreit H, Kirinjuk S,
cinoma: incremental value versus planar imaging. J Nucl Med. Baranowski O, et al. Twenty-five years after Chernobyl: outcome
2009;50:184–90. of radioiodine treatment in children and adolescents with very
84. Kohlfuerst S, Igerc I, Lobnig M, et al. Posttherapeutic high-risk radiation-induced differentiated thyroid carcinoma. J
131
ISPECT-CT offers high diagnostic accuracy when the findings Clin Endocrinol Metab. 2013;98:3039–48.
on conventional planar imaging are inconclusive and allows a tai- 102. Pilli T, Brianzoni E, Capoccetti F, et al. A comparison of 1850 (50
lored patient treatment regimen. Eur J Nucl Med Mol Imaging. mCi) and 3700 MBq (100 mCi) 131-iodine administered doses for
2009;36:886–93. recombinant thyrotropin-stimulated postoperative thyroid rem-
85. Akram K, Parker JA, Donohoe K, Kolodny G. Role of single nant ablation in differentiated thyroid cancer. J Clin Endocrinol
photon emission computed tomography/computed tomogra- Metab. 2007;92:3542–6.
phy in localization of ectopic parathyroid adenoma: a pictorial 103. Santini F, Pinchera A, Marsili A, et al. Lean body mass is a major
case series and review of the current literature. Clin Nucl Med. determinant of levothyroxine dosage in the treatment of thyroid
2009;34:500–2. diseases. J Clin Endocrinol Metab. 2005;90:124–7.
86. Barwick T, Murray I, Megadmi H, et al. Single photon emission 104. Sawin CT, Geller A, Hershman JM, Castelli W, Bacharach P. The
computed tomography (SPECT)/computed tomography using aging thyroid. The use of thyroid hormone in older persons.
Iodine-123 in patients with differentiated thyroid cancer: addi- JAMA. 1989;261:2653–5.
tional value over whole body planar imaging and SPECT. Eur J 105. Bartalena L, Martino E, Pacchiarotti A, et al. Factors affecting sup-
Endocrinol. 2010;162:1131–9. pression of endogenous thyrotropin secretion by thyroxine treat-
87. Mustafa M, Kuwert T, Weber K, et al. Regional lymph node ment: retrospective analysis in athyreotic and goitrous patients. J
involvement in T1 papillary thyroid carcinoma: a bicentric Clin Endocrinol Metab. 1987;64:849–55.
prospective SPECT/CT study. Eur J Nucl Med Mol Imaging. 106. Marcocci C, Golia F, Bruno-Bossio G, Vignali E, Pinchera
2010;37:1462–6. A. Carefully monitored levothyroxine suppressive therapy is
88. Grewal RK, Tuttle RM, Fox J, et al. The effect of posttherapy 131I not associated with bone loss in premenopausal women. J Clin
SPECT/CT on risk classification and management of patients with Endocrinol Metab. 1994;78:818–23.
differentiated thyroid cancer. J Nucl Med. 2010;51:1361–7. 107. Cooper DS, Specker B, Ho M, et al. Thyrotropin suppression and
89. Wong KK, Sisson JC, Koral KF, Frey KA, Avram AM. Staging disease progression in patients with differentiated thyroid cancer:
of differentiated thyroid carcinoma using diagnostic 131I SPECT/ results from the National Thyroid Cancer Treatment Cooperative
CT. AJR Am J Roentgenol. 2010;195:730–6. Registry. Thyroid. 1998;8:737–44.
108. Mazzaferri EL, Jhiang SM. Differentiated thyroid cancer long- 125. Pacini F. Follow-up of differentiated thyroid cancer. Eur J Nucl
term impact of initial therapy. Trans Am Clin Climatol Assoc. Med Mol Imaging. 2002;29(Suppl 2):492–4966.
1995;106:151–68. 126. Chiovato L, Latrofa F, Braverman LE, et al. Disappearance of
109. Schlumberger M. Can iodine-131 whole-body scan be replaced humoral thyroid autoimmunity after complete removal of thyroid
by thyroglobulin measurement in the post-surgical follow-up of antigens. Ann Intern Med. 2003;139:346–51.
differentiated thyroid carcinoma? J Nucl Med. 1992;33:172–3. 127. Schlumberger M, Challeton C, De Vathaire F, et al. Radioactive
110. Mazzaferri EL. Changing paradigms in the follow-up of patients iodine treatment and external radiotherapy for lung and bone metas-
with differentiated thyroid cancer: an alternative to [18F]fluoro- tases from thyroid carcinoma. J Nucl Med. 1996;37:598–605.
deoxyglucose positron emission tomographic scanning. Endocr 128. Urken ML, Milas M, Randolph GW, et al. Management of recur-
Pract. 2003;9:324–6. rent and persistent metastatic lymph nodes in well-differentiated
111. Van Herle AJ. Serum thyroglobulin levels in patients with differ- thyroid cancer: a multifactorial decision-making guide for the
entiated thyroid carcinoma. Ann Radiol (Paris). 1977;20:743–5. Thyroid Cancer Care Collaborative. Head Neck. 2015;37:605–14.
112. Spencer CA, Lopresti JS. Measuring thyroglobulin and thyroglob- 129. Ceccarelli C, Pacini F, Lippi F, Elisei R, Arganini M, Miccoli P,
ulin autoantibody in patients with differentiated thyroid cancer. Pinchera A. Thyroid cancer in children and adolescents. Surgery.
Nat Clin Pract Endocrinol Metab. 2008;4:223–33. 1988;104:1143–8.
113. Pacini F, Mariotti S, Formica N, Elisei R, Anelli S, Capotorti 130. Vassilopoulou-Sellin R, Klein MJ, Smith TH, et al. Pulmonary
E, Pinchera A. Thyroid autoantibodies in thyroid cancer: inci- metastases in children and young adults with differentiated thy-
dence and relationship with tumour outcome. Acta Endocrinol. roid cancer. Cancer. 1993;71:1348–52.
1988;119:373–80. 131. Piekarski JD, Schlumberger M, Leclere J, Couanet D, Masselot J,
114. Pacini F, Molinaro E, Lippi F, et al. Prediction of disease status by Parmentier C. Chest computed tomography (CT) in patients with
recombinant human TSH-stimulated serum Tg in the postsurgical micronodular lung metastases of differentiated thyroid carcinoma.
follow-up of differentiated thyroid carcinoma. J Clin Endocrinol Int J Radiat Oncol Biol Phys. 1985;11:1023–7.
Metab. 2001;86:5686–90. 132. Schlumberger M, Arcangioli O, Piekarski JD, Tubiana M,
115. Mazzaferri EL, Robbins RJ, Spencer CA, et al. A consensus report Parmentier C. Detection and treatment of lung metastases of
of the role of serum thyroglobulin as a monitoring method for low- differentiated thyroid carcinoma in patients with normal chest
risk patients with papillary thyroid carcinoma. J Clin Endocrinol X-rays. J Nucl Med. 1988;29:1790–4.
Metab. 2003;88:1433–41. 133. Sumimura J, Nakagawa K, Kawamura J, Tayama M, Takahashi E,
116. Schlumberger M, Hitzel A, Toubert ME, et al. Comparison of Moritomo T, Miyata M. Thyroid cancer metastasis to the lumbar
seven serum thyroglobulin assays in the follow-up of papillary spine successfully treated by embolization and radioiodine. A case
and follicular thyroid cancer patients. J Clin Endocrinol Metab. report. Nippon Geka Gakkai Zasshi. 1990;91:910–3.
2007;92:2487–95. 134. Court C, Noun Z, Gagey O, Nordin JY. Surgical treatment of
117. Chindris AM, Diehl NN, Crook JE, Fatourechi V, Smallridge metastases from thyroid cancer in the axial skeleton. A retrospec-
RC. Undetectable sensitive serum thyroglobulin (<0.1 ng/ml) in tive study of 18 cases. Acta Orthop Belg. 2000;66:345–52.
163 patients with follicular cell-derived thyroid cancer: results 135. Smit JW, Links TP, Hew JM, Goslings BM, Vielvoye GJ,
of rhTSH stimulation and neck ultrasonography and long-term Vermey A. Embolization of skeletal metastases in patients
biochemical and clinical follow-up. J Clin Endocrinol Metab. with differentiated thyroid carcinoma. Ned Tijdschr Geneeskd.
2012;97:2714–23. 2000;144:1406–10.
118. Pacini F, Capezzone M, Elisei R, Ceccarelli C, Taddei D, Pinchera 136. Van Tol KM, Hew JM, Jager PL, Vermey A, Dullaart RP, Links
A. Diagnostic 131-iodine whole-body scan may be avoided TP. Embolization in combination with radioiodine therapy for
in thyroid cancer patients who have undetectable stimulated bone metastases from differentiated thyroid carcinoma. Clin
serum Tg levels after initial treatment. J Clin Endocrinol Metab. Endocrinol (Oxf). 2000;52:653–9.
2002;87:1499–501. 137. van Tol KM, Hew JM, Links TP. Images in thyroidology.
119. Gerard SK, Cavalieri RR. I-123 diagnostic thyroid tumor whole- Embolization of a bone metastasis of follicular thyroid carcinoma.
body scanning with imaging at 6, 24, and 48 hours. Clin Nucl Thyroid. 2000;10:621–2.
Med. 2002;27:1–8. 138. Klain M, Ricard M, Leboulleux S, Baudin E, Schlumberger
120. Yan W, Roach PJ, Bautovich GJ, Learoyd DL, Robinson M. Radioiodine therapy for papillary and follicular thyroid carci-
BG. Timing of iodine-123 scintigraphy following use of recombi- noma. Eur J Nucl Med Mol Imaging. 2002;29:479–85.
nant human thyrotropin in differentiated thyroid carcinoma. Clin 139. Franzius C, Dietlein M, Biermann M, et al. Procedure guide-
Nucl Med. 2007;32:375–7. line for radioiodine therapy and 131iodine whole-body scintig-
121. Pacini F, Lippi F, Formica N, Elisei R, Anelli S, Ceccarelli C, raphy in paediatric patients with differentiated thyroid cancer.
Pinchera A. Therapeutic doses of iodine-131 reveal undiagnosed Nuklearmedizin. 2007;46:224–2231.
metastases in thyroid cancer patients with detectable serum thyro- 140. Jarzab B, Handkiewicz-Junak D, Wloch J. Juvenile differentiated
globulin levels. J Nucl Med. 1987;28:1888–91. thyroid carcinoma and the role of radioiodine in its treatment: a
122. Pineda JD, Lee T, Ain K, Reynolds JC, Robbins J. Iodine-131 qualitative review. Endocr Relat Cancer. 2005;12:773–803.
therapy for thyroid cancer patients with elevated thyroglobu- 141. Van Nostrand D, Wartofsky L. Radioiodine in the treatment of
lin and negative diagnostic scan. J Clin Endocrinol Metab. thyroid cancer. Endocrinol Metab Clin N Am. 2007;36:807–22.
1995;80:1488–92. 142. Benua RS, Leeper RD. A method and rationale for treating meta-
123. Elisei R, Pinchera A, Romei C, et al. Expression of thyrotropin static thyroid carcinoma with the largest safe dose of I-131. In:
receptor (TSH-R), thyroglobulin, thyroperoxidase, and calcitonin Medeiros-Neto G, Gaitan G, editors. Frontiers in thyroidology.
messenger ribonucleic acids in thyroid carcinomas: evidence of New York: Plenum; 1986. p. 1317–21.
TSH-R gene transcript in medullary histotype. J Clin Endocrinol 143. Traino AC, Ferrari M, Cremonesi M, Stabin MG. Influence of total-
Metab. 1994;78:867–71. body mass on the scaling of S-factors for patient-specific, blood-
124. Arturi F, Russo D, Schlumberger M, du Villard JA, Caillou B, based red-marrow dosimetry. Phys Med Biol. 2007;52:5231–48.
Vigneri P, Wicker R, et al. Iodide symporter gene expression in 144. Durante C, Haddy N, Baudin E, et al. Long-term outcome of 444
human thyroid tumors. J Clin Endocrinol Metab. 1998;83:2493–6. patients with distant metastases from papillary and follicular thy-
roid carcinoma: benefits and limits of radioiodine therapy. J Clin and 3-dimensional internal dosimetry software. J Nucl Med.
Endocrinol Metab. 2006;91:2892. 2007;48:143–9.
145. Lippi F, Capezzone M, Angelini F, Taddei D, Molinaro E, 154. Lassmann M, Hänscheid H, Verburg FA, Luster M. The use of
Pinchera A, Pacini F. Radioiodine treatment of metastatic differ- dosimetry in the treatment of differentiated thyroid cancer. Q J
entiated thyroid cancer in patients on l-thyroxine, using recombi- Nucl Med Mol Imaging. 2011;55:107–15.
nant human TSH. Eur J Endocrinol. 2001;144:5–11. 155. Jentzen W, Verschure F, van Zon A, van de Kolk R, Wierts
146. Jarzab B, Handkiewicz-Junak D, Roskosz J, et al. Recombinant R, Schmitz J, et al. Response assessment of bone metastases
human TSH-aided radioiodine treatment of advanced differenti- from differentiated thyroid cancer patients in the initial radio-
ated thyroid carcinoma: a single-centre study of 54 patients. Eur J iodine treatment using iodine-124 PET imaging. J Nucl Med.
Nucl Med Mol Imaging. 2003;30:1077–86. 2016;57:1499–504.
147. Luster M, Lippi F, Jarzab B, Perros P, Lassmann M, Reiners C, 156. Schlumberger M, Brose M, Elisei R, et al. Definition and manage-
Pacini F. rhTSH-aided radioiodine ablation and treatment of dif- ment of radioactive iodine-refractory differentiated thyroid can-
ferentiated thyroid carcinoma: a comprehensive review. Endocr cer. Lancet Diabetes Endocrinol. 2014;2:356–8.
Relat Cancer. 2005;12:49–64. 157. Samaan NA, Schultz PN, Hickey RC, Goepfert H, Haynie TP,
148. Eschmann SM, Reischl G, Bilger K, et al. Evaluation of dosimetry Johnston DA, Ordonez NG. The results of various modalities
of radioiodine therapy in benign and malignant thyroid disorders of treatment of well differentiated thyroid carcinomas: a ret-
by means of iodine-124 and PET. Eur J Nucl Med. 2002;29:760–7. rospective review of 1599 patients. J Clin Endocrinol Metab.
149. Freudenberg LS, Jentzen W, Görges R, Petrich T, Marlowe RJ, 1992;75:714–20.
Knust J, Bockisch A. 124I-PET dosimetry in advanced differ- 158. Matuszczyk A, Petersenn S, Bockisch A, Gorges R, Sheu SY, Veit
entiated thyroid cancer: therapeutic impact. Nuklearmedizin. P, Mann K. Chemotherapy with doxorubicin in progressive med-
2007;46:121–8. ullary and thyroid carcinoma of the follicular epithelium. Horm
150. Erdi YE, Macapinlac H, Larson SM, Erdi AK, Yeung H, Furhang Metab Res. 2008;40:210–3.
EE, Humm JL. Radiation dose assessment for I-131 therapy of 159. Sherman SI. Cytotoxic chemotherapy for differentiated thyroid
thyroid cancer using I-124 PET imaging. Clin Positron Imaging. carcinoma. Clin Oncol (R Coll Radiol). 2010;22:464–8.
1999;2:41–6. 160. Shimaoka K, Schoenfeld DA, DeWys WD, Creech RH, DeConti
151. Jentzen W, Weise R, Kupferschläger J, et al. Iodine-124 PET R. A randomized trial of doxorubicin versus doxorubicin plus
dosimetry in differentiated thyroid cancer: recovery coefficient cisplatin in patients with advanced thyroid carcinoma. Cancer.
in 2D and 3D modes for PET/CT systems. Eur J Nucl Med Mol 1985;56:2155–60.
Imaging. 2008;35:611–23. 161. Williams SD, Birch R, Einhorn LH. Phase II evaluation

152. Jentzen W, Freudenberg L, Eising EG, Sonnenschein W, Knust of doxorubicin plus cisplatin in advanced thyroid cancer: a
J, Bockisch A. Optimized 124I PET dosimetry protocol for radio- Southeastern Cancer Study Group Trial. Cancer Treat Rep.
iodine therapy of differentiated thyroid cancer. J Nucl Med. 1986;70:405–7.
2008;49:1017–723. 162. Santini F, Bottici V, Elisei R, et al. Cytotoxic effects of carbo-
153. Kolbert KS, Pentlow KS, Pearson JR, Sheikh A, Finn RD, Humm platinum and epirubicin in the setting of an elevated serum thy-
JL, Larson SM. Prediction of absorbed dose to normal organs rotropin for advanced poorly differentiated thyroid cancer. J Clin
in thyroid cancer patients treated with 131I by use of 124IPET Endocrinol Metab. 2002;87:4160.
Hybrid Imaging in Non-thyroidal
Endocrinological Disorders 28
Duccio Volterrani, Federica Guidoccio,
Contents
28.1 Adrenal Glands 749
28.1.1 Radionuclide Imaging of the Adrenal Cortex 750
28.1.2 Radionuclide Imaging of the Adrenal Medulla 755
28.1.3 Radionuclide Imaging Protocols for Pheochromocytomas 755
28.2 Parathyroid Glands 757
28.2.1 Role of Preoperative Imaging 759
28.2.2 Radionuclide Parathyroid Imaging 759
28.2.3 Imaging of Parathyroid Carcinoma 763
References 764
Learning Objectives 28.1 Adrenal Glands

• Acquire the basic pathophysiologic and clinical notions
of medullary and cortical adrenal diseases. Adrenal glands are two organs of the endocrine system that
• Understand the methodologies and techniques for adrenal secrete different types of hormones. They are located in the
gland imaging with 131I norcholesterol and 123I/131I MIBG. retroperitoneal space, surrounded by a thick fibrous capsule,
• Understand the pathophysiology of hyperparathyroidisms placed immediately below the diaphragm, superiorly and
and the different methodologies used for parathyroid medially to the kidneys, between the 12th thoracic and the
imaging in patients with hyperparathyroidism. 1st lumbar vertebra. They have a cone shape with an inferior
• Understand the principles of image interpretation for base, a weight of about 8 g, and sizes of about 2–3 cm in
nuclear medicine imaging in medullary and cortical adre- width and 4–6 cm in length. They are richly vascularized,
nal diseases and hyperparathyroidism. innervated by the autonomous system, mainly through the
• Know when SPECT and SPECT/CT can be useful. adrenal plexus. Each adrenal consists of two distinct parts: a
peripheral part, called adrenal cortex, and a central part,
called adrenal medulla.
The cortical and the medullary components of the adre-
nals have completely different roles and a different embryo-
logical origin. The cortex originates from the embryonic
mesoderm, and in adults it constitutes about 3/4 of the adre-
nal mass. It consists of three histologically distinct areas: the
D. Volterrani (*) · F. Guidoccio · G. Mariani glomerular zone (more external), the fascicular zone, and the
Regional Center of Nuclear Medicine, Department of Translational reticular zone (more internal). The glomerular zone produces
mineralocorticoid hormones, mainly aldosterone—a hor-
e-mail: duccio.volterrani@med.unipi.it mone involved in the control of blood pressure, whose function

is to promote reabsorption of sodium and excretion of potas- circulating LDLs and are thus transported into the tissues
sium in the renal tubules. The fascicular zone (which consti- through a mechanism mediated by LDL receptors. These
tutes 75% of the cortex) produces glucocorticoids, essentially radiolabeled compounds are then stored within the intracellular
cortisol, mainly involved in glycemic control. In fact, corti- lipid droplet pool, without undergoing any other metabolic
sol increases blood glucose by enhancing hepatic gluconeo- pathways, as esters of long-chain fatty acids [1]. Adrenal uptake
genesis, stimulating glucagon secretion, and reducing the of cholesterol is regulated by two main control systems: the
activity of insulin receptors. The reticular zone produces the ACTH-cortisol axis and the renin- angiotensin-aldosterone
main adrenal androgens, DHEA, DHEA-S, and axis. Up to 50% of the cholesterol that enters the adrenals is
Δ-androstenedione, and to a lesser extent cortisol. modulated by ACTH, whereas the renin-angiotensin system
While cortisol production occurs in response to the pitu- modulates about 30% of such process. The remaining 20% is
itary hormone ACTH, modulated by a negative feedback sys- not under the control of any known hormonal system.
tem with cortisol itself, aldosterone is produced in response
to stimulation by the renin-angiotensin system. Renin pro- 28.1.1.1 Radiopharmaceuticals
duced by the renal dense macula converts the inactive pep- After the introduction of the first radiolabeled cholesterol
tide, angiotensinogen, into angiotensin-I, which is converted analog, 131I-19-iodocholesterol, two compounds with higher
in turn into angiotensin-II by the angiotensin-converting extraction by adrenocortical tissue became available for clini-
enzyme (ACE); finally, angiotensin-II promotes the adrenal cal purposes: 131I-6-β-iodo-methyl-norcholesterol (NP-59)
production of aldosterone. and 75Se-6-β-seleno-methyl-norcholesterol (Scintadren).
The adrenal medulla originates from the entoderm, which Typical administered activities are 37 MBq (1 mCi) for 131I-6-
makes up about 25% of the total mass of the gland, and is β-iodo-methyl-norcholesterol and about 6–8 MBq (0.16–
responsible for the production and release of catecholamines 0.22 mCi) for Scintadren. The two agents have very similar
(adrenaline and noradrenaline), under the direct control of features with a relatively high radiation dosimetry burden due
the nervous system or hormonal systems such as renin- to their long half-life, both physical and biological. In base-
angiotensin system. Catecholamines, which accumulate line physiologic conditions, uptake in the adrenals accounts
within cytoplasmic vesicles, are secreted by exocytosis and for about 0.15 ± 0.04% of the administered activity [1].
are then recycled by cells of the adrenal medulla through a Timing of image acquisition post-administration varies
“reuptake” mechanism, thanks to the presence of transporter depending on the protocols used. Generally, images are
sites on the cell membrane, called “uptake-1.” In peripheral acquired at the second, fourth, and seventh day after the i.v.
tissues, catecholamines are recycled by cells through injection of norcholesterol, using a wide-field camera range
“uptake-2” transporter sites, to be subsequently metaboli- and collimators for high energies or medium energies, in
cally degraded to metanephrines. case of the 75Se-labeled agent. When the l31I-labeled tracer is
used, thyroid blockade, starting 3 days before tracer admin-
istration and prolonged for the entire imaging procedure, is
Key Learning Points suggested in order to avoid thyroidal uptake of free iodide
• The cortical and the medullary components of the released during metabolic degradation [2].
adrenals have completely different roles and a dif- Static images are acquired with the patient lying supine
ferent embryological origin. and the gamma camera detector positioned posteriorly of the
• The glomerular zone of adrenal cortex produces lumbar region so as to include in the field of view the hepatic
mineralocorticoid hormones, mainly aldosterone, dome. A total of 500K–1M counts (about 20–30 min for each
the fascicular zone produces glucocorticoids. view) are acquired, employing a 20% window centered on
• The adrenal medulla is responsible for the produc- the photopeak of 131I (364 keV).
tion and release of catecholamines. Normal biodistribution of the radiopharmaceutical is
characterized by a physiological hepatic uptake (due to the
presence of LDL receptors); occasionally the gallbladder is
visualized. Often there is important accumulation of activity
28.1.1 Radionuclide Imaging of the Adrenal within the colon (due to biliary-enteric excretion of radio-
Cortex cholesterol), which may require the use of laxatives to accel-
erate its clearance. Due to the presence of high-hepatic
So far, steroid hormones per se have not been successfully activity (especially on the second day), which sometimes
radiolabeled for adrenocortical imaging. Therefore, imaging of prevents clear visualization of the right adrenal gland, it may
the adrenal cortex is based on the use of radiolabeled precur- be necessary to acquire supplementary images after injecting
sors of steroid hormones, such as cholesterol analogs that, a 99mTc-colloid, which is taken up by the reticuloendothelium
mimicking the fate of endogenous cholesterol, associate within of the liver. Keeping the patient in the same position and
28 Hybrid Imaging in Non-thyroidal Endocrinological Disorders 751
Fig. 28.1 Adrenocortical
scintigraphy: 131I-6-β-iodo-
methyl-norcholesterol
(NP-59) and 99mTc-
radiocolloid. Weighted
subtraction of hepatic activity
after the first acquisition
(day 2) of NP59
NP59 Radiocolloid Substraction image
using the same collimator, a static image is acquired on the pituitary; in this form pituitary ACTH production is sup-
99m
Tc peak (140 keV, 20% window) for about 5–10 min. It is pressed because of the negative feedback modulated by the
thus possible, during post-acquisition processing, to remove increased levels of circulating cortisol.
the hepatic activity by performing a “weighted” image sub- The scintigraphic pattern observed in patients with
traction of the two acquisitions (radiocholesterol image Cushing’s syndrome depends on the specific pathophysiologic
minus radiocolloid image, normalized for the counts calcu- abnormalities that cause the two different forms of the syn-
lated within a hepatic ROI) (Fig. 28.1). drome. In the ACTH-dependent disease, there is intense and
Uptake of the adrenal glands in normal conditions should be symmetrical uptake of the radiopharmaceutical in both adrenal
symmetrical, already appreciable on the second day; the right glands, which are commonly increased in size. Whereas, the
adrenal gland often appears slightly more active than the left, scintigraphic pattern of a cortisol- hypersecreting adenoma
due to some scatter effect because of its proximity to the liver. (which represents the most appropriate clinical indication for
corticoadrenal scintigraphy) consists in the monolateral visual-
ization of the adrenal gland bearing the adenoma; in this case,
the contralateral adrenal gland is not visualized because it is
Key Learning Points
hypofunctioning following inhibition of ACTH secretion
• I-6-β-iodo-methyl-norcholesterol (NP-59) and
131
(Fig. 28.2). On the other hand, cortical nodular hyperplasia is
Se-6-β-seleno-methyl-norcholesterol (Scintadren)
75
characterized by a bilateral, often asymmetric, uptake
are two radiopharmaceuticals with high extraction
(Fig. 28.3). Finally, in the rare cases of a corticoadrenal carci-
by adrenocortical tissue that became available for
noma, imaging is characterized by the lack of visualization of
clinical purposes.
the adrenal glands. In fact, in most cases uptake of radiocholes-
• Their adrenal uptake is regulated by two main con-
terol within the carcinoma is too low to visualize the tumor
trol systems: the ACTH-cortisol axis and the renin-
mass, while uptake in the normal contralateral adrenal is inhib-
angiotensin-aldosterone axis.
ited anyway because of the increased secretion of cortisol.
28.1.1.2 Adrenal Imaging in Cushing’s Syndrome Key Learning Points

The Cushing’s syndrome (initially described in 1932) is • Two forms of Cushing’s syndrome can be distin-
characterized by the set of signs and symptoms deriving guished, ACTH-dependent and ACTH-independent,
from an excess of glucocorticoid hormones (hypercorti- respectively.
solism). Two fundamental forms are recognized: (1) ACTH- • In the ACTH-dependent disease there is intense and
dependent and (2) ACTH-independent disease. The former symmetrical uptake of the radiopharmaceutical in
(which is often defined as “Cushing’s disease”) is caused by both adrenal glands.
increased pituitary secretion of ACTH, due in general to an • The scintigraphic pattern of a cortisol-hypersecreting
ACTH-secreting adenoma (resulting in bilateral adrenal adenoma (the most appropriate clinical indication
hyperplasia) or to circulating ACTH-like compounds as part for corticoadrenal scintigraphy) consists in the
of a paraneoplastic syndrome caused by, e.g., pulmonary monolateral visualization of the adrenal gland bear-
microcytoma, carcinoid tumor, and medullary thyroid carci- ing the adenoma.
noma. The ACTH-independent syndrome is caused by • In a rare corticoadrenal carcinoma causing a
increased secretion of cortisol due to a primary adrenal dis- Cushing’s syndrome, imaging is characterized by
ease (adrenal adenoma, bilateral adrenal nodular hyperpla- the lack of visualization of the adrenal glands.
sia, adrenal carcinoma) functionally autonomous from the
Fig. 28.2 Adrenocortical
scintigraphy with NP-59 in a
patient with Cushing’s
syndrome. Images show
accumulation of the tracer in
the right adrenal (with no
visualization of the
contralateral adrenal gland).
This scintigraphic pattern is
consistent with a right adrenal
adenoma
day 2 day 4 day 7
28.1.1.3 Imaging in Hyperaldosteronism

and Hyperandrogenism
Primary hyperaldosteronism (initially described in 1954) is
defined as an excess of aldosterone secretion, in most cases
caused by a corticoadrenal adenoma. Less frequently hyper-
aldosteronism is caused by bilateral hyperplasia of the glo-
merular zone of the adrenal cortex or, even more rarely, by a
carcinoma. Secondary hyperaldosteronism is generally
linked to renal hyperproduction of the hormone renin, which
activates the renin-angiotensin-aldosterone system. The typi-
cal symptoms include arterial hypertension and hypokalemia
with muscular disorders, asthenia, and adynamia. Diagnosis
is often delayed because arterial hypertension can be under-
estimated and/or treated with the common antihypertensive
drugs, thus concealing the diagnosis of Conn’s disease.
Blood shows elevated plasma and urinary aldosterone levels,
associated with suppressed renin activity.
Hyperandrogenism can be caused by excessive produc-
tion of androgens from the ovary or adrenals or by increased
peripheral conversion of steroid precursors into androgens.
Fig. 28.3 Adrenocortical scintigraphy with NP-59 in a patient with Since in these two hypersecretory syndromes the adrenal
ACTH- independent Cushing’s syndrome. Images show high tracer uptake of radiocholesterol must be investigated without inter-
uptake in both adrenals. This scintigraphic pattern is consistent with
bilateral adrenal hyperplasia. Kidney contours in the lower right image ference from ACTH stimulation (which is quantitatively the
are obtained by manually drawing the contours in the image obtained most relevant), adrenocortical scintigraphy is performed with
after administration of 99mTc-DMSA and transferring the resulting sil-
houette on the NP-59 scintigraphic image
a suppression test with dexamethasone. In particular, in order

day 2
to enhance the detection of adenomas secreting mineralocor-
ticoids, dexamethasone suppression of ACTH is usually
applied [2]. Suppression of ACTH secretion reduces radio-
cholesterol uptake in the fascicular zone of the adrenal cortex
(where cortisol is produced), thus optimizing imaging of the
ACTH-independent metabolic abnormalities occurring in the
glomerular or reticular zone. In this way, it is possible to rec-
Liver subtraction
ognize mineralocorticoid and androgen secretory adenomas
and to distinguish them from bilateral hyperplasia.
day 4 day 7
ACTH-suppressing regimens range from 1 to 4 mg of
dexamethasone administered orally for 7 days before tracer
injection and continued during the entire imaging period.
Although no data are available to select the best regimen for
effective suppression, the higher dosage is assumed to avoid
the possibility of escape from ACTH suppression. Moreover,
it is advisable to discontinue drugs interfering with the renin-
angiotensin-aldosterone system (ACE inhibitors, spironolat-
tone, and most diuretics) in order to reduce the possibility of
misdiagnoses [2, 3].
The normal adrenals are usually visualized as quite sym-
metrical areas (the right adrenal may be more evident due to Fig. 28.4 Adrenocortical scintigraphy with dexamethasone suppres-
sion test in a patient with primary hyperaldosteronism. Images show
liver overlap) starting from day 4 to 5 after tracer injection. tracer uptake in the right adrenal starting on day 2. Although with lower
Adrenal uptake accounts for 0.04 ± 0.01% of administered intensity, some tracer uptake is detectable also on the left side (more
activity in the dexamethasone-suppressed state. Unilateral or obvious after liver subtraction). Images on the 4th and 7th day confirm
bilateral adrenal visualization occurring before the fourth asymmetric bilateral adrenal hyperplasia
day after tracer administration indicates the presence of ade-
noma or bilateral hyperplasia, respectively (Fig. 28.4). 28.1.1.4 Imaging of Adrenal Incidentalomas
The exact protocol for adrenocortical scintigraphy using The wide application in the routine of high-resolution imag-
radiolabeled cholesterol analogs remains to be further defined. ing techniques (US, CT, MRI) for the evaluation of patients
It should be pointed out that this imaging modality has the with extra-adrenal abdominal symptoms is leading more and
potential limitation of not correctly identifying an adenoma more frequently to the identification of incidental adrenal
when bilateral hyperplasia coexists. Nevertheless, this tech- masses. In patients who do not have hypersecretory syn-
nique is capable of demonstrating the bilateral nature of dromes and who do not have a history of cancer, most of the
hyperfunction, which constitutes a negative predictive factor adrenal masses (70–90%) are nonfunctioning benign adeno-
for clinical improvement after unilateral adrenalectomy [4]. mas. Instead, in patients with a positive history of primary
Despite the relatively worse spatial resolution compared to non-adrenal neoplasm, about 50% of the lesions are of meta-
other imaging techniques, adrenocortical scintigraphy with static origin, arising more frequently from lung carcinomas,
radiocholesterol should be considered in patients with pri- but also mammary, renal, ovarian, gastrointestinal, lympho-
mary hyperaldosteronism and normal or equivocal CT scans. mas, and melanomas. On the other hand, incidentalomas
may be adrenal carcinomas (3%), pheochromocytomas
(4–10%), ganglioneuromas or ganglioneuroblastomas
Key Learning Points (<6%), adrenal cysts (5%), myelolipomas (7–15%), hemato-
• Primary hyperaldosteronism is defined as an excess mas and hemorrhages (<4%), and finally false incidentalo-
of aldosterone secretion, in most cases caused by a mas (<10%), originating from structures adjacent to the
corticoadrenal adenoma. adrenal glands (kidney, pancreas, spleen, lymph nodes).
• Adrenocortical scintigraphy is performed with a When an incidental adrenal mass is found, it is necessary
suppression test of ACTH with dexamethasone, in to distinguish the benign-nonfunctioning lesions (for which
order to enhance the detection of adenomas secret- conservative therapy is sufficient) from those for which sur-
ing mineralocorticoids. gical removal or a combined therapeutic approach may be
• Unilateral or bilateral adrenal visualization occur- necessary. The diagnostic process starts, therefore, with bio-
ring before the fourth day after tracer administra- chemical evaluation of the hormonal system, in order to
tion suggests the presence of adenoma or bilateral identify hyperfunctioning lesions.
hyperplasia. Although CT and MRI can often provide indications on the
etiology of incidental adrenal mass (based on densitometric or
Fig. 28.5 [18F]FDG PET/CT performed for restaging in a patient treated for small-cell lung cancer. Left adrenal mass with high [18F]FDG uptake
consistent with the clinical suspicion of metastasis
signal intensity information related to the presence/absence of

fat tissue within the lesion), scintigraphy has been used for the Key Learning Points
differential diagnosis of incidentalomas. The diagnostic para- • US, MR, and CT led to an increase of the identifica-
digm relies on the assumption that an adenoma, although non- tion of incidental adrenal masses. It is necessary to
hyperfunctioning, should concentrate radiocholesterol at the distinguish benign nonfunctioning from malignant
same level or more than the normal contralateral adrenal; incidental adrenal masses.
instead, in case of metastasis or other diseases that replace the • Adrenocortical scintigraphy has been used in the
normal corticoadrenal tissue, radiocholesterol uptake is absent. past for the differential diagnosis of incidentalo-
On the other hand, it should be considered that this approach mas, however, when CT or MRI cannot provide
has almost been abandoned in favor of using PET/CT with indications on the etiology of incidental adrenal
[18F]FDG, which can further characterize an adrenal mass by mass PET/CT with [18F]FDG, which can further
highlighting its glucose hypermetabolic state in case of pri- characterize an adrenal mass.
mary or metastatic malignant lesions (Fig. 28.5).
28.1.2 Radionuclide Imaging of the Adrenal

Medulla Key Learning Points
• Pheochromocytomas and paragangliomas are
Pheochromocytomas and paragangliomas are catecholamine-producing tumors arising from chro-
catecholamine-producing tumors arising from chromaffin maffin cells.
cells, which are located in the adrenal medulla—but also • MIBG is an analog of the neuron-blocking agent
widely present at several sites in the body. These lesions guanethidine. It is taken up through the amine
(which can be multiple, extra-adrenal, and also malignant) uptake-1 mechanism on the cell membrane of sym-
are included in the group of neuroendocrine tumors and can pathomedullary tissues and accumulates within the
be familial, either alone or in combination with other disor- cytoplasmic catecholamine storage vesicles.
ders (MEN type 2A and MEN type 2B). CT, MRI, and • Superior diagnostic performance of [123I]MIBG ver-
radionuclide imaging can all be used for imaging pheochro- sus [I31I]MIBG is due to the more favorable physical
mocytomas. Radionuclide procedures are gaining an emission properties, the important added informa-
increasing role for a theranostic management of these neu- tion that can be provided by SPECT and SPECT/CT
roendocrine tumors [5]. imaging and the more favorable dosimetry.
28.1.2.1 Radiopharmaceuticals
Meta-iodobenzylguanidine (MIBG, which can be labeled
with either 123I or 131I) is an analog of the neuron-blocking 28.1.3 Radionuclide Imaging Protocols
agent guanethidine that was first introduced in 1981 for clini- for Pheochromocytomas
cal use [6]. It is taken up mainly through the active amine
uptake-1 mechanism present on the cell membrane of sym- 28.1.3.1 MIBG
pathomedullary tissues, after which it accumulates within The radiopharmaceutical is injected i.v. after adequate prep-
the cytoplasmic catecholamine storage vesicles. The supe- aration to block thyroidal uptake of free radioiodide released
rior diagnostic performance of [123I]MIBG versus [I31I]MIBG during metabolic degradation of the tracer. Injection should
for the imaging characterization of pheochromocytomas has not be too fast, in order to minimize the risk of a hypertensive
largely been validated, due to the more favorable physical crisis due to possible release into the circulation of catechol-
emission properties for gamma camera imaging, also consid- amines displaced from their intracellular storage sites, espe-
ering the important added information provided by SPECT cially when low-specific activity tracers are administered to
(preferably SPECT/CT) imaging. Moreover, the radiation patients with pheochromocytoma. Planar whole-body and
dosimetry of [123I]MIBG is far more favorable than that of spot images as well as SPECT images (preferably with
[123I]MIBG. SPECT/CT acquisitions) are usually acquired 4 and 24 h
Several drugs may interfere with MIBG uptake (such as, after tracer administration; delayed images may be acquired
e.g., tricyclic antidepressants, antipsychotics, beta-blockers, also beyond the 24-h time-point when [13II]MIBG is
reserpine, calcium channel blockers, cocaine, sympathomi- employed (Fig. 28.6). Quantitative evaluation of MIBG
metics), and therefore they should be discontinued before uptake for increasing diagnostic accuracy and for dosimetric
imaging for a period adequate to ensure complete pharmaco- purposes before therapy with [13II]MIBG treatment is also
logical washout. possible.
111
In-DTPA-d-Phe1-octreotide (111In-pentetreotide or
Octreoscan®), a somatostatin analog, has recently been 28.1.3.2 Octreoscan®
introduced for diagnosing a wide range of neuroendocrine Similarly as for scintigraphy with MIBG, planar whole-body
tumors, including pheochromocytomas [7]. This imaging and spot images as well as SPECT images (preferably with
modality is based on the expression of octreotide binding by SPECT/CT acquisitions) are usually acquired 4 and 24 h
somatostatin receptors on most neuroendocrine cells. The after tracer administration; delayed images beyond the 24-h
octreotide-receptor complex is internalized into intracellu- time-point are sometimes required to better clarify sites of
lar vesicles and then delivered to lysosomes for metabolic uptake in the abdomen.
degradation. Adrenal and extra-adrenal pheochromocytomas are visu-
There is an ongoing debate whether or not to discontinue alized as foci of MIBG and/or Octreoscan® uptake not related
treatment with unlabeled somatostatin analogs, which are to physiologic site of uptake/excretion. More details on how
used to control symptoms caused by neuroendocrine tumors, to interpret and report scintigraphy with MIBG or
before imaging with Octreoscan®. Octreoscan® are given in Chap. 29.
Fig. 28.6 Scintigraphy with [123I]MIBG in a patient with suspected left obtained by manually drawing the contours in the image obtained after
adrenal pheochromocytoma. Images show tracer uptake in the left adre- administration of 99mTc-DMSA and transferring the resulting silhouette
nal, obvious both at 4 h and 24 h after tracer injection (left panel and on the [131I]MIBG scintigraphic image
right panel, respectively). Kidney contours in the lower right image is
Fig. 28.7 Scintigraphy with

111
In-pentetreotide. Bulky
paragangliomas with high
expression of somatostatin
receptors in the neck and in
the cervico-mediastinal region
Anteriore Posteriore
28.1.3.3 Current Role of MIBG and Octreoscan® high physiological uptake of the labeled somatostatin analog
Scintigraphy for Pheochromocytoma in the kidneys. On the other hand, higher sensitivity of scin-
MIBG scintigraphy is probably the optimal approach to tigraphy with Octreoscan® versus MIBG has been observed
radionuclide imaging of pheochromocytomas. In fact, for paragangliomas located in the head and neck region
Octreoscan® imaging is less sensitive than MIBG scintigra- (Fig. 28.7) [9]. Furthermore, Octreoscan® scintigraphy might
phy, especially for adrenal lesions [8], mainly because of the be more accurate than MIBG scintigraphy for detecting dis-
tant metastases, thus suggesting a complementary role of Better diagnostic performance has instead been reported
both imaging agents MIBG when distant metastases are sus- for PET/CT with 18F-DOPA than with [18F]FDG. As an ana-
pected in patients with pheochromocytoma. log of the dopamine precursor, 18F-DOPA is concentrated
Sensitivity of MIBG scintigraphy for detecting primary by the chromaffin cells, where it is transformed by the
pheochromocytomas ranges from 82% to 100%, with enzyme amino acid decarboxylase (AADC) to
88–100% specificity [10, 11]. The sensitivity of [123I] 18
F-fluorodopamine, which is stored in the same storage
MIBG scintigraphy for adrenal tumors is quite similar to vesicles containing catecholamines—where MIBG is also
the sensitivities of CT and MR imaging [12], although CT stored. Therefore, 18F-DOPA has the same specificity as
provides superior anatomical information and thus consti- MIBG for the c hromaffin tissues, with all the advantages of
tutes an important guide for the surgeon. Nevertheless, an better spatial resolution and greater signal-to-noise ratios
advantage of MIBG scintigraphy over other imaging tech- of PET/CT imaging compared to single-photon imaging.
niques is its ability to identify extra-adrenal pheochromo- Furthermore, the intrinsic hybrid nature of PET/CT imag-
cytomas, due to the whole-body nature of radionuclide ing allows precise anatomical localization of the lesions
imaging. For the same reason, MIBG scintigraphy is par- detected with the use of the PET tracer. Current literature
ticularly useful for localizing distant metastases from data demonstrate better sensitivity and specificity of
malignant pheochromocytomas. Moreover, also due to its 18
F-DOPA PET/CT compared to MIBG scintigraphy, in
high specificity, MIBG scintigraphy is to be preferred over particular for extra-adrenal pheochromocytomas of the
other imaging modalities in case of suspected recurrences head and neck region (Fig. 28.8).
post-surgery, especially when scars and/or other anatomi- Other PET tracers, such as 18F-Fluorodopamine, [11C]
cal distortions produced by surgery make interpretation of hydroxy-ephedrine, and [11C]epinephrine, are currently
purely morphologic imaging difficult [13–15]. Finally, under clinical validation [19].
hybrid SPECT/CT imaging considerably improves ana-
tomical localization of extra-adrenal lesions. Familial
pheochromocytoma and pheochromocytoma in MEN and
other syndromes have also been successfully localized Key Learning Points
with MIBG scintigraphy in a preclinical stage, although a • MIBG scintigraphy is probably the optimal
higher tracer uptake has been reported in the sporadic ver- approach to radionuclide imaging of
sus the familial forms [8]. pheochromocytomas.
• The sensitivity of [123I]MIBG scintigraphy for adre-
28.1.3.4 [123I]MIBG Versus [131I]MIBG nal tumors is quite similar to the sensitivities of CT
The non-negligible false-negative rates that had been and MR imaging.
reported in early experience using [131l]MIBG have been • An advantage of MIBG imaging is its ability to
markedly reduced with the use of [123I]MIBG, due to a com- identify extra-adrenal pheochromocytomas and to
bination of factors that include (1) higher photon flux of 123I localize distant metastases from malignant
versus 131I, (2) greater detection efficiency of gamma cam- pheochromocytomas.
eras for 123I versus 131I, (3) better spatial resolution of 123I • Hybrid SPECT/CT imaging improves anatomical
versus 131I imaging, (4) lower radiation dosimetry burden of localization of extra-adrenal lesions.
123
I versus 131I (which allows for administration of greater • 18F-DOPA has the same specificity as MIBG for the
activities—yielding better counting statistics), and (5) easier chromaffin tissues, with all the advantages of better
acquisition of SPECT/CT imaging. Therefore, whenever spatial resolution and greater signal-to-noise ratios
possible [I23I]MIBG is to be preferred to [131I]MIBG for of PET/CT imaging.
imaging [15, 16].
28.1.3.5 PET Tracers for Pheochromocytomas

The diagnostic performance of [18F]FDG PET/CT is in gen- 28.2 Parathyroid Glands
eral lower than that of [123I]MIBG scintigraphy in the identi-
fication of adrenal pheochromocytomas, with variable After originating from the endoderm of the third and fourth
sensitivities—higher in malignant than in benign pheo- pharyngeal pouches, the parathyroid glands migrate during
chromocytomas (82% versus 58%), as expected because of fetal life toward their final location in the neck. In particular,
the higher glucose metabolic activity in more actively grow- the glands that are originally positioned more cranially (in
ing tumors [17, 18]. Nonetheless, so far [18F]FDG PET/CT is the third pharyngeal pouch) follow in part the descent of the
not part of the common diagnostic protocol for thymus to reach their final location at the posterolateral sur-
pheochromocytoma. face of the lower thyroid lobes, whereas, the glands that are
Fig. 28.8 F-DOPA PET/CT showing a right laterocervical mass with intense uptake consistent with a paraganglioma
18
originally positioned more caudally (in the fourth pharyngeal Histologically, the adult parathyroid tissue is constituted
pouch) maintain a close association with the embryologic by a predominant type of cells called chief cells, followed by
and fetal thyroid gland until the descent of both structures to the oxyphilic cells (whose histologic appearance is linked to
their final location in the neck. This complex pattern of their abundance in mitochondria) and by the intermediate
migration of the parathyroids, which brings one pair from a stage of transitional oxyphilic cells.
cranial to a caudal position relative to the other pair, explains The parathyroid hormone (PTH) acts on different tissues
why the surgical approach to the parathyroid glands can be and apparatuses with multifaceted functions, the combined
intricate and variable. effect of which is to increase calcium concentration in the
There are normally two pairs of parathyroid glands in blood. In order to achieve this effect, PTH stimulates bone
adults, each one measuring approximately 6 × 4 × 2 mm and resorption (thereby mobilizing calcium from bone to circu-
weighing approximately 30–50 mg, although supernumerary lating fluids) and promotes calcium resorption in the kid-
parathyroid glands can be found in 2–5% of the general pop- neys; at the same time, it decreases tubular resorption of
ulation. While the upper parathyroid glands are in most cases phosphate. Moreover, PTH stimulates the synthesis of the
located posteriorly to the middle-upper third of the thyroid active form of vitamin D, which in turn stimulates calcium
lobe, location of the inferior parathyroid glands is more vari- absorption in the gastrointestinal tract.
able, probably because of their more complex migration pro- Primary hyperparathyroidism (HPTH) is characterized
cess. In approximately 50% of surgeries, the lower by inappropriate PTH production, which causes hypercalce-
parathyroid glands can be found within a 2 cm radius poste- mia. Primary HPTH can be caused by either a single parathy-
riorly or laterally to the lower pole of the thyroid lobe. Other roid adenoma (80–85% of the cases), hyperplasia or multiple
possible locations include the thyrothymic ligament, the thy- adenomas (10–15% of the cases), or parathyroid carcinoma
mus, and even an intrathyroidal location—with decreasing (0.5–1% of the cases). Familial forms of primary HPTH
frequency. include multiple endocrine neoplasia type I (MEN I) and
type II (MEN II), the HPTH-jaw tumor syndrome, familial presurgical imaging not only reduces the surgical time by
hypocalciuric hypercalcemia, and familial isolated HPTH. limiting surgical exploration to the affected side but, above
Secondary HPTH is the result of long-standing hypocal- all, allows the detection of parathyroid adenomas in ectopic
cemia, as a response of the parathyroid glands to maintain sites. In this regard, while 80–85% of parathyroid adenomas
calcium homeostasis. The most frequent cause of secondary are found in their normal location adjacent to the thyroid
HPTH is chronic renal failure, secondary HPTH developing gland, 15–20% are ectopic. Overall ectopic locations include
in about 90% of the patients undergoing hemodialysis. anterior–superior mediastinum (either within the thymus or
Tertiary HPTH is defined as the condition of persisting in a juxta-thymic position), posterior–superior mediastinum
HPTH after successfully treating the cause of a secondary along the esophagus, lower thyroid lobe (2–3% of all para-
form of HPTH, as it can happen following, e.g., correction of thyroid adenomas), and middle mediastinum (very rarely).
chronic renal failure with kidney transplantation. Tertiary
HPTH is usually sustained by hyperplasia of all four glands,
but in over 20% of the patients single or double adenomas 28.2.2 Radionuclide Parathyroid Imaging
are present.
In 5–10% of patients who undergo surgery for PHPT, per- The lack of a parathyroid-specific radiopharmaceutical has
sistent or recurrent HPTH can occur. Hyperparathyroidism stimulated in the 1980s the development of double-tracer sub-
presenting after a period of >6 months of normocalcemia fol- traction parathyroid scintigraphy. The rationale underlying
lowing surgery is defined as “recurrent hyperparathyroid- this approach relies on the concurrent utilization of a tracer
ism” and is usually linked to regrowth of the remaining taken up both by thyroid and parathyroid tissue and of a tracer
parathyroid tissue. “Persistent hyperparathyroidism” is taken up only by the thyroid; after suitable digital subtraction
instead defined as the immediate failure of surgical treatment processing, it is thus possible to visualize the pathological
to restore normal PTH and calcium levels and is generally parathyroid tissue. The earliest version of this technique was
caused by inaccurate/incomplete localization of adenomas, based on the use of 75Se-methionine (a radiolabeled amino
insufficient resection of easily recognized multigland dis- acid actively concentrated both in the thyroid and the parathy-
ease, and (though rarely) by the presence of (unrecognized) roid tissues because of their high metabolic activity) and on
metastatic parathyroid carcinoma. simultaneous or sequential administration of 131I-iodide
(which selectively localizes in the thyroid gland) [21, 22].
Nevertheless, this technique did not gain popularity because
of poor performance (linked to technical limitations typical of
Key Learning Points
the imaging equipment available at the time) and because of
• Primary hyperparathyroidism is characterized by
concerns about a relatively high radiation dosimetry burden
inappropriate PTH production and can be caused by
associated with the use of 75Se. Better results were observed
either a single parathyroid adenoma, hyperplasia or
when 75Se-methionine was replaced with 201Tl-chloride and
multiple adenomas, or a parathyroid carcinoma. 131
I-iodide was replaced with 99mTc-pertechnetate (99mTcO4−).
• Secondary hyperparathyroidism develops in about
Nevertheless, also this double-isotope scintigraphic proce-
90% of the patients with chronic renal failure under-
dure failed to perform better than other imaging modalities,
going hemodialysis.
as a result of the degree of intrinsic variability and because of
• Tertiary HPTH is defined as the condition of per-
low reproducibility in different centers.
sisting HPTH after successfully correcting chronic
The real turning point in parathyroid scintigraphy when
renal failure with kidney transplantation.
Coakley and colleagues in the late 1980s incidentally
• Persistent or recurrent HPTH can occur in 5–10%
observed that, during myocardial perfusion scintigraphy
of patients who undergo surgery.
with 99mTc-Sestamibi, this radiopharmaceutical was taken up
by abnormal parathyroid tissue [23]. Following their seminal
report, the more favorable imaging and dosimetric character-
28.2.1 Role of Preoperative Imaging istics of 99mTc-Sestamibi and its lower cost compared with
201
Tl-chloride quickly led to the replacement of 201Tl-chloride
Preoperative imaging is important to identify the parathyroid with 99mTc-Sestamibi. Parathyroid scintigraphy with
glands that are enlarged/hyperfunctioning because of hyper- 99m
Tc-Sestamibi became the standard imaging procedure to
plasia and/or adenoma [20]. Visualization of normal parathy- evaluate patients with primary PHPT.
roid glands using conventional imaging is virtually The mechanism of 99mTc-Sestamibi uptake in the parathyroid
impossible; in fact, because of their tissue characteristics and tissue depends upon the fast transmembrane diffusion of this
small size and weight (less than 40 mg), they cannot be dis- lipophilic compound, further enhanced by its cationic nature
tinguished from the adjacent thyroid parenchyma. Adequate that produces an electrical gradient between the extracellular
compartment (positively charged) and the intracellular compart- ent timings of administration of 99mTcO4− relative to
ment (negatively charged). Once inside the cell, 99mTc-Sesta- 99m
Tc-Sestamibi, different activities of the two radiopharma-
mibi accumulates in mitochondria as a function of metabolic ceuticals, and possible use of potassium perchlorate to
activity. Uptake in hyperplastic or adenomatous parathyroid induce faster washout of 99mTcO4− from the thyroid
glands is therefore linked to a combination of factors including parenchyma.
blood flow, gland size, and mitochondrial activity. Irrespective of the specific double-tracer protocol
Similar to 75Se-methionine and 201Tl-chloride, 99mTc- adopted, parathyroid imaging consists of planar acquisitions
Sestamibi accumulates both in thyroid and in parathyroid tis- of the neck and upper chest in the anterior view, extending
sue within minutes after intravenous administration. from the submandibular salivary glands to the upper part of
However, 99mTc-Sestamibi is released much faster from thy- myocardium in order to ensure the detection of ectopic para-
roid than from parathyroid tissue, a pathophysiologic pattern thyroid tissue. The use of a LEHR parallel-hole collimator
possibly related to downregulation in the parathyroid of the is generally recommended, while a pinhole collimator can
membrane-associated P-glycoprotein system related to mul- be used to obtain a magnified, higher-resolution image
tidrug resistance, a system for which 99mTc-Sestamibi is also focused on the neck. During post-acquisition processing, by
a substrate [24]. This differential washout rate of 99mTc- employing a suitable normalization factor, the thyroid-only
Sestamibi from the thyroid relative to the parathyroid tissue image is subtracted from the combined thyroid + parathy-
permits parathyroid scintigraphy to be performed with this roid image, in order to visualize the abnormal parathyroid
single tracer, according to the so-called dual-phase 99mTc- tissue only (Fig. 28.9). Protocols employing a simultaneous
Sestamibi acquisition protocol (see further below) [25]. acquisition using separate energy windows for two radio-
pharmaceuticals, 99mTc-sestamibi and 123I-iodide, are also
possible.
Key Learning Points
• In double-tracer subtraction parathyroid scintigraphy,
a first tracer is taken up both by thyroid and parathy- Key Learning Points
roid tissues and a second tracer only by the thyroid. • The double-tracer protocols currently in use are
After digital subtraction processing, it is possible to based on sequential image acquisition of two differ-
visualize only the pathological parathyroid tissue. ent tracers: 99mTcO4− and 99mTc-Sestamibi.
• 99mTc-Sestamibi accumulates both in thyroid and in • Parathyroid imaging consists of planar acquisitions
parathyroid tissues, but it is released much faster of the neck and upper chest in the anterior view.
from thyroid. This differential washout rate of • Protocols employing a simultaneous acquisition
99m
Tc-Sestamibi from the thyroid relative to the using separate energy windows for two radiophar-
parathyroid tissue permits parathyroid imaging maceuticals, 99mTc-sestamibi and 123I-iodide, are
with the so-called dual-phase 99mTc-Sestamibi also possible.
acquisition protocol.
28.2.2.1 Double-Tracer Parathyroid Scintigraphy 28.2.2.2 Dual-Phase Parathyroid Scintigraphy

Preparation of patients for double-tracer parathyroid scintig- There is no need for a specific preparation of patients before
raphy is identical to that for a thyroid scan, requiring preven- dual-phase parathyroid scintigraphy. The dual-phase proto-
tion of expansion of the body pool of iodine. In particular, col is currently based on planar imaging of the neck and
administration of iodinated contrast medium should be chest at 15 min, then 2–3 h after the i.v. injection of 740 MBq
avoided for at least 2 weeks before double-tracer parathyroid 99m
Tc-Sestamibi. As for the double-tracer protocol, the use of
scintigraphy; furthermore, whenever possible thyroid hor- a LEHR parallel-hole collimator is generally recommended,
mone therapy should be withheld for 2–3 weeks, and treat- but a pinhole collimator can be also used to obtain higher-
ment with methimazole or propylthiouracil should be resolution images focused on the neck.
discontinued for at least 1 week. A hyperfunctioning parathyroid adenoma appears as an
The double-tracer protocols currently in use are based on area of early radiopharmaceutical uptake that persists on late
sequential image acquisition of two different tracers, one for imaging (Fig. 28.10). Thyroid nodules that concentrate
the thyroid and one for combined thyroid and parathyroid 99m
Tc-Sestamibi avidly can cause difficulties in the interpre-
tissues. Variations in parathyroid imaging protocols using tation of images. To reduce this drawback, 99mTc-pertechnetate
99m
TcO4− (thyroid-only imaging) and 99mTc-Sestamibi (com- or 123I-iodide scintigraphy can be performed, for correlation
bined thyroid and parathyroid imaging) are based on differ- with the 99mTc-Sestamibi images, in patients with thyroid
Fig. 28.9 Double-tracer
parathyroid scintigraphy. Left
panel: 99mTcO4− thyroid-only
imaging. Center panel:
99m
Tc-Sestamibi combined
thyroid and parathyroid
imaging. Right panel:
subtraction image normalized
for thyroid counts, showing a
parathyroid adenoma at the
right inferior pole of the
thyroid
99mTcO - 99mTc-Sestamibi 99mTc-Sestamibi
4
–
99mTcO -
4
Tc-Sestamibi
99m
Tc-Sestamibi
99m
TcO4-
99m
early (15 min) delayed (2.5 h)
Fig. 28.10 Dual-phase parathyroid scintigraphy with 99mTc-Sestamibi. acquisition. Two focal areas of persistent uptake at the lower and upper
Left panel: early (15 min) 99mTc-Sestamibi image. Center panel: late poles of the left thyroid lobe, respectively, are depicted on late 99mTc-
(2.5 h) 99mTc-Sestamibi image. Right panel: visualization of the thyroid Sestamibi image that are consistent with two parathyroid adenomas
obtained after the administration of 99mTcO4− at the end of late image
nodular goiter after completing the delayed acquisition. It 28.2.2.3 SPECT and Hybrid SPECT/CT
should also be considered that some parathyroid adenomas A SPECT acquisition of the neck and thorax after planar
exhibit a rapid washout of 99mTc-Sestamibi, thus possibly imaging is routinely performed after the early 99mTc-
resulting in false-negative results. Sestamibi acquisition in the dual-phase protocol. SPECT
imaging of the neck and thorax improves both sensitivity and
accuracy of double-tracer parathyroid scintigraphy as well,
especially in case of ectopic parathyroid glands. Although
SPECT is currently considered a useful integration modality
for parathyroid scintigraphy, when employed as a stand-
Key Learning Points alone technique, it does not provide reliable anatomical land-
• The dual-phase protocol is based on planar imaging marks due to the lack of precise tomographic orientation in
of the neck and chest at 15 min, then 2–3 h after the the neck and mediastinum compartments. With the availabil-
i.v. injection of 99mTc-Sestamibi. ity of new hybrid SPECT/CT systems, it is possible to co-
• A hyperfunctioning parathyroid adenoma appears register in the same session SPECT and CT images of neck
as an area of early radiopharmaceutical uptake that and mediastinum, thus allowing fusion of functional data
persists on late imaging. (SPECT component) with the precise anatomical location
• Thyroid nodules that concentrate 99mTc-Sestamibi (CT component). As a result, SPECT/CT not only improves
can cause difficulties in the interpretation of images. sensitivity (through better attenuation correction based on
Moreover, some parathyroid adenomas exhibit a the patient-specific tissue attenuation data obtained with CT)
rapid washout of 99mTc-Sestamibi. but also improves specificity of parathyroid scintigraphy
with 99mTc-Sestamibi (Fig. 28.11).
Fig. 28.11 Early SPECT/CT imaging with 99mTc-Sestamibi showing a bulky parathyroid adenoma developing in the upper mediastinum. Hybrid
imaging can add information that helps surgical planning
In particular, SPECT/CT imaging reduces the number of

false positive findings linked to physiological and/or para- Key Learning Points
physiological sites of 99mTc-Sestamibi uptake/accumula- • SPECT is currently considered a useful integration
tion, especially in patients with recurrent PHPT, in whom modality for parathyroid scintigraphy, but when
normal anatomy can be distorted by scar tissue. Early employed as a stand-alone technique, it does not
SPECT/CT in combination with any modality for delayed provide reliable anatomical landmarks.
imaging emerged as the best overall protocol. In particular, • SPECT/CT improves both sensitivity and specificity
the main advantage of SPECT/CT is its ability to distin- of parathyroid scintigraphy with 99mTc-Sestamibi.
guish inferior from infero-posterior glands (superior glands • Whenever possible, dual-phase imaging with early
migrated in the tracheoesophageal groove). As expected, SPECT/CT is recommended.
for all three techniques (planar imaging, SPECT, and
SPECT/CT), dual-phase imaging is superior to single-
phase imaging. Thus, whenever possible, dual-phase imag- 28.2.2.4 PET/CT of the Parathyroids
ing with early SPECT/CT is recommended, especially The use of PET/CT with different tracers has also been pro-
when a minimally invasive surgical approach is being posed for localizing hyperfunctioning parathyroid tissue,
considered. particularly in the setting of recurrent HPTH and/or failure
Subtraction protocols using SPECT/CT images have of conventional radionuclide imaging. Although the perfor-
recently been proposed with 123I-iodide and 99mTc-Sestamibi, mance of these newer techniques is reported to be highly
simultaneously acquiring dual-isotope images. Then, utiliz- promising, clinical experience is still limited to a few spe-
ing the attenuation-corrected SPECT sections, it is possible cialized centers. The PET tracers employed in this scenario
to obtain 99mTc-Sestamibi SPECT subtracted of the 123I-iodide include [18F]FDG, [11C]methionine, 18F-fluorocholine, and
SPECT. 18
F-DOPA as the most widely employed agents [26–28]. On
Fig. 28.12 18F-fluorocholine PET/CT performed in a patient with uptake of 18F-fluorocholine is observed also in the thyroid parenchyma,
prostate cancer shows an incidental small nodule with radiotracer consistent with a known chronic thyroiditis with nodules mainly located
uptake inferior to the left thyroid lobe. Clinical follow-up and surgery in the left thyroid lobe
confirmed the presence of a parathyroid adenoma. Inhomogeneous
the basis of recent systematic reviews and meta-analyses, the 28.2.3 Imaging of Parathyroid Carcinoma
most promising PET tracers appear to be [11C]methionine
and 18F-fluorocholine (Fig. 28.12). Less than 1% of the cases of primary HPTH are sustained by
a parathyroid carcinoma, a very rare endocrine malignancy.
Since histology of parathyroid tumors can be equivocal or
frankly misleading, the differential diagnosis between a
parathyroid carcinoma and an adenoma is often made only
Key Learning Points when local recurrence or metastases occur.
• PET/CT with different tracers has also been pro- The same imaging techniques as those used for localizing
posed for localizing hyperfunctioning parathyroid benign parathyroid disease are helpful also for imaging para-
tissue. thyroid cancers. The accuracy of radionuclide imaging
• The most promising PET tracers appear to be [11C] depends on the size and site of the parathyroid carcinoma,
methionine and 18F-fluorocholine. while the suspected tissue always requires histologic
confirmation.
Ultrasound examination of the neck is useful especially References

for detecting a space-occupying lesion, while it can also help
to distinguish a parathyroid carcinoma from an adenoma, as 1. Gross BH, Shapiro B. Adrenocortical scintigraphy. In: Khalkhali I,
Maublant J, Goldsmith S, editors. Nuclear oncology. Philadelphia,
well as invasion in the surrounding tissue and metastasis PA: Lippincott Williams & Wilkins; 2001. p. 461–74.
in local lymph nodes. CT and MR have been variously used 2. Gross MD, Shapiro B, Shreve P. Radionuclide imaging of the adre-
to localize parathyroid carcinomas and to detect mediastinal nal cortex. Q J Nucl Med. 1999;43:224–32.
and thoracic recurrences or distant metastases. 3. Francis IR, Gross MD, Shapiro B, Korobkin M, Quint LE. Integrated
imaging of adrenal disease. Radiology. 1992;184:1–13.
99m
Tc-Sestamibi scintigraphy can be successful for preop- 4. Nocaudie-Calzada M, Huglo D, Lambert M, Ernst O, Proye C,
erative localization of a parathyroid malignancy (with 80–90% Wemeau JL, et al. Efficacy of iodine-131 6-beta-methyl-iodo-19-
sensitivity). However, it cannot distinguish a parathyroid ade- norcholesterol scintigraphy and computed tomography in patients
noma from a carcinoma, since 99mTc-Sestamibi localizes in with primary aldosteronism. Eur J Nucl Med. 1999;26:1326–32.
5. Troncone L, Rufini V. Nuclear medicine therapy of pheochromocy-
both parathyroid adenomas and parathyroid carcinoma. The toma and paraganglioma. Q J Nucl Med. 1999;43:344–55.
National Cancer Database report by the American Cancer 6. Sisson JC, Frager MS, Valk TW, Gross MD, Swanson DP, Wieland
Society estimates that 15–17% of parathyroid carcinomas, DM, et al. Scintigraphic localization of pheochromocytoma. N
which look identical to parathyroid adenomas on 99mTc-Sesta- Engl J Med. 1981;305:12–7.
7. Krenning EP, Kwekkeboom DJ, Bakker WH, Breeman WA,
mibi scintigraphy, have concomitant lymph node metastasis at Kooij PP, Oei HY, et al. Somatostatin receptor scintigraphy with
the time of diagnosis [29]. Once parathyroid carcinoma has [111ln-DTPA-D-Phel]- and [1231- Tyr3]-octreotide: the Rotterdam
been diagnosed and resected, 99mTc-Sestamibi scintigraphy experience with more than 1000 patients. Eur J Nucl Med.
can identify recurrences in the surgical bed, as well as metas- 1993;20:716–31.
8. van der Harst E, de Herder WW, Bruining HA, Bonjer HJ, de
tases in the contralateral neck lymph nodes, and metastasis at Krijger RR, Lamberts SW, et al. [123I]metaiodobenzylguanidine and
distant sites such as lung and bone [30]. The evaluation of [11lIn]octreotide uptake in benign and malignant pheochromocyto-
bone lesions can be challenging from the diagnostic point of mas. J Clin Endocrinol Metab. 2001;86:685–93.
view, because skeletal brown tumors (a possible consequence 9. Kwekkeboom DJ, van Urk H, Pauw BK, Lamberts SW, Kooij PP,
Hoogma RP, et al. Octreotide scintigraphy for the detection of para-
of HPTH) accumulate 99mTc-Sestamibi as well, thus mimick- gangliomas. J Nucl Med. 1993;34:873–8.
ing metastatic parathyroid carcinoma [31]. 10. Maurea S, Cuocolo A, Reynolds JC, Tumeh SS, Begley MG,

PET with [18F]FDG has also shown to be very helpful for Linehan WM, et al. Iodine-131-metaiodobenzylguanidine scintig-
detecting metastatic parathyroid cancers, although brown raphy in preoperative and postoperative evaluation of paraganglio-
mas: comparison with CT and MRI. J Nucl Med. 1993;34:173–9.
tumors can be [18F]FDG-avid as well, thus being possibly 11. Berglund AS, Hulthen UL, Manhem P, Thorsson O, Wollmer P,
incorrectly interpreted as metastatic bone lesion [32]. Due to Törnquist C. Metaiodobenzylguanidine (MIBG) scintigraphy and
the rarity of parathyroid carcinoma, there is very limited computed tomography (CT) in clinical practice. Primary and sec-
experience with the use of PET agents other than [18F]FDG, ondary evaluation for localization of phaeochromocytomas. J Intern
Med. 2001;249:247–51.
occasional reports being usually included in case study pop- 12. Shapiro B, Copp JE, Sisson JC, Eyre PL, Wallis J, Beierwaltes
ulations of HPTH patients in general. Similarly as for benign WH. Iodine-131 metaiodobenzylguanidine for the locating of sus-
forms of HPTH, the performance of [11C]methionine PET/ pected pheochromocytoma: experience in 400 cases. J Nucl Med.
CT is expected to be especially promising in patients with 1985;26:576–85.
13. Velchik MG, Alavi A, Kressel HY, Engelman K. Localization of
parathyroid carcinoma. pheochromocytoma: MIGB, CT, and MRI correlation. J Nucl Med.
1989;30:328–36.
14. Francis IR, Glazer GM, Shapiro B, Sisson JC, Gross
BH. Complementary roles of CT and 131I-MIBG scintigraphy in diag-
nosing pheochromocytoma. AJR Am J Roentgenol. 1983;141:719–25.
Key Learning Points 15. Furuta N, Kiyota H, Yoshigoe F, Hasegawa N, Ohishi Y. Diagnosis
• Parathyroid carcinoma is a rare endocrine malignancy. of pheochromocytoma using [1231]- compared with [1311]-metaiodo-
benzylguanidine scintigraphy. Int J Urol. 1999;6:119–24.
• The same imaging techniques as those used for
16. Lynn MD, Shapiro B, Sisson JC, Beierwaltes WH, Meyers LJ,
localizing benign parathyroid disease are helpful Ackerman R, et al. Pheochromocytoma and the normal adrenal
for imaging parathyroid cancers. medulla: improved visualization with I-123 MIBG scintigraphy.
• 99mTc-Sestamibi scintigraphy cannot distinguish a Radiology. 1985;155:789–92.
17. Pasquali C, Rubello D, Sperti C, Gasparoni P, Liessi G, Chierichetti
parathyroid adenoma from a carcinoma, but it can
F, et al. Neuroendocrine tumor imaging: can 18F-fluorodeoxyglucose
be used to identify recurrences, as well as lymph positron emission tomography detect tumors with poor prognosis
node and distant metastases. and aggressive behavior? World J Surg. 1998;22:588–92.
18. Shulkin BL, Thompson NW, Shapiro B, Francis IR, Sisson

However, skeletal brown tumors accumulate 99mTc- JC. Pheochromocytomas: imaging with 2-[fluorine-18]fluoro-2-
Sestamibi mimicking a metastatic parathyroid deoxy-D-glucose PET. Radiology. 1999;212:35–41.
carcinoma. 19. Bergström M, Bonasera TA, Lu L, Bergström E, Backlin C,

Juhlin C, et al. In vitro and in vivo primate evaluation of carbon-
Also PET with [18F]FDG can be helpful for detect- 11-etomidate and carbon-11-metomidate as potential tracers for
ing metastatic parathyroid cancers. PET imaging of the adrenal cortex and its tumors. J Nucl Med.
1998;39:982–9.
20. Mariani G, Gulec SA, Rubello D, Boni G, Puccini M, Pelizzo MR, (Baltimore). 2015;94(41):e1701. https://doi.org/10.1097/
et al. Preoperative localization and radioguided parathyroid sur- MD.0000000000001701.
gery. J NUcl Med. 2003;44:1443–58. 27. Caldarella C, Treglia G, Isgrò MA, Giordano A. Diagnostic per-
21. Goodwin DA, Crowley LG, Camargo CA. Localization of a medi- formance of positron emission tomography using 11C-methionine
astinal adenoma by selenomethionine Se75 scanning. JAMA. in patients with suspected parathyroid adenoma: a meta-analysis.
1969;208:2333–5. Endocrine. 2013;43:78–83.
22. DiGiulio W, Morales JO. The value of the selenomethionine Se 75 28. Kluijfhout WP, Pasternak JD, Drake FT, Beninato T, Gosnell JE,
scan in preoperative localization of parathyroid adenomas. JAMA. Shen WT, et al. Use of PET tracers for parathyroid localization:
1969;209:1873–80. a systematic review and meta-analysis. Langenbecks Arch Surg.
23. Coakley AJ, Kettle AG, Wells CP, O’Doherty MJ, Collins RE. 2016;401(7):925–35.
99
Tcm sestamibi – a new agent for parathyroid imaging. Nucl Med 29. Hundahl SA, Fleming ID, Fremgen AM, Menck HR. Two hun-
Commun. 1989;10:791–4. dred eighty-six cases of parathyroid carcinoma treated in the U.S.
24. Mitchell BK, Cornelius EA, Zoghbi S, Murren JR, Ghoussoub between 1985–1995: a National Cancer Data Base Report. The
R, Flynn SD, Kinder BK. Mechanism of technetium 99m sesta- American College of Surgeons Commission on Cancer and the
mibi parathyroid imaging and the possible role of p-glycoprotein. American Cancer Society. Cancer. 1999;86:538–44.
Surgery. 1996;120:1039–45. 30. Al-Sobhi S, Ashari LH, Ingemansson S. Detection of metastatic
25. Taillefer R, Boucher Y, Potvin C, Lambert R. Detection and
parathyroid carcinoma with Tc-99m sestamibi imaging. Clin Nucl
localization of parathyroid adenomas in patients with hyperpara- Med. 1999;24:21–3.
thyroidism using a single radionuclide imaging procedure with 31. Santiago Chinchilla A, Ramos Font C, Murosde Fuentes MA,

technetium-99m-sestamibi (double-phase study). J Nucl Med. Navarro-Pelayo Láinez M, Palacios Gerona H, et al. False negative
1992;33:1801–17. of the scintigraphy with 99mTc-sestamibi in parathyroid carcinoma
26. Michaud L, Balogova S, Burgess A, Ohnona J, Huchet V,
with associated brown tumors. Contributions of the 18F-FDG-PET/
Kerrou K, et al. A pilot comparison of 18F-fluorocholine PET/ CT. Rev Esp Med Nucl. 2011;30:174–9.
CT, u ltrasonography and 123I/99mTc-sestaMIBI dual-phase 32. Kemps B, van Ufford HQ, Creyghton W, de Haas M, Baarslag HJ,
dual-isotope scintigraphy in the preoperative localization of Rinkes IB, et al. Brown tumors simulating metastases on FDG
hyperfunctioning parathyroid glands in primary or secondary PET in a patient with parathyroid carcinoma. Eur J Nucl Med Mol
hyperparathyroidism: influence of thyroid anomalies. Medicine Imaging. 2008;35:850.
Therapy of Neuroendocrine Tumors 29
Duccio Volterrani, Lisa Bodei, and Federica Guidoccio
Contents
29.2 Rationale of Radionuclide Imaging for Neuroendocrine Tumors 768
29.3 Somatostatin Receptor (SSTR) Imaging 769
29.3.1 111
In-Pentetreotide Scintigraphy 769
29.3.2 PET/CT with 68Ga-DOTA-Peptides 772
29.4 Imaging of Neuramine Uptake 774
29.4.1 MIBG Scintigraphy 774
29.4.2 PET/CT with 18F-DOPA 776
29.4.3 PET/CT with [18F]FDG 779
29.5 Targeted Radionuclide Therapy of NETs 780
29.5.1 Peptide Receptor Radionuclide Therapy 780
29.5.2 [131I]MIBG Therapy 782
References 783
Learning Objectives • Understand the principles of image interpretation for

• Acquire the basic pathophysiologic and clinical notions nuclear medicine imaging of neuroendocrine tumors.
on neuroendocrine tumors as they relate to the principles • Acquire the basic notions on targeted radionuclide ther-
of radionuclide targeting. apy of neuroendocrine tumors.
• Learn the characteristics of somatostatin receptor imag- • Understand the principles and the rationale of MIBG
ing and neuroamine uptake imaging. therapy and peptide receptor radionuclide therapy
• Learn the properties of radiotracers used for neuroendo- (90Y-DOTA-TOC and 177Lu-DOTA-TATE).
crine tumors and the rationale to their use in different
clinical settings (GEP-NETs, pulmonary NETs, paragan-
gliomas, medullary thyroid carcinoma). 29.1 Introduction
Neuroendocrine tumors (NETs) constitute a heterogeneous

group of rare tumors that arise from cells of the neuroendo-
crine system. The term “neuroendocrine” relates to the abil-
D. Volterrani (*) · F. Guidoccio
Regional Center of Nuclear Medicine, Department of Translational ity of these cells to produce and store amines and peptide
Research and Advanced Technologies in Medicine and Surgery, hormones produced by both the endocrine system and the
University of Pisa, Pisa, Italy nervous systems: gastrin, insulin, serotonin, somatostatin,
glucagon, pancreatic polypeptide, vasoactive intestinal pep-
L. Bodei tide (VIP), catecholamines, ACTH, GH, prolactin, FSH, LH,
Molecular Imaging and Therapy Service, Department of
TSH, and PTH. Although NETs differ widely in their biol-
Radiology, Memorial Sloan Kettering Cancer Center,
New York, NY, USA ogy and clinical presentation, they share the capability to

produce certain biological compounds (chromogranins and tinct clinical syndrome, symptoms go frequently unrecog-
synaptophysin) that are considered specific markers of neu- nized, and diagnosis is often delayed until the tumor has
roendocrine cells [1]. advanced to reach the metastatic phase.
Moreover, NETs frequently express a high density of Several diagnostic imaging modalities can be used for
somatostatin receptors (SSTRs) on their cell membranes. localizing and assessing the extent of NETs, including X-ray
Somatostatin, a small cyclic peptide which circulates in the computed tomography (CT), magnetic resonance imaging
blood in two biologically active forms (somatostatin-14 and (MRI), ultrasonography (US), contrast-enhanced US
somatostatin-28), plays an important role in modulating neu- (CEUS), endoscopic US (EUS), intraoperative US (IOUS),
rotransmission and secretion and in controlling proliferation selective angiography with hormonal sampling, and radionu-
of both normal and tumor cells. Six SSTR subtypes have clide imaging. Nevertheless, no single imaging technique
been identified by molecular analysis (SSTR1, SSTR2a, represents a real gold standard, and specific sequences of
SSTR2b, SSTR3, SSTR4, and SSTR5), each one exerting its exams are needed for each tumor type. It should also be
action by inhibiting adenylyl cyclase activity. The predomi- emphasized that up to 50% of metastatic NETs remain with
nant expression of SSTR2 on the neuroendocrine cells of the an unknown primary site.
pancreas (α cells), gastrointestinal tract, and adrenals is the Radionuclide imaging can disclose enhanced expression
basis for the successful clinical application of somatostatin of specific targets and/or functions of NETs that are charac-
analogs in molecular imaging and in therapy [1]. terized by, e.g., abundant cell-membrane-bound somatosta-
Although most NETs are sporadic, they can also be part tin receptors and/or the presence of neuroamine uptake
of multiple endocrine neoplasia types 1 and 2 (MEN1 and mechanisms. Moreover, since high glucose metabolism of
MEN2). Germline mutations in at least five genes have been tumors is directly related with cell proliferation (depending
recognized to be responsible for familial pheochromocyto- on the genetic expression of glucose transporters), increased
mas: the von Hippel-Lindau gene (VHL) causing von uptake of [18F]FDG by NET cells is typically related to
Hippel-Lindau syndrome, the RET gene associated with enhanced cell dedifferentiation and high malignancy. On the
MEN2, the neurofibromatosis type 1 gene (NF1) associated other hand, similarly as for other tumors, [18F]FDG PET
with von Recklinghausen’s disease, and the genes encoding does not invariably provide useful information for most of
the B and D subunits of mitochondrial succinate dehydroge- NETs that are relatively slow growing and therefore biologi-
nase (SDHB and SDHD), which are associated with familial cally less aggressive.
paragangliomas and pheochromocytomas.
According to an anatomic “surgical” criterion, NETs can
be divided into those arising in neuroendocrine organs (such
as medullary thyroid cancer, pancreatic endocrine tumors, Key Learning Points
pheochromocytomas, and paragangliomas), those arising • NETs are rare tumors that arise from cells of the
from dispersed neuroendocrine cells (such as pulmonary or neuroendocrine system.
gastroenteropancreatic NETs), and finally those arising from • These tumors share the capability either to produce
non-neuroendocrine organs, such as thymic or cutaneous and store amines and peptide hormones or to
NETs. express a high density of somatostatin receptors on
According to clinical presentation, NETs can be divided cell membranes.
into gastroenteropancreatic NETs (GEP-NETs), pulmonary • Because of an overall indolent pattern of growth,
NETs, and paragangliomas. most NETs are diagnosed in an advanced stage,
when the presence of distant metastases renders
radical treatment with curative intents unfeasible.
29.2 R
ationale of Radionuclide Imaging • Specific features of NETs (either expression of
for Neuroendocrine Tumors somatostatin receptors and/or the expression of
neuroamine uptake mechanisms) constitute the
Since NETs, and particularly GEP-NETs, are in general basis for radionuclide imaging using specific
well-differentiated and slow growing (only a minority of molecular targeting agents.
them exhibiting an aggressive pattern of growth), they are • While [18F]FDG PET is of little value in case of
often diagnosed when they have already metastasized, usu- slow-growing NETs, increased [18F]FDG uptake is
ally to the liver; in this stage, radical treatment with curative observed in case of tumors taking on an aggressive
intents is no longer possible. Even when the tumors release pattern of growth.
biologically active compounds, despite the presence of a dis-
29 Hybrid Imaging and Radionuclide Therapy of Neuroendocrine Tumors 769
29.3 S
omatostatin Receptor (SSTR) improved tumor-to-background ratios following pretreat-
Imaging ment with nonradioactive octreotide; this issue is therefore
not definitely clarified. The time interval between discontin-
Radionuclide imaging of NETs involves the use of radiola- uation of therapy and 111In-pentetreotide scintigraphy
beled analogs of somatostatin; these analogs have been depends on the type of drugs used (1 day for short-lived mol-
developed because the very short biological half-life in ecules, 3–4 weeks for long-acting analogs).
plasma of native somatostatin precludes efficient tumor tar- Administration of a mild oral laxative is advised on the
geting—for either diagnostic or therapeutic purposes. The day before injection and continued throughout the days of
somatostatin analog octreotide (Fig. 29.1) has the highest imaging to minimize the potential for visualizing artifacts in
affinity for SSTR2a–SSTR2b (Kd 0.1–1 nM) and is therefore the intestine when abdominal lesions are suspected.
particularly suitable for imaging, since the subtype 2 recep-
tor is most frequently expressed in NET cells. SSTR density 29.3.1.2 Image Acquisition
is markedly elevated in malignant tissues, from 80 to The radiopharmaceutical should be administered using an
2000 fmol/mg protein, whereas SSTR expression in normal indwelling i.v. catheter. The recommended activity to obtain
neuroendocrine tissues is relatively low [2]. good quality imaging is about 200 MBq, although activities
reported in the literature range from 120 to 220 MBq.
A large FOV gamma camera fitted with medium-energy,
29.3.1 111In-Pentetreotide Scintigraphy parallel-hole collimators is required. Dual peak energy
acquisitions with 20% windows are generally used (111In
123
I-[Tyr3]-octreotide was the first radiopharmaceutical used to photopeaks, 172 and 245 keV).
this purpose, but its fast deiodination in vivo resulted in subop- Planar and SPECT images should be acquired at 4 and
timal imaging properties. The more metabolically stable 24 h postinjection. Four-hour images benefit from a lower
111
In-[DTPA-d-Phe]-octreotide (or 111In-pentetreotide) has bowel activity, but radiopharmaceutical concentration at the
subsequently been developed and registered by Mallinckrodt sites of disease usually becomes significant later; for this rea-
Inc. with the trade name of OctreoScan®. This radiopharma- son, it is important to acquire two sets of images, with at
ceutical binds with high affinity to the SST2 receptors, while least one SPECT acquisition.
it has moderate affinity for the SST3 and SST5 receptors.
1. An anterior and posterior whole-body scan (scanning

29.3.1.1 Patient Preparation speed 3–7 cm/min) should be acquired at 4 h postinjec-
Some authors recommend discontinuation of therapy with tion, followed by the acquisition of anterior and posterior
“cold” octreotide therapy (when possible and not contraindi- spot images of the head/neck and/or chest and/or abdo-
cated by clinical conditions) to avoid possible saturation of men/pelvis and/or lower extremities, depending on the
somatostatin receptors. However, some reports describe even clinical indication. Spot images increase the relatively
low sensitivity for lesion detection achievable with the
octreoscan whole-body acquisition. Planar spot images should be
acquired for at least 15 min per view, using a large FOV
pentetreotide and a 128 × 128 matrix. SPECT acquisitions should be
performed over a 360° of rotation (circular or elliptical, or
octreotide
with body contouring), with 120 angular views, 30–45 s/
view, and a 128 × 128 acquisition matrix. SPECT scans
111ln DTPA D-
Phe Cys Phe are preferably performed 24 h after radiopharmaceutical
D-
injection, although the acquisition at 4 h over the abdo-
S Trp men/pelvis should be considered to obtain images with
the lowest background activity within the bowel and to
S minimize the potential for visualizing artifacts (Figs. 29.2
and 29.3).
Thr- Cys Thr 2. SPECT/CT imaging can be used for attenuation correc-
ol
tion and may improve the localization of SSTR-expressing
lesions [3, 4] (Fig. 29.4).
3. Sometimes additional acquisition at 48 h is indicated
Fig. 29.1 Schematic structure of 111
In-DTPA-octreotide
( In-pentetreotide or OctreoScan ). Octreotide, a small peptide consti-
111 ® when the 24 h scan shows accumulation in the abdomen
tuted by eight amino acids, is labeled with 111In by coupling with DTPA, of uncertain interpretation, which may represent radioac-
a chelating agent tive bowel content.
Fig. 29.2 Scintigraphy with 4h

OctreoScan®. Normal
biodistribution at 4 and 24 h:
blood pool, thyroid, kidneys
and bladder, liver, spleen, and
bowel uptake. In the 24 h
image, background activity is
reduced versus the earlier
imaging time point, although
nonspecific accumulation in 24h
the bowel increases
4h
24h
4h
anterior posterior anterior posterior
29.3.1.3 Image Interpretation • Variable degrees of tumor differentiation, with conse-

The final report of scintigraphy with 111
In-pentetreotide quent heterogeneous SSTR expression throughout differ-
should take into account: ent metastatic lesions in the same patient.
• Insulinomas are often small lesions not expressing
• Physiologic tissue distribution and timing (111In- SSTR2 [5].
pentetreotide is rapidly cleared from the blood, excretion • Liver metastases are sometimes not detected because
is almost entirely through the kidneys, and hepatobiliary SSTR expression by the tumor cells is isointense to that of
excretion accounts for 2% of the total administered normal liver cells.
activity).
• SSTRs are physiologically expressed in different organs In summary, since 111In-pentetreotide uptake is not specific
which may be imaged during SSTR scintigraphy, includ- for tumors, positive findings on the scan require consideration
ing the liver, spleen, pituitary, and thyroid. of the possibility that other diseases characterized by high
• Other sites that are visible on the scan as a result of 111In- local somatostatin receptor concentrations may be present.
pentetreotide clearance include the kidneys, renal collect-
ing system, ureters, urinary bladder, gallbladder, and 29.3.1.4 Clinical Indications
bowel. The main indication for 111In-pentetreotide scintigraphy is
imaging of GEP-NETs and pulmonary NETs. Less frequent
Anatomical localization of tumor-related indications include NETs originating from the skin, sympa-
111
In-pentetreotide uptake according to other imaging data thoadrenal system (paragangliomas), thyroid (medullary car-
and the intensity of uptake should also be noted. cinoma), and genital tract. In the management of patients
Semiquantitative assessment of the lesions with tumor-to- with NETs, 111In-pentetreotide scintigraphy can be used to
background ratio may be useful in selected cases. localize primary tumors, to detect sites of metastatic disease
Several potential pitfalls should be considered when (staging), and to detect residual, recurrent, or progressive
interpreting 111In-pentetreotide scintigraphy, as follows: disease (restaging). SSTR scintigraphy with
111
In-pentetreotide is also used to select potentially respon-
• Possible tracer accumulation in areas of recent surgery. sive patients for treatment with unlabeled octapeptide soma-
34 35 36 37 38
39 40 41 42 43
Head to Feet Transversal Slice thickness 9,04 mm
29 30 31 32 33
34 35 36 37 38
Right to Left Sagittal Slice thickeness 9,04 mm
16 17 18 19 20
21 22 23 24 25
Anterior to Posterior Coronal Slice thickeness 9,04 mm
Fig. 29.3 Transaxial, coronal, and sagittal SPECT sections from a Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
study with OctreoScan® in a patient with intestinal neuroendocrine car- Applicazioni. Milan: Springer Italy; 2010)
cinoma (reproduced with permission from: Volterrani D, Erba PA,
tostatin analogs, as well as with tumor-targeting radiolabeled is associated with a higher probability of response to PRRNT
somatostatin analogs (the so-called peptide receptor radionu- [6, 7].
clide therapy or PRRNT—see further below in this chapter). [99mTc-EDDA-HYNIC-d-Phe1,Tyr3]octreotide (or 99mTc-
In fact, intense uptake of 111In-pentetreotide in tumor lesions EDDA/HYNIC-TOC), a novel agent for SSTR scintigraphy,
Fig. 29.4 Transaxial slice of

a SPECT/CT study with
OctreoScan® (CT component
in upper left panel, SPECT
component in upper right
panel, fused SPECT/CT
image in lower panel)
revealing a focal area of
increased uptake within the
right iliac fossa in a patient
with appendicular carcinoid
CT Transaxials NM Transaxials
is commercially available in some European countries. The

suggested activity range is 370–740 MBq, and images are
acquired at 1–2 and 4 h after i.v. administration. More favor-
P able distribution pattern, higher tumor-non-tumor ratios,
lower radiation dose to patients, and better spatial resolution
are reported as the main advantages of this 99mTc-labeled
T compound over 111In-pentetreotide.
L 29.3.2 PET/CT with 68Ga-DOTA-Peptides

S
A
Somatostatin analogs labeled with the positron emitter 68Ga
K have been developed, allowing the possibility to perform SSTR
K
l
PET; the pattern of physiologic distribution of such NET imag-
PH ing agents is shown in Fig. 29.5. With respect to single-photon
gamma camera imaging, this novel imaging approach is char-
B
acterized by significantly better spatial resolution (4–5 mm ver-
sus 8–10), higher signal-to-noise ratios within a quite shorter
time after administration, easier image quantification, and
lower radiation dose to patients [8, 9] (Figs. 29.6 and 29.7).
Fig. 29.5 Normal distribution of 68Ga-DOTA-TOC (MIP, maximum The 68Ga-peptides most commonly used in clinical practice
intensity projection image) with physiological visualization of the pitu-
itary (P), thyroid (T), liver (L), adrenals (A), kidneys (K), spleen (S),
are based on radiolabeling obtained with the metal chelating
and pancreatic head (PH) and elimination of the activity in the urinary agent 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic
bladder (B) and intestine (I) (reproduced with permission from Bodei L, acid (or DOTA). Among several different DOTA-octapeptides
Kidd M, Gilardi L, Volterrani D, Paganelli G, Grana CM, Modlin that have been investigated for targeting of NETs, the most
IM. Diagnostic applications of nuclear medicine: neuroendocrine
tumours. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
commonly employed are DOTA-Tyr3 octreotide (DOTA-
Nuclear Oncology – From Pathophysiology to Clinical Applications, TOC), DOTA-Thr8 octreotide (DOTA-TATE), and DOTA-1-
vol 2. New York: Springer; 2017: 799–838) Nal3 octreotide (DOTA-NOC). Their SSTR affinity profiles
these kits is expected to take place in the near future, and

certainly it will promote a more widespread use of SSTR
PET/CT imaging in patients with confirmed or suspected
well-differentiated GEP-NETs.
29.3.2.1 Patient Preparation

The same considerations listed above for “cold” octreotide
therapy as it relates to 111In-pentetreotide scintigraphy apply
for PET/CT with 68Ga-DOTA-peptides.
29.3.2.2 Image Acquisition and Interpretation

The activity administered ranges from 100 to 200 MBq, and
scans are usually acquired between 45 and 90 min after i.v.
administration. The field of view usually extends from the
base of the skull to the mid-thighs, although whole-body
acquisition can be performed, depending on the clinical set-
ting. Due to technical differences among the different PET/
CT scanners, there is no univocal protocol for acquisition
times and parameters.
The pattern of physiological distribution of 68Ga-DOTA-
peptides is similar to that of 111In-pentetreotide. The images
should be analyzed considering sites of anatomical localiza-
tion of radiotracer uptake according to the paired CT infor-
mation (or other imaging data, e.g., if PET/MR is performed).
SUV can be used to assess the intensity of 68Ga-DOTA-
peptide uptake for tumor characterization.
Fig. 29.6 Transaxial CT and fused PET/CT slice of a PET/CT scan with Key Learning Points
68
Ga-DOTA-peptide localizing an ileal carcinoid tumor (left of midline) • Imaging of NETs involves the use of radiolabeled
with likely metastasis in a mesenteric lymph node (right of midline)
(images provided by courtesy of Prof. Stefano Fanti, Nuclear Medicine analogs of somatostatin.
Service, “Malpighi – S. Orsola” University Hospital, Bologna, Italy) • OctreoScan® binds with high affinity to the SST2
receptors, while it has moderate affinity for the
SST3 and SST5 receptors.
have been characterized as follows: -TOC and -TATE have high • The main indication for OctreoScan® is the imaging
binding affinity to SSTR2, while -NOC has high affinity for of GEP and pulmonary NETs. Less frequent indica-
SSTR2, SSTR3, and SSTR5 [10]. Despite such differences in tion includes NETs originating from the skin, sym-
binding affinity profiles, the few studies so far performed compar- pathoadrenal system (paragangliomas), thyroid
ing different 68Ga DOTA-octapeptides in the same patients have (medullary carcinoma), and genital tract.
reported comparable diagnostic accuracy for all three tracers. • SSTR scintigraphy can be used to localize primary
Nevertheless, 68Ga-DOTA-TOC and 68Ga-DOTA-TATE have tumors, detect sites of metastatic disease (staging),
become the new standard somatostatin analogs for imaging of and detect residual, recurrent, or progressive dis-
NETs, since they mimic very closely the in vivo distribution pat- ease (restaging).
tern of their 90Y- or 177Lu-labeled counterparts used for PRRNT. • SSTR scintigraphy with 111In-pentetreotide is also
So far, 68Ga-DOTA-peptides have been available as in- used to select potentially responsive patients for
house preparations, provided they are prepared fulfilling all treatment with unlabeled octapeptide somatostatin
the strict criteria laid down in good radiopharmaceutical analogs, as well as with tumor-targeted radiolabeled
practice. However, two kits for the preparation of 68Ga-DOTA- somatostatin analogs.
peptides (SomaKit-TATE and SomaKit-TOC) have very • Two different kits for 68Ga-DOTA-peptide prepara-
recently been approved by the US Food and Drug tion, SomaKit-TATE and SomaKit-TOC, have been
Administration (FDA) and by the European Medicines recently approved by the FDA and EMA.
Agency (EMA), respectively. Commercial availability of
Fig. 29.7 PET/CT shows liver metastases from a primary pancreatic Stefano Fanti, Nuclear Medicine Service, “Malpighi – S. Orsola”
NET (centered by a red cross in lower right panel) with increased University Hospital, Bologna, Italy)
uptake of 68Ga-DOTA-peptide (images provided by courtesy of Prof.
29.4 Imaging of Neuramine Uptake tracers that target the catecholamine pathway, such as [11C]
ephedrine, [11C]hydroxyephedrine, 18F-dopamine, 18F-l-
A biological feature of some NETs is their ability to take up dihydroxyphenylalanine (18F-DOPA), and radioiodinated
amino acids and transform them into biogenic amines meta-iodobenzylguanidine (MIBG), which is a catechol-
through the decarboxylation pathway. Due to this peculiar amine analog taken up by the cells through the norepineph-
metabolic pattern, NETs have been in the past denominated rine transporters and stored in the synaptic vesicles.
also as APUD tumors (the acronym for “amine precursor
uptake and decarboxylation”). Through a sequential enzy-
matic pathway, which begins with the conversion of tyrosine 29.4.1 MIBG Scintigraphy
into l-DOPA (which is then decarboxylated to dopamine,
oxydated to norepinephrine, and methylated to yield epi- MIBG has historically been introduced as the first option for
nephrine), NET cells can synthesize catecholamines that are radionuclide imaging of NETs. Although both [123I]MIBG
transported and stored into the synaptic vesicles by means of and [131I]MIBG are approved for diagnostic use in the
the vesicular transporters for catecholamines (VMAT1 and European Union and in the United States, the diagnostic per-
VMAT2). After their release in the synaptic space, reuptake formance of [123I]MIBG is clearly better than that of [131I]
occurs by norepinephrine transporters located in the presyn- MIBG [10, 11], mostly because of the superior imaging qual-
aptic membrane. Thus, several NETs can be imaged with ity due to the better spatial resolution provided by the 159 keV
γ-rays of 123I and to the higher signal-to-noise ratio achievable ful to improve the anatomic localization of MIBG avid lesions
because of the greater injected activity of [123I]MIBG. The and for attenuation correction.
higher count rate made possible by the administration of [123I]
MIBG also enables to easily perform SPECT and SPECT/CT 29.4.1.3 Image Interpretation
acquisitions. Moreover, the patient’s effective dose due to MIBG physiologically accumulates in the liver, spleen,
[123I]MIBG is significantly lower than that due to [131I]MIBG. lungs, heart, and salivary glands. Tracer uptake can be
detected also in normal adrenal glands at late imaging times.
29.4.1.1 Patient Preparation Most of the administered MIBG is excreted through the kid-
1. Before MIBG scintigraphy, whenever clinically possible, neys, with minor fecal excretion. In some patients with para-
administration of drugs that can interfere with the specific gangliomas, physiologic uptake in the heart can be very low
MIBG uptake mechanism must be discontinued (usually due to downregulation of norepinephrine transporters in
1–3 days before imaging), including both cardiovascular response to increased serum levels of circulating catechol-
and sympathomimetic drugs (such as anti-arrhythmics for amines (Fig. 29.8).
ventricular arrhythmias, combined α- and β-blockers, High and/or inhomogeneous MIBG uptake within an
adrenergic neuron blockers, α-blockers, calcium channel enlarged adrenal gland is considered an abnormal finding
blockers, inotropic sympathomimetic drugs, vasocon- consistent with an adrenal paraganglioma. Extra-adrenal foci
strictor sympathomimetics, β-2 stimulants, and systemic of uptake that cannot be explained by normal physiological
and local nasal decongestants, sympathomimetics for distribution should be considered abnormal findings. SPECT/
glaucoma) and neurological drugs (such as antipsychot- CT or correlative imaging improves anatomical localization
ics, sedating antihistamines, opioid analgesics, tricyclic and helps toward the final diagnosis.
antidepressants, tricyclic-related antidepressants, and Suboptimal sensitivity can be due to small size of lesions
CNS stimulants) [12]. and/or to their location. In particular, sensitivity can be
2. Thyroid blockade with Lugol’s iodine (in an amount
reduced in case of extra-adrenal lesions close to sites of
equivalent to 100 mg of iodine) should be started 1 day physiological uptake/excretion. On the other hand, foci of
before tracer injection and continued for 2 days when
using [123I]MIBG. Alternatively, potassium perchlorate
can be used starting 4 h before tracer injection and contin-
ued thereafter for 2 days (400–600 mg/day).
The recommended activity in adults is 200–400 MBq for

[123I]MIBG. In order to minimize possible acute side effects
(such as tachycardia, pallor, vomiting, or abdominal pain),
MIBG should be administered by slow i.v. injection.
29.4.1.2 Image Acquisition

A large FOV gamma camera fitted with low-energy high-
resolution, parallel-hole collimators is required. However,
several centers prefer medium-energy collimators because
they reduce septal penetration of the high-energy photons
that are part of the 123I decay. The energy peak should be set
to 159 keV, with a 20% window.
Planar and SPECT images should be acquired at least at
24 h postinjection, with optional sets of images being acquired
at 4–6 h after injection. An anterior and posterior whole-body
scan (scanning speed of 5–7 cm/min) should be performed,
followed by the acquisition of anterior and posterior spot
images of the head/neck and/or chest and/or abdomen/pelvis
and/or lower extremities, depending on the clinical indica-
tion. As for 111In-pentetreotide, spot images enable to increase
sensitivity over the whole-body acquisition. Planar spot
ANT POST
images should be acquired for 10 min per view, using a large
FOV and a 128 × 128 matrix. SPECT acquisitions should be Fig. 29.8 A typical distribution of [123I]MIBG (4 h post injection) which
performed over a 360° rotation (circular or elliptical, or with includes clear visualization of the salivary glands, lungs, liver, spleen, and
body contouring), with 120 views, time per view of 30 s, and urinary collecting system. Due to the presence of a high-uptake adrenal pheo-
a 128 × 128 acquisition matrix. SPECT/CT is especially use- chromocytoma, no significant myocardial uptake is visualised in this patient
physiologic accumulation of MIBG within the urinary tract 29.4.2 PET/CT with 18F-DOPA
or bowel can cause false-positive results.
PET/CT with 18F-DOPA is being increasingly employed
29.4.1.4 Indications because of several advantages over MIBG scintigraphy, par-
MIBG scintigraphy is frequently used in patients with para- ticularly higher spatial resolution and low background activ-
gangliomas; NETs that arise from cells are derived from the ity allowing the detection of smaller adrenal and extra-adrenal
neural crest stem. Paragangliomas may arise anywhere along lesions (Fig. 29.11). 18F-DOPA is currently commercially
the paraganglial system and can be associated with the sympa- available in the European Union, but not in the United States.
thetic nervous system (adrenal medulla, the organ of PET/CT can be performed shortly after the administration of
Zuckerkandl, or other chromaffin cells) or the parasympa- 18
F-DOPA, as opposed to waiting at least 24 h as necessary
thetic nervous system (chemoreceptors mainly located in the for MIBG scintigraphy. In addition, thyroid blocking and dis-
head and neck). A whole-body MIBG scan is particularly continuation of medications are not necessary for 18F-DOPA
helpful in the preoperative identification of multiple primary PET. While VMAT 1 expression is necessary for MIBG
lesions, especially occurring in patients with a familial syn- uptake, both VMAT 1 and VMAT 2 seem to be involved in
drome or of metastases from malignant tumors (Figs. 29.9 and 18
F-DOPA uptake. The predominant VMAT 2 expression in
29.10). MIBG scintigraphy is also useful to detect small tumor head and neck parasympathetic paraganglia might explain the
recurrences, because the radiotracer accumulates specifically better performance of 18F-DOPA PET/CT in paragangliomas,
within the tumor and is not affected by postsurgical or post- usually derived from parasympathetic ganglia [13, 14].
radiotherapy changes.
MIBG imaging provides valuable information when plan- 29.4.2.1 Patient Preparation
ning targeted radionuclide therapy with radiolabeled Patients should fast for at least 4 h before 18F-DOPA admin-
MIBG—by confirming uptake and aiding for personalized istration. There is no univocal consensus concerning pre-
dosimetric evaluation. MIBG imaging is also helpful for treatment with 200 mg of carbidopa 1 h prior to 18F-DOPA
assessing metabolic tumor response to treatment in patients injection—aimed at increasing tumor uptake.
with inoperable metastatic paragangliomas.
The sensitivity of MIBG scintigraphy is somewhat lower 29.4.2.2 Image Acquisition and Interpretation
for parasympathetic paragangliomas in the head and neck Scans are usually acquired between 30 and 60 min after
region when compared with both SSTR scintigraphy and tracer injection from the base of the skull to the mid-thighs,
18
F-DOPA PET/CT imaging. although whole-body acquisition can be performed accord-
Fig. 29.9 [123I]MIBG planar

whole-body scan revealing
multiple bone lesions from
malignant pheochromocytoma
NM Coronals NM Sagittals NM Transaxials
MIP Navigate
Fused Coronals Fused Sagittals Fused Transaxials
Fig. 29.10 [123I]MIBG SPECT/CT shows multiple focal sites of tracer D, Paganelli G, Grana CM, Modlin IM. Diagnostic applications of
uptake within the abdomen consistent with tumor seeding from a pheo- nuclear medicine: neuroendocrine tumours. In: Strauss HW, Mariani G,
chromocytoma of the right adrenal ruptured during surgery 7 years earlier Volterrani D, Larson SM, Eds. Nuclear Oncology – From Pathophysiology
(reproduced with permission from: Bodei L, Kidd M, Gilardi L, Volterrani to Clinical Applications, vol 2. New York: Springer; 2017: 799–838)
Fig. 29.11 PET/CT with

18
F-DOPA. Two small foci of
uptake close to the right
external ear consistent with
local recurrent paraganglioma
ing to the clinical indication. Physiological uptake can be SUV and metabolic tumor burden (MTB, calculated as
observed within the striata in the brain, kidneys, pancreas, the product of metabolic tumor volume and SUVmean) can be
liver, gallbladder, biliary tract, and duodenum. Also the adre- used to quantify 18F-DOPA uptake in tumor lesions.
nal glands can physiologically show a slight tracer uptake.
Any high uptake within an enlarged adrenal and any non- 29.4.2.3 Indications
physiological extra-adrenal focal uptake should be consid- PET/CT with 18F-DOPA is highly sensitive for detecting
ered consistent with the presence of a paraganglioma. adrenal and extra-adrenal paragangliomas, constituting an
Fig. 29.12 PET/CT with 18F-DOPA in a patient with multiple paragangliomas of the head and neck and thoracic districts
excellent first-line imaging option especially in head and

neck paragangliomas (Fig. 29.12). 18F-DOPA PET/CT may Key Learning Points
also detect residual medullary thyroid carcinoma lesions • The diagnostic accuracy of [123I]MIBG is clearly
postsurgery and other syndromic tumors such as pancreatic better than that of [131I]MIBG due to the better spa-
neuroendocrine tumors that may develop in patients with tial resolution provided by the 159 keV γ-rays of
VHL syndrome. 18F-DOPA PET/CT can be used for detect- 123
I and to the higher signal-to-noise ratio achiev-
ing metastatic disease in patients with malignant paragan- able as a consequence of the greater administered
gliomas, although scarce and non-univocal data are activity of [123I]MIBG.
available on metastatic and recurrent forms (Fig. 29.13). • MIBG scintigraphy is frequently used for imaging
In patients with GEP-NETs, 18F-DOPA PET/CT is cur- paragangliomas.
rently recommended mainly in the case of insulinomas, for • The sensitivity of MIBG scintigraphy results is lower
preoperative localization in the case of congenital hyperinsu- for parasympathetic paragangliomas in the head and
linism [15–17]. In fact, the ability to take up DOPA and to neck region when compared with both SSTR scintig-
convert it into dopamine is increased in the hyperfunctioning raphy and 18F-DOPA PET/CT imaging.
of the affected pancreatic tissue in comparison with normally • 18F-DOPA PET/CT is highly sensitive for detecting
functioning pancreas, which exhibits low levels of aromatic adrenal and extra-adrenal paragangliomas, being an
amino acid decarboxylation. Conversely, in adult subjects excellent first-line imaging tool especially in head
pancreatic uptake of 18F-DOPA is not statistically different and neck paragangliomas.
between normal subjects and hyperinsulinemic patients, and • In GEP-NETs, 18F-DOPA PET/CT is currently rec-
the main limitation for identifying insulinomas or β-cell ommended for preoperative localization of insulin-
hyperplasia is the relatively high uptake of 18F-DOPA in the omas in patients with congenital hyperinsulinism.
whole pancreas.
Fig. 29.13 MIP of PET/CT

with 18F-DOPA before and
after treatment with MIBG in
a patient with a metastatic
paraganglioma, revealing
partial response after
treatment
23/4/2015 30/11/2015
29.4.3 PET/CT with [18F]FDG
Changes in tumor biology can be characterized by modifica- SSTR GLUT

differentiation
tions in receptors’ or transporters’ expression on cell mem-
aggressivity
branes. While well-differentiated NETs do not overexpress
glucose transporters (GLUT) and overexpress SSTRs, poorly GLUT SSTR
differentiated tumors may show a decline in SSTR density
associated with increased GLUT expression (Fig. 29.14).
Thus, GEP-NETs with a positive [18F]FDG PET are gen-
erally more aggressive irrespective of the classical markers Fig. 29.14 Diagrammatic representation of the inverse correlation
of aggressive pattern. Therefore, [18F]FDG PET may help in between SSTR expression (an index of cell differentiation generally
stratifying patients according to their risk for proper therapy associated with relatively low biologic aggressivity) and GLUT expres-
planning [18, 19]. sion (an index of cell dedifferentiation almost invariably associated
with high biologic aggressivity) in NETs. This phenomenon is similar
The usefulness of [18F]FDG PET/CT for pulmonary NETs to the so-called “flip-flop” phenomenon that can be observed in patients
is still debated. In fact, these tumors show variable [18F]FDG with differentiated thyroid cancer regarding the inverse correlation
uptake, depending on mitotic index and tumor proliferation between NIS expression (in more differentiated, less aggressive tumors)
rate. Furthermore, PET/CT seems to perform better than CT and GLUT expression
in detecting distant metastases, thus leading to changes in
clinical management of patients. approaching 100%, [18F]FDG PET is the preferred functional
Typical carcinoids that usually overexpress SSTRs dis- imaging modality for staging and for treatment monitoring
play high 68Ga-DOTA-TOC and low [18F]FDG uptake, con- of metastatic adrenal paragangliomas associated with the
sistent with their low proliferative index. It is likely that SDHB mutation [20].
avidity of [18F]FDG or 68Ga-DOTA-peptides could help to
identify, at initial staging (as well as during follow-up and
after therapy), aggressive tumors or sites of possible dedif- Key Learning Points
ferentiation within otherwise indolent tumors. • GEP-NETs with a positive [18F]FDG PET are more
Although [18F]FDG PET/CT is not recommended as a aggressive, irrespective of the classical markers of
first-line approach to localize paragangliomas, it is of special aggressive pattern.
value for localizing paragangliomas when other imaging • [18F]FDG PET could help in stratifying patients
methods have failed. In fact, most paragangliomas accumu- according to their risk, for proper therapy planning.
late [18F]FDG, especially when malignant. With sensitivity
29.5 Targeted Radionuclide Therapy TOC and 177Lu-DOTA-TATE are the most widely employed
of NETs radiopeptides and can be used in the European Union for
therapy trials. Choice of these two radiopharmaceuticals
Therapy of NETs is typically multidisciplinary and should derives from physical and pharmacokinetic considerations.
be tailored to the individual patient according to the tumor Yttrium-90 emits higher energy-range β− particles than 177Lu,
histotype and extension, as well as symptoms. Different which makes these particles more penetrating within tumor
newly proposed treatments are being suggested and applied lesions an important physical feature in the case of bulky
by various centers. The current validated therapeutic and/or lesions with inhomogeneous SSTR expression.
approaches for patients with NETs include surgery, interven- Moreover, the shorter physical half-life of 90Y results in a
tional radiology, medical treatments, chemotherapy, new tar- higher dose rate to the tumor than 177Lu. On the other hand,
geted drugs, and targeted radionuclide therapy. the lower energy and penetrating range of its β− particles
Surgery can be crucial in different phases of the disease, make 177Lu more suitable for treating smaller tumors.
from potential curative resection of the primary tumor to The pharmacokinetic profile of DOTA-TOC and DOTA-
debulking surgery for metastatic lesions (in the perspective TATE is characterized by a rapid plasma clearance after
of employing other therapies, in order to control debilitating administration, with relevant renal excretion. Since the resi-
symptoms due to overproduction of biologically active com- dence times of DOTA-TATE in the tumor and kidney are lon-
pounds) or to surgery with pure palliative intents. The main ger than those of DOTA-TOC, 177Lu appears more beneficial
limitation of surgery is the frequent presence of synchronous when labeling DOTA-TATE, while in view of the higher
metastatic disease. renal radiation burden, 90Y appears more convenient to label
Interventional radiology techniques are often used in DOTA-TOC.
patients with GEP-NETs because of their frequent metastatic
diffusion to the liver. Besides radiofrequency ablation and 29.5.1.1 Indications
high-intensity focused ultrasound ablation, radioemboliza- According to guidelines jointly issued by the International
tion of liver metastases with 90Y-labeled microspheres has Atomic Energy Agency (IAEA), the European Association
been tested in several clinical trials. of Nuclear Medicine (EANM), and the Society of Nuclear
Medical therapy is aimed at treating symptoms and/or Medicine and Molecular Imaging (SNMMI) [21], candidates
reducing tumor growth. Traditional chemotherapy has little for PRRNT are basically all patients affected by metastatic/
place in well-differentiated NETs, since most of them are unresectable SSTR2-expressing NETs; this population
slow-growing tumors. One of the basic treatments for NETs includes mainly patients with GEP-NETs and NETs of the
is biotherapy with somatostatin analogs that are generally bronchial tract and possibly also patients with pheochromo-
well tolerated; long-acting formulations are used success- cytoma, paraganglioma, neuroblastoma, or medullary thy-
fully to control tumor hypersecretion and symptoms in up to roid carcinoma—provided that such tumors express SSTR2,
70% of patients, although tachyphylaxis frequently occurs as most frequently demonstrated by radionuclide imaging
early after initiation of therapy. Objective partial responses with labeled somatostatin analogs (Fig. 29.15).
are observed in <10% of patients. SSTR scintigraphy with 111In-pentetreotide is currently
Targeted radionuclide therapy offers the potential advan- considered the most accurate and validated method to assess
tage of systemic administration with minimal toxicity to for SSTR overexpression in tumor lesions. Nevertheless,
nontarget healthy tissues. Success of treatment depends on PET/CT with either 68Ga-DOTA-TOC or 68Ga-DOTA-TATE
the specificity of the targeting carrier molecule and choice of is becoming the new standard for selecting patients for
an appropriate radioisotope label. PRRNT, since these radiolabeled analogs mimic very closely
the in vivo behavior of their 90Y- or 177Lu-labeled counter-
parts used for PRRNT (Fig. 29.16).
29.5.1 Peptide Receptor Radionuclide Therapy
29.5.1.2 Contraindications
Peptide receptor radionuclide therapy (PRRNT) is based on There are very few absolute contraindications to PRRNT
the use of radiolabeled peptides for treating unresectable or with 90Y- or 177Lu-labeled somatostatin analogs that are gen-
metastatic NETs. The rationale for such therapy is to convey erally common to all radionuclide-based therapies or treat-
radioactivity inside the tumor cells, where the sensitive tar- ments that require a certain life expectancy and/or a certain
gets, such as DNA, can be hit as a result of internalization of degree of patient’s compliance. They include pregnancy,
the somatostatin receptor and radiolabeled analog complex. severe acute concomitant illnesses, or severe unmanageable
Different peptides labeled with radiometals emitting β− psychiatric disorders.
particles suitable for therapy, such as 90Y or Lutetium-177, Also, relative contraindications to PRRNT include the
have been explored for therapeutic purposes. Y-DOTA- following conditions:
90
Fig. 29.15 Tumors candidate

to PRRT are those with an a b c d
uptake on planar images at
least equal to the one of the
normal liver (grade 1, image
b), higher than that (grade 2,
image c), or higher than the
one of the kidneys and spleen,
the “hottest” organs at
111
Inpentetreotide scintigraphy
(grade 3, image d). Patients
with tumour uptake less than
in the liver (image a) are not
good candidate to PRRT
from: Bodei L, Gilardi L,
Volterrani D, Paganelli G,
Grana CM, Kidd M, Modlin
IM. Neuroendocrine tumours:
therapy with radiolabeled
peptides. In: Strauss HW,
Mariani G, Volterrani D,
Larson SM, Eds. Nuclear
Oncology – From
Applications, vol 2. New York:
Springer; 2017: 1243–67)
Fig. 29.16 Scintigraphy with

OctreoScan® before and after
therapy with 90Y-DOTA-TOC
in a patient with metastatic
NET
pre post
• Breastfeeding (it should be discontinued). • Severely reduced bone marrow reserve, i.e., erythrocyte
• Severely reduced renal function, especially if a 90Y-labeled count <3,000,000/μL, leukocyte count <3000/μL (or
peptide is being considered for PRRNT (mild to moderate absolute neutrophil count <1000/μL), and platelet count
renal impairment—serum creatinine ≤1.7 mg/dL—is <75,000/μL for 177Lu-DOTA-TATE or <90,000/μL for
compatible with treatment). 90
Y-DOTA-TOC.
PRRNT with either 90Y-DOTA-TOC or 177Lu-DOTA- authorization was based on efficacy and safety data from a
TATE is generally well tolerated, without serious acute side phase 1/phase 2 trial conducted by Erasmus Medical Center
effects [22–24]. However, kidneys are the critical organs and on the results of a randomized pivotal phase 3 study,
(particularly after 90Y-DOTA-TOC administration), due to NETTER-1, that compared treatment using Lutathera® to a
proximal tubular reabsorption and interstitial retention of the double dose of unlabeled long-acting repeatable octreotide
radiolabeled peptides. Some decline in creatinine clearance (octreotide-LAR) in patients with inoperable midgut NETs
was reported after initial studies on 90Y-DOTA-TOC and that were progressive under treatment with standard-dose
177
Lu-DOTA-TATE therapy, especially in the case of con- octreotide-LAR and overexpressing SSTRs. The FDA is
comitant risk factors such as arterial hypertension and/or dia- scheduled to make a final approval decision within the year
betes mellitus [25]. 2018.
29.5.1.3 Recommended Protocols

and Administration Schedules 29.5.2 [131I]MIBG Therapy
The standard protocol for PRRNT includes pretreatment
renal protection, which is achieved by administration of the Radionuclide therapy with [123I]MIBG has been investigated
positively charged amino acids l-lysine and/or l-arginine— for treating different types of NETs. While this radiopharma-
with the purpose of competitively inhibiting proximal tubu- ceutical is approved for treatment in the European Union, it
lar reabsorption of the radiopeptide and thus reducing the is still considered as an investigational new drug in the
renal absorbed dose [26, 27]. Unless contraindicated by United States, where it can be used for therapy trials in
other clinical conditions (such as cardiac failure), the patient accordance with USP regulations.
is hydrated with infusion of normal saline containing the Adequate cellular uptake and retention of MIBG and the
amino acid; the protocol recommended by the IAEA/EANM/ emission from 131I of β− particles with adequate energy con-
SNMMI guidelines [21] suggests infusion over 4 h (starting stitute the rationale of this treatment approach.
30–60 min before PRRNT) of 25 g of lysine and 25 g of argi-
nine in 2 L of normal saline. Reduction of the renal absorbed 29.5.2.1 Indications
radiation dose can be achieved also by pretreatment infusion As recommended by guidelines issued by the European
of the plasma expander Gelofusine—in association with the Association of Nuclear Medicine, candidates for therapy
amino acid infusion [28]. Variations in administration proto- with [131I]MIBG include the following conditions:
cols are described in details in the abovementioned IAEA/
EANM/SNMMI guidelines [21]. • Inoperable pheochromocytoma, paraganglioma, or carci-
For PRRNT based on either 90Y-DOTA-TATE or noid tumor
90
Y-DOTA-TOC, the activity administered is usually 3.7 GB • Stage II or IV neuroblastoma
(100 mCi) per m2 of body surface, repeating a second cycle • Metastatic or recurrent medullary thyroid carcinoma
of administration after 6–12 weeks. An alternative regimen is
based on administration of 2.78–4.44 GBq (75–120 mCi) in Similarly as described above for PRRNT, therapy with
two to four cycles at 6–12 week intervals. [131I]MIBG is reserved to patients whose tumor lesions for
When using 177Lu-DOTA-TATE or 177Lu-DOTA-TOC for which treatment is intended exhibit sufficient uptake of this
PRRNT, the activity administered is usually 5.55–7.4 GBq radiopharmaceutical as demonstrated by scintigraphy with a
(150–200 mCi) in three to five cycles and 6–12 week diagnostic activity of either [123I]MIBG or [131I]MIBG—
intervals. although there is no univocal consensus on what should be
Recently, combination therapies using both 177Lu- and considered as adequate uptake.
90
Y-labeled radiopeptides have been proposed in order to
improve efficacy of treatment in patients with tumors of vari- 29.5.2.2 Contraindications
ous sizes and inhomogeneous SSTR distribution within the Absolute contraindications to therapy with [131I]MIBG
tumor lesions. These combination treatments should be per- include pregnancy, life expectancy <3 months (unless treat-
formed in centers with extensive experience with the use of ment is intended for palliation of bone pain due to skeletal
PRRNT. metastases), and impending severe renal failure for which
Although 90Y-DOTA-TOC has been extensively investi- dialysis is planned in the short term.
gated, the European Medicines Agency recently approved Relative contraindications include the following
the marketing authorization of 177Lu-DOTA-TATE conditions:
(177Lu-oxodotreotide or Lutathera®) for the treatment of
unresectable or metastatic, progressive, well-differentiated • Unacceptable medical risk for isolation
(G1 and G2), SSTR-positive GEP-NETs in adults. The • Unmanageable urinary incontinence
• Glomerular filtration rate <30 mL/min published data are limited and with few prospective trials. In
• Severe hematological and/or renal toxicities from prior these patients symptomatic and biochemical improvements
treatments are frequently observed after treatment, even though the
• Reduced bone marrow reserve (i.e., leukocyte count rates of complete response are low.
<3000/μL and/or platelet count <100,000/μL) There is currently no consensus on the optimal dosing
strategy. In patients treated with median activities of 9.9 GBq
Administered activity should be reduced, and close fol- (268 mCi) versus 5.5 GBq (149 mCi), higher activities seem
low-up should be adopted to anticipate toxicities in the case to deliver the desired dose faster with a modest increase in
of low leukocyte/platelet count or in patients with reduced toxicity but similar overall response rates [30].
renal function. Since somatostatin analogs exhibit better targeting prop-
erties, the role of [131I]MIBG therapy in patients with GEP-
29.5.2.3 Recommended Protocols NETs is limited, as a palliative option, to SSTR-negative
and Administration Schedules patients, to those with borderline renal function, and to
Preparation of patients includes discontinuation of drugs patients resistant to somatostatin therapy.
interfering with MIBG uptake, as detailed previously in this In the case of neuroblastomas, most studies have used
chapter when describing preparation for MIBG scintigraphy. [131I]MIBG with therapeutic intent; good responses have
There is a wide range of drugs that potentially interfere with been obtained in selected patients with advanced disease in
MIBG uptake, as follows: whom first-line therapy failed. The response rates in these
studies using single doses or large cumulative doses of [131I]
• Drugs for ventricular arrhythmias and sympathomimetic MIBG were between 25% and 46% [31–34]. The actual clin-
drugs ical benefit of multiple treatments is still debated.
• α-Blockers
• Calcium channel blockers
• Inotropic sympathomimetic drugs Key Learning Points
• Vasoconstrictor sympathomimetic drugs • Tumor candidates for PRRNT with radiolabeled
• β2 Stimulants (sympathomimetic drugs) somatostatin analogs are basically all SSTR2-
• Adrenoreceptor stimulant (orciprenaline) expressing NETs.
• Systemic and local nasal decongestants • SSTR scintigraphy with 111In-pentetreotide is cur-
• Sympathomimetic drugs for glaucoma rently the most accurate and validated method to
• Antipsychotic drugs assess for the presence of SSTR overexpression.
• Sedating antihistamine drugs • 90Y-DOTA-TOC and 177Lu-DOTA-TATE are the
• Opioid analgesic drugs most widely employed radiopeptides for therapy.
• Tricyclic antidepressant drugs • Kidneys are the critical organs, particularly after
• Stimulants of the central nervous system administration of 90Y-DOTA-TOC.
• The European Medicines Agency recently approved
Thyroid blockade using potassium iodide, also in combi- the marketing authorization of 177Lu-DOTA-TATE
nation with potassium perchlorate, is necessary, generally (Lutathera®) for the treatment of unresectable or
starting 1 or 2 days prior to treatment and lasting over 14 days metastatic, progressive, well-differentiated (G1 and
after treatment. G2), SSTR-positive GEP-NETs in adults.
Slow i.v. infusion of [131I]MIBG (suggested range from • Unresectable adrenal and extra-adrenal paragangli-
45 min to 4 h) is mandatory in order to minimize the pharma- omas, or carcinoids, advanced neuroblastomas, and
cologic mass effects of MIBG. For the same reason, the spe- metastatic medullary thyroid carcinomas are the
cific activity of [131I]MIBG for therapy should be higher than tumors that are most frequently treated with [131I]
that used for diagnostic purpose (up to 1.48 GBq/mg) [29], MIBG therapy.
in order to minimize the incidence of pharmacologic mass
effects of MIBG, such as nausea, vomiting, and hyperten-
sion. Vital signs (heart rate, blood pressure) must be moni-
tored during therapy administration and over the next References
2–3 days, especially in the case of catecholamine-secreting
tumors and administering α- or β-blockers on an emergency 1. Polak JM, Bloom SR. Regulatory peptides of the gastrointestinal
and respiratory tracts. Arch Int Pharmacodyn Ther. 1986;280:16–49.
basis—if necessary. 2. Wong F, Kim E. Peptide receptor imaging. In: Kim E, Yang D, edi-
Although the role of MIBG therapy has been best defined tors. Targeted molecular imaging in oncology. New York: Springer;
in patients with adrenal and extra-adrenal paragangliomas, 2011. p. 102–10.
3. Hillel PG, van Beek EJ, Taylor C, Lorenz E, Bax ND, Prakash V, Mol Imaging. 2011:524130. https://doi.org/10.1155/2011/524130.
et al. The clinical impact of a combined gamma camera/CT imag- Epub 2011 Nov 9
ing system on somatostatin receptor imaging of neuroendocrine 19. Severi S, Nanni O, Bodei L, Sansovini M, Ianniello A, Nicoletti
tumours. Clin Radiol. 2006;61:579–87. S, et al. Role of 18FDG PET/CT in patients treated with
4. Perri M, Erba P, Volterrani D, Lazzeri E, Boni G, Grosso M, et al. 177
Lu-DOTATATE for advanced differentiated neuroendocrine
Octreo-SPECT/CT imaging for accurate detection and localization tumours. Eur J Nucl Med Mol Imaging. 2013;40:881–8.
of suspected neuroendocrine tumors. Q J Nucl Med Mol Imaging. 20. Pacak K, Eisenhofer G, Ilias I. Diagnosis of pheochromocytoma
2008;52:323–33. with special emphasis on MEN2 syndrome. Hormones (Athens).
5. Kwekkeboom DJ, Krenning EP, Scheidhauer K, Lewington V, 2009;8:111–6.
Lebtahi R, Grossman A, et al. ENETS consensus guidelines for 21. Bodei L, Mueller-Brand J, Baum RP, Pavel ME, Hörsch D,

the standards of care in neuroendocrine tumors: somatostatin O’Dorisio TM, et al. The joint IAEA, EANM, and SNMMI prac-
receptor imaging with 111In-pentetreotide. Neuroendocrinology. tical guidance on peptide receptor radionuclide therapy (PRRNT)
2009;90:184–9. in neuroendocrine tumours. Eur J Nucl Med Mol Imaging.
6. Kwekkeboom D, Krenning EP, de Jong M. Peptide receptor imag- 2013;40:800–16.
ing and therapy. J Nucl Med. 2000;41:1704–13. 22. Esser JP, Krenning EP, Teunissen JJ, Kooij PP, van Gameren AL,
7. Kwekkeboom DJ, Kam BL, van Essen M, Teunissen JJ, van Eijck Bakker WH, et al. Comparison of [177Lu-DOTA0,Tyr3]octreotate
CH, Valkema R, et al. Somatostatin receptor-based imaging and and [177Lu-DOTA0,Tyr3]octreotide: which peptide is preferable for
therapy of gastroenteropancreatic neuroendocrine tumors. Endocr PRRT? Eur J Nucl Med Mol Imaging. 2006;33:1346–51.
Relat Cancer. 2010;17:R53–73. 23. Virgolini I, Britton K, Buscombe J, Moncayo R, Paganelli G, Riva
8. Buchmann I, Henze M, Engelbrecht S, Eisenhut M, Runz A, Schäfer P. In- and Y-DOTA-lanreotide: results and implications of the
M, et al. Comparison of 68Ga-DOTATOC PET and 111In-DTPAOC MAURITIUS trial. Semin Nucl Med. 2002;32:148–55.
(Octreoscan) SPECT in patients with neuroendocrine tumours. Eur 24. Wehrmann C, Senftleben S, Zachert C, Muller D, Baum RP. Results
J Nucl Med Mol Imaging. 2007;34:1617–26. of individual patient dosimetry in peptide receptor radionuclide
9. Junik R, Drobik P, Malkowski B, Kobus-Blachnio K. The role of therapy with 177Lu-DOTA-TATE and 177Lu-DOTA-NOC. Cancer
positron emission tomography (PET) in diagnostics of gastroen- Biother Radiopharm. 2007;22:406–16.
teropancreatic neuroendocrine tumours (GEP NET). Adv Med Sci. 25. Bodei L, Cremonesi M, Ferrari M, Pacifici M, Grana CM,

2006;51:66–8. Bartolomei M, et al. Long-term evaluation of renal toxicity after
10. Cecchin D, Lumachi F, Marzola MC, Opocher G, Scaroni C,
peptide receptor radionuclide therapy with 90Y-DOTATOC and
Zucchetta P, et al. A meta-iodobenzylguanidine scintigraphic scor- 177
Lu-DOTATATE: the role of associated risk factors. Eur J Nucl
ing system increases accuracy in the diagnostic management of Med Mol Imaging. 2008;35:1847–56.
pheochromocytoma. Endocr Relat Cancer. 2006;13:525–33. 26. Jamar F, Barone R, Mathieu I, Walrand S, Labar D, Carlier P, et al.
11. Nakatani T, Hayama T, Uchida J, Nakamura K, Takemoto Y,
(86YDOTA0)-D-Phe1-Tyr3-octreotide (SMT487) – a phase 1 clinical
Sugimura K. Diagnostic localization of extra-adrenal pheochromo- study: pharmacokinetics, biodistribution and renal protective effect
cytoma: comparison of 123I-MIBG imaging and 131I-MIBG imaging. of different regimens of amino acid co-infusion. Eur J Nucl Med
Oncol Rep. 2002;9:1225–7. Mol Imaging. 2003;30:510–8.
12. Bombardieri E, Giammarile F, Aktolun C, Baum RP, Bischof
27. Bodei L, Cremonesi M, Grana C, Rocca P, Bartolomei M, Chinol
Delaloye A, Maffioli L, et al. 131I/123I-metaiodobenzylguanidine M, et al. Receptor radionuclide therapy with 90Y-[DOTA]0-Tyr3-
(mIBG) scintigraphy: procedure guidelines for tumour imaging. octreotide (90Y-DOTATOC) in neuroendocrine tumors. Eur J Nucl
Eur J Nucl Med Mol Imaging. 2010;37:2436–46. Med Mol Imaging. 2004;31:1038–46.
13. Fottner C, Helisch A, Anlauf M, Rossmann H, Musholt TJ, Kreft 28. Rolleman EJ, Melis M, Valkema R, Boerman OC, Krenning EP,
A, et al. 6-18F-fluoro-L-dihydroxyphenylalanine positron emission de Jong M. Kidney protection during peptide receptor radionuclide
tomography is superior to 123I-metaiodobenzyl-guanidine scintigra- therapy with somatostatin analogues. Eur J Nucl Med Mol Imaging.
phy in the detection of extraadrenal and hereditary pheochromocy- 2010;37:1018–31.
tomas and paragangliomas: correlation with vesicular monoamine 29. Giammarile F, Chiti A, Lassmann M, Brans B, Flux G. EANM pro-
transporter expression. J Clin Endocrinol Metab. 2010;95:2800–10. cedure guidelines for 131I-meta-iodobenzylguanidine (131I-mIBG)
14. Peter D, Liu Y, Sternini C, de Giorgio R, Brecha N, Edwards therapy. Eur J Nucl Med Mol Imaging. 2008;35:1039–47.
RH. Differential expression of two vesicular monoamine transport- 30. Castellani MR, Seghezzi S, Chiesa C, Aliberti GL, Maccauro M,
ers. J Neurosci. 1995;15:6179–88. Seregni E, et al. 131I-MIBG treatment of pheochromocytoma: low
15. Ambrosini V, Tomassetti P, Castellucci P, Campana D, Montini versus intermediate activity regimens of therapy. Q J Nucl Med Mol
G, Rubello D, et al. Comparison between 68Ga-DOTA-NOC and Imaging. 2010;54:100–13.
18
FDOPA PET for the detection of gastro-entero-pancreatic and 31. Hoefnagel CA, Voute PA, De Kraker J, Valdes Olmos RA. [131I]
lung neuro-endocrine tumours. Eur J Nucl Med Mol Imaging. metaiodobenzylguanidine therapy after conventional therapy for
2008;35:1431–8. neuroblastoma. J Nucl Biol Med. 1991;35:202–6.
16. Haug A, Auernhammer CJ, Wangler B, Tiling R, Schmidt G, 32. Howard JP, Maris JM, Kersun LS, Huberty JP, Cheng SC, Hawkins
Goke B, et al. Intraindividual comparison of 68Ga-DOTA-TATE RA, et al. Tumor response and toxicity with multiple infusions of
and 18F-DOPA PET in patients with well-differentiated m etastatic high dose 131I-MIBG for refractory neuroblastoma. Pediatr Blood
neuroendocrine tumours. Eur J Nucl Med Mol Imaging. Cancer. 2005;44:232–9.
2009;36:765–70. 33. Lashford LS, Lewis IJ, Fielding SL, Flower MA, Meller S,

17. Mohnike K, Blankenstein O, Minn H, Mohnike W, Fuchtner F, Kemshead JT, et al. Phase I/II study of iodine 131 metaiodoben-
Otonkoski T. [18F]-DOPA positron emission tomography for pre- zylguanidine in chemoresistant neuroblastoma: a United Kingdom
operative localization in congenital hyperinsulinism. Horm Res. Children’s Cancer Study Group investigation. J Clin Oncol.
2008;70:65–72. 1992;10:1889–96.
18. Oh S, Prasad V, Lee DS, Baum RP. Effect of peptide receptor 34. Matthay KK, Yanik G, Messina J, Quach A, Huberty J, Cheng SC,
radionuclide therapy on somatostatin receptor status and glucose et al. Phase II study on the effect of disease sites, age, and prior
metabolism in neuroendocrine tumors: intraindividual compari- therapy on response to iodine-131-metaiodobenzylguanidine ther-
son of Ga-68 DOTANOC PET/CT and F-18 FDG PET/CT. Int J apy in refractory neuroblastoma. J Clin Oncol. 2007;25:1054–60.
Radionuclide Imaging
of the Nephro-Urinary Tract 30
Duccio Volterrani, Federica Orsini,
and Federica Guidoccio
Contents
30.1.1 Anatomy and Physiology of the Kidney and Urinary Tract 786
30.1.2 Radiopharmaceuticals in Nephrourology 787
30.2 Dynamic or Sequential Renal Scintigraphy 788
30.2.1 Methodology 788
30.2.2 Image Interpretation 789
30.2.3 Data Processing 790
30.2.4 Mean Transit Times 791
30.2.5 GFR Measurement 793
30.3 Evaluation of GFR with Blood Sampling 794
30.3.1 Single-Injection “Bolus” Techniques 794
30.3.2 Practical Protocol to Estimate GFR with 99mTc-DTPA (Two-Compartment Model) 795
30.4 Dynamic Renal Scintigraphy with Frusemide 795
30.4.1 Test Design 796
30.4.2 Interpretation Criteria 797
30.5 Dynamic Renal Scintigraphy with Captopril Test 798
30.5.1 Preparation and Test Design 798
30.5.2 Interpretation Criteria 798
30.6 Static Renal Scintigraphy 800
30.6.1 Technical Protocol 801
30.6.2 Image Analysis and Interpretation 801
30.7 Radionuclide Cystography for Vesicoureteral Reflux 802
30.7.1 Clinical Indication 802
30.7.2 Patient Preparation 802
30.7.3 Technique 803
30.7.4 Image Interpretation 803
30.8 [18F]FDG PET/CT for Tumors of the Nephro-Urinary Tract 804
30.9 PET Tracers Other Than [18F]FDG for Renal Cancers 805
30.9.1 18
F-Fluorothymidine 805
30.9.2 18F-FMISO 805
30.9.3 124I-cG250 805
30.9.4 18F-Fluoride 806
D. Volterrani (*) · F. Guidoccio

F. Orsini
Nuclear Medicine Unit, “Maggiore della Carità” University

30.10 PET Tracers Other Than [18F]FDG for Bladder Cancer 806

30.10.1 [11C]Choline 806
30.10.2 [11C]Methionine 807
30.10.3 [11C]Acetate 807
30.10.4 F-Fluoride
18
807
References 807
Learning Objectives The kidneys are wrapped by four different layers that are
• Understand the rationale of exploring renal function using arranged as follows (from deep to superficial):
radionuclide imaging.
• Describe the methodology of performing dynamic renal • The renal capsule composed of a tough fibrous tissue.
scintigraphy. • The perirenal fat, a gathering of extraperitoneal fat.
• Understand how to analyze and interpret the results of • The renal fascia that encloses the kidneys and the adrenal
dynamic renal scintigraphy. glands.
• Describe how to measure the glomerular filtration rate • The pararenal fat, mainly located on the posterolateral
with radioactive tracers. side of the kidney.
• Understand the rationale of the captopril renographic test
and of diuresis renography. As to the kidney per se, the renal parenchyma can be
• Describe the methodology of performing the captopril divided into two main areas: the outer cortex and the inner
renographic test and diuresis renography. medulla. The cortex extends into the medulla (columns of
• Understand the rationale and describe the methodology of Bertin), dividing it into triangular shapes also known as
static renal scintigraphy. renal pyramids. The apex of each renal pyramid is called
• Describe how to analyze and interpret the scintigraphic renal papilla and is associated with a minor calyx, which
findings obtained with static renal scintigraphy. collects urine from the pyramids. Several minor calices
• Describe the methodology for performing radionuclide merge to form a major calyx. Urine passes through the
cystography. major calices into the renal pelvis, from which it drains
• Understand how to interpret the results of radionuclide through the ureter into the bladder. All the urinary tract
cystography in the evaluation of vesicoureteral reflux. including the renal pelvis, the ureters, the bladder, and parts
• Summarize the main indications of [18F]FDG PET/CT in of the urethra have a type of epithelium consisting of
renal cell and bladder cancer. approximately 3–5 cell layers, also called urothelium or
• Summarize the results of clinical studies on the use of “transitional epithelium.”
non-[18F]FDG PET tracers in patients with renal or blad- The kidneys serve important functions, including filtra-
der cancers. tion and excretion of metabolic breakdown products (urea
and ammonia); regulation of necessary electrolytes, fluid,
and the acid-base balance; and stimulation of red blood cell
30.1 Introduction production. They also help to regulate blood pressure via the
renin-angiotensin-aldosterone system, controlling reabsorp-
30.1.1 Anatomy and Physiology of the Kidney tion of water and maintaining intravascular volume.
and Urinary Tract Each kidney contains over one million nephrons. Each
nephron is a functional unit composed of the renal (or
The kidneys are located in the retroperitoneal space at the Malpighian) corpuscle containing the glomerulus (which acts
sides of the psoas muscle (D11-L3 tract), about 7 cm in depth as a plasma filtration unit) and of a tubular system that has the
from the posterior abdominal wall. The left kidney is typi- task of producing the final urine. The renal corpuscle consists
cally somewhat more superior in position than the right kid- of the glomerulus (a cluster of afferent and efferent arterioles
ney. The upper poles are normally oriented more medially and capillaries, having a fenestrated endothelium that allows
and posteriorly than the lower poles. The medial margin of the plasma ultrafiltration) and of the Bowman’s capsule (cor-
each kidney is marked by a deep fissure, known as the renal responding to the initial dead-end part of the renal tubule
hilum that acts as a gateway to the kidney. wrapping the glomerulus that collects the glomerular filtrate).
30 Radionuclide Imaging of the Nephro-Urinary Tract 787
Glomerular filtration is driven by the high hydrostatic

pressure of the blood in the capillary loops of the glomerulus • The basic anatomic and functional renal unit is the
(70 mmHg, about twofold versus the common capillary net- nephron, which includes a glomerulus (where ultra-
works, due to the resistance offered by the efferent arteriole). filtration of plasma occurs through a fenestrated
The capillary hydrostatic pressure is opposed to the capsular endothelium) and a tubular system (which produces
pressure and plasma osmotic pressure; the resultant glomer- the final urine through a complex series of reab-
ular filtration pressure is about 35 mm Hg. The glomerular sorption and secretion of various compounds).
filtration rate (GFR) is about 120 mL/min in a healthy adult • The renal vascular resistance is regulated through
individual. Thus, in 1 day about 180 L of plasma are filtered, numerous factors including the autonomic nervous
99% of which are reabsorbed, while the metabolites to be system, the renin-angiotensin system, endothelin,
eliminated are concentrated in the urine. vasopressin, natriuretic peptides, kinins, prosta-
The renal arteries originate from the abdominal aorta at glandins, and nitric oxide.
L1–L2 level and, after passing the hilum, divide into segmen- • Either parenchymal, vascular, or obstructive disease
tal terminal-type arteries. These are divided into interlobar can acutely or chronically affect the nephro-urinary
arteries, which run in the middle of the columns of Bertin up apparatus.
to the cortical border, where they continue in the arcuate • Non-radionuclide techniques allowing anatomo-
arteries from which several interlobular arteries originate giv- functional exploration of the nephro-urinary appa-
ing rise to numerous afferent arterioles that feed the glomeru- ratus include ultrasound (with color Doppler), CT,
lar capillaries. Each afferent arteriole divides into numerous and MRI.
capillaries (where the ultrafiltration of plasma occurs, with
formation of the so-called pre-urine within the Bowman’s
capsule), which then merge giving rise to the efferent arteri- 30.1.2 Radiopharmaceuticals
ole from which the peritubular capillary network originates. in Nephrourology
The renal vascular resistance is regulated through numerous
factors including the autonomic nervous system, the renin- The radiopharmaceuticals employed to explore different
angiotensin system, endothelin, vasopressin, natriuretic pep- functions (such as glomerular filtration, renal plasma flow,
tides, kinins, prostaglandins, and nitric oxide. Most of them are and tubular function) [1, 2] can be classified according to
involved in the final integrated adjustment of pre- and post-glo- their renal extraction mechanism, as summarized below
merular resistors (efferent and afferent arterioles, respectively), (see Chap. 2 for details).
which allows the self-regulation of renal blood flow and GFR
over a wide range of variations in the renal perfusion pressure. 30.1.2.1 Radiopharmaceuticals Excreted
Renal blood flow normally amounts to about 1200 mL/min, by Glomerular Filtration
thus accounting for about 20% of the overall cardiac output. The most frequently used radiopharmaceutical of this group
The different diseases involving acutely or chronically the is 99mTc-DTPA, which in the normal subject has an extraction
nephro-urinary apparatus are due to either parenchymal, vas- fraction of about 20%, corresponding to the filtration frac-
cular, or obstructive diseases that all involve functional alter- tion; its low, but non-negligible binding to plasma proteins
ations of perfusion and glomerular and/or tubular function. (3–5%) results in a marginal error in the GFR estimate com-
Ultrasound, color Doppler, CT, and MRI can aid in the diag- pared to reference values measured using inulin or
nostic work-up of different kidney and urinary tract diseases, 51
Cr-EDTA. 99mTc-DTPA is commonly employed for sepa-
while nuclear medicine techniques (in particular dynamic rate GFR determinations (i.e., the split renal function) with
and static renal scintigraphy) primarily explore functional gamma camera dynamic scintigraphy.
aspects to evaluate the degree of kidney impairment.
30.1.2.2 R adiopharmaceuticals Excreted by
Tubular Secretion
In addition to glomerular filtration, these radiopharmaceuticals
Key Learning Points undergo tubular secretion, with a higher extraction fraction from
• The kidneys filtrate and excrete metabolic break- the blood, therefore providing better-quality imaging than 99mTc-
down products, regulate electrolyte concentration DTPA in case of impaired renal function. For instance, the extrac-
in body fluids as well as the acid-base balance, tion fraction for 123I-hippuran is ≥85%; for this reason, its plasma
stimulate red blood cell production, and contribute clearance approximates renal plasma flow. Currently, the tracer
to regulate arterial blood pressure via the renin- with tubular excretion most frequently used is 99mTc-MAG3,
angiotensin-aldosterone system. which is mainly cleared from the blood by tubular secretion and, in
a very small fraction, by hepatobiliary excretion. Its overall
e xtraction rate is approximately 60%. Such high extraction enables

to obtain better-quality images than those obtained with 99mTc-
DTPA. Its clearance mirrors the tubular extraction rate (TER), xyphoid
which approximates the effective renal plasma flow (ERPF).
99m
Tc-MAG3 is currently considered the radiopharmaceutical of
choice for functional imaging in children and for transplanted kid-
ney evaluation and when kidney function is significantly impaired.
30.1.2.3 Radiopharmaceuticals with Cortical
FOV
FOV
Binding
These radiopharmaceuticals have a prolonged retention in
the renal parenchyma and can therefore be used to assess the
functionally active renal mass. The most frequently used
parenchymal tracer is 99mTc-DMSA, which has a prolonged
retention in the renal cortex, by binding to the proximal tubu- pubis
lar cells. This imaging is mainly used for the diagnosis of
pyelonephritis in pediatric patients with vesicoureteral reflux
Fig. 30.1 Anatomical landmarks for correctly centering the patient
or with recurrent urinary tract infections.
within the FOV of the gamma camera for complete imaging of the kid-
neys and bladder (modified from: Volterrani D, Erba PA, Mariani G,
Eds. Fondamenti di Medicina Nucleare – Tecniche e Applicazioni.
Key Learning Points Milan: Springer Italy; 2010)
Radiopharmaceuticals allowing estimation/measure-
ment of glomerular filtration include 99mTc-DTPA and generally properly centered, even when ptosis or ectopia is
51
Cr-EDTA. present. In case of transplanted kidney evaluation, the detector
Radiopharmaceuticals allowing estimation of the must be positioned anteriorly centering the pelvis, because in
tubular excretion function include 123I-hippuran and these patients the kidney is located in the iliac fossa.
99m
Tc-MAG3. The recommended administered activities in adults are as
99m
Tc-DMSA allows estimation of the functionally follows:
active renal mass.
• 100–150 MBq for 99mTc-MAG3.
• 200–300 MBq for 99mTc-DTPA.
30.2 D
ynamic or Sequential Renal • 50–100 MBq for 123I-Hippuran.
Scintigraphy
For patients in the pediatric age range, activities must be
30.2.1 Methodology reduced (generally based on body weight), as suggested,
e.g., in the “pediatric dosage card” of the European
Adequate hydration is required for this procedure [3], which Association of Nuclear Medicine (accessible at “http://www.
is generally achieved by asking the patient to drink water eanm.org/publications/dosage-card/”).
(about 7 mL/kg) over the hour preceding the acquisition; in Although there is no unique standardized acquisition pro-
fact, dehydration can cause delay in intraparenchymal transit tocol, the dynamic acquisition is generally performed over
and excretion of the radiopharmaceutical. Immediately 15–30 min (usually 20 min), which starts at the exact moment
before positioning on the imaging table, the patient is asked of injecting the radiopharmaceutical as a small-volume sin-
to urinate in order to void the bladder. The scan is generally gle bolus [4]. The number of frames acquired depends on
performed with the patient in the supine position. their length, which can be constant (single-phase acquisition,
A large FOV gamma camera detector, equipped with e.g., all frames with a duration of 10 s) or variable (acquisi-
LEHR parallel-hole collimator (HEGP when using tion with multiple phases) as in the following example:
131
I-Hippuran), is positioned posteriorly at the level of the lum-
bar region. In order to center the two kidneys in the FOV, it is • Phase 1: 1 frame/s for the first min (called first pass or
advisable to look for anatomical landmarks with the help of a angiogram), to characterize renal perfusion.
radioactive point source: the xiphoid should be in the proxim- • Phase 2: 1 frame/10 s for the next 4 min, to characterize
ity of the upper edge of the FOV, the space between the iliac active renal extraction (parenchymal uptake) of the
crest and costal arch should correspond to half the height of radiopharmaceutical.
the FOV, and the pubis should correspond to the lower portion • Phase 3: 1 frame/20 s for the next 15 min, to characterize
of the FOV (Fig. 30.1). Using these landmarks, the kidneys are renal excretion.
The acquisition matrix is normally 64 × 64, although 30.2.2 Image Interpretation

the 128 × 128 matrix can be used, especially in chil-
dren. At the end of the dynamic acquisition, a static Qualitative interpretation of the images constitutes the first
image can be obtained (if possible in the upright posi- phase in the analysis of a dynamic renal scintigraphy; it is based
tion) with the aim to evaluate washout of the radiophar- on visualization of the entire sequence of the images acquired,
maceutical from the urinary tract and to identify a correlating the images to the various distribution phases of the
mobile kidney. radiotracer within the kidney (Fig. 30.2), as follows:
Fig. 30.2 Dynamic sequential renal scintigraphy. Sequence of 1-min seen in the heart) to the excretion phase (reproduced with permission
frames describing the various phases of distribution of 99mTc-DTPA, from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
from the vascular phase (when the radioactive blood pool can be clearly Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
• Vascular phase (first min): reduced uptake in one kidney nephropathies. Abnormal persistence of activity in a dilated
indicates reduced perfusion relative to the contralateral pelvis indicates hydronephrosis.
kidney.
• Parenchymal phase (2–3 min): in this phase the tracer
accumulation in the parenchyma is proportional to renal 30.2.3 Data Processing
function (GFR, ERPF, or TER).
• Excretory phase (remainder of the acquisition): renal The standard processing of sequential renal scintigraphy
chalices and the pelvis begin to accumulate the radiophar- generates activity/time curves obtained from the kidney
maceutical 3–5 min after administration; over the next ROIs, which can be manually drawn using a summed image
10–15 min, activity in the renal parenchyma progressively relative to the parenchymal phase, where the margins of the
declines, while activity in the bladder increases. kidneys are better detectable. It is also possible to use soft-
ware that automatically or semiautomatically defines the
Delayed visualization of the chalices and pelvis indicates kidney contours (Fig. 30.3). Background activity/time curves
a slow parenchymal transit time, as it occurs in several (usually obtained by drawing a perirenal ROI below the
Fig. 30.3 Processing of dynamic renal scintigraphy. The upper middle activity/time curves of the vascular phase with starting and vascular peak
image corresponds to the parenchymal phase (3 min), with overlaid limits in yellow. Bottom right: the two renograms (right kidney in purple,
drawing of the kidney and background ROIs. Top left: image of the vas- left kidney in blue) (reproduced with permission from: Volterrani D,
cular phase (sum of frames from 0–30 sec) with the corresponding kid- Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche
ney and aorta ROIs. Top right: image of the excretion phase. Bottom left: e Applicazioni. Milan: Springer Italy; 2010)
In studies that include acquisition of the angiogram, it is

parenchymal possible to obtain indices of renal perfusion, by comparing
phase
the curve of the kidney vascular phase with that obtained
over the aorta or an iliac artery. Assessment of the slope or
the area from 0 to time to peak allows calculation of the renal
perfusion indices. When evaluating a transplanted kidney,
the Hilson perfusion index is calculated from the ratio
excretion
phase
between the area (from 0 to vascular peak) under the curve
obtained over the iliac artery and the kidney curve: a ratio
counts
>1.5 is consistent with kidney hypoperfusion, as often

observed in case of acute rejection [6].
vascular
phase
30.2.4 Mean Transit Times
Following its extraction from the circulating plasma, the

radiopharmaceutical crosses the renal parenchyma up to the
urinary tract. The time for passing through a single nephron
min is called transit time; the mean transit time is the average
time that the radiopharmaceutical spends to transit through
Fig. 30.4 Diagrammatic representation of the renogram: vascular,
the whole system of nephrons that forms the kidney
parenchymal, and excretion phase (modified from: Volterrani D, Erba
PA, Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e parenchyma.
Applicazioni. Milan: Springer Italy; 2010) Ideally, assuming injection of the radiopharmaceutical as
a bolus directly into the renal artery and assuming no recir-
culation through the venous system (i.e., as an impulsive
lower pole or from the upper to the lower pole laterally of the input), the kidney’s response (or impulsive output) would be
kidneys) are subtracted from the renal activity/time curves as shown in Fig. 30.6 and would include the following
[5], thus obtaining the so-called renographic curves or reno- components:
grams (Fig. 30.4) in which a vascular phase (first rapid rising
phase), a parenchymal phase (second slow rising phase), and • A curve with an early and narrow peak (within 5–10 s),
an excretion phase (third declining phase) can be corresponding to the fraction of the tracer that does not
recognized. accumulate in the parenchyma in a single pass but only
The following quantitative and semiquantitative indices has one intravascular transit.
can be obtained from the renograms, usually included in the • Plateau phase (time during which the radiopharmaceuti-
report (Fig. 30.5): cal passes through the parenchyma).
• Excretion phase.
• Tmax, or parenchymal peak time (time needed to reach the
parenchymal peak). In reality, the input is not impulsive but has a shape that is
• Excretion T½: time required for counts to decline by 50% determined both by fragmentation of the bolus from the
during the excretion phase. point of i.v. injection and then through the lung circulation
• Mean transit time (see “Calculation of mean transit and by the curve of plasma clearance (I(t)), while the kidney
time”). output is the renogram (R(t)). What relates the plasma disap-
• Relative parenchymal indices of each kidney (or split pearance curve with the renogram is the impulsive output
renal function), expressed as % of the overall renal func- (kidney response), also called transfer function (H(t)). The
tion based on either integral counts over a certain time renogram is obtained through convolution of the plasma dis-
interval (e.g., 2–3 min) or on the slope of the parenchymal appearance curve with the transfer function:
phase.
R(t ) = I (t ) ⊗ H (t )

It is also possible to define ROIs that do not include the
whole kidney but only the parenchyma, i.e., excluding cali- In practice, from the transfer function H(t) (that can be cal-
ces and pelvis. This approach, which can be useful in case of culated through the inverse operation, called deconvolution,
urinary stasis, produces the so-called parenchymal or by deconvoluting the renogram R(t) for the plasma curve
renograms. I(t)), obtained by drawing an ROI on the left ventricle:
Fig. 30.5 Final report of a dynamic sequential renal scintigraphy with reported (reproduced with permission from: Volterrani D, Erba PA,
99m
Tc-DTPA. Representative images of the vascular, parenchymal, and Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
excretion phases are displayed, with the corresponding vascular curves Applicazioni. Milan: Springer Italy; 2010)
and renograms. Semiquantitative and quantitative indices are also
H (t ) = R(t ) ⊗ I (t ) H (ω ) = R(ω ) / I (ω )

(where symbol ⊗ denotes the convolution operation). From Thus, after calculating H (ω ) , the inverse Fourier transform
a mathematical point of view, deconvolution is performed is performed to obtain H (t ) , from which the mean transit time
based on the Fourier transform of R(t ) and I (t ) that are thus can be calculated after subtracting the early peak of intravas-
transferred from the time domain to frequency domain (R(𝜔) cular phase of the radiopharmaceutical (Fig. 30.7) [7].
and I(𝜔)), where the operation of deconvolution becomes a From the clinical point of view, estimation of the mean
simple division: transit times is especially useful when evaluating patients
MTT =
∫ H’ (t )
h h
H(t ) H’(t )
input output
transfer function
Fig. 30.7 Calculation of mean transit time. After removing the “spike”
of the initial transfer function, the mean transit time (MTT) is obtained by
dividing the area under the curve by the height (h) of the plateau (modified
from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
30.2.5 GFR Measurement
The main clinical indication of dynamic renal scintigraphy is

quantitative estimation of global and separate renal function,
the most important global parameter of renal function being
blood clearance curve renogram GFR. Sequential renal scintigraphy with 99mTc-DTPA is cur-
rently the only method that allows calculation of the global
Fig. 30.6 Calculation of the mean transit time. The output curve (top
right) represents the theoretical response of the kidney to the input (top
and separate GFR (split renal function). The most commonly
left) but also the transfer function that relates the plasma clearance employed method to estimate GFR relies on the computa-
curve (bottom left) with the renogram (bottom right). The transfer functional approach published by Gates [8], an indirect method
tion is obtained by deconvoluting the renogram by the plasma clearance based on calculation of integral kidney counts during the
curve (modified from: Volterrani D, Erba PA, Mariani G, Eds.
third minute from the time of arrival of the tracer bolus to the
Springer Italy; 2010) kidneys. Net kidney counts (i.e., subtracted of background
counts) are corrected for attenuation (where renal depth is
calculated using the Tønnesen algorithm based on weight
with suspected renal artery stenosis, obstructive disease of and height of the patient) and expressed as a fraction of the
the urinary tract, and rejection of kidney transplant. In these injected activity. GFR is calculated (in mL/min) using the
conditions, mean transit times are abnormally prolonged, equation derived from the linear regression between the renal
although the threshold value for a transit time to be consid- integral counts and creatinine clearance, which Gates had
ered as delayed remains controversial. used as a reference (see the diagram below formula):
cpm Right Kidney cpm Left Kidney

−0.153 13.3( weight / height ) + 0.7 
+ −0.153 13.2 ( weight / height ) + 0.7 
GFR = e e × 100 × 9.75621 − 6.19843
Dose
The Gates method has an intrinsic inaccuracy mainly due to

the inadequacy of its gold standard, the plasma creatinine Key Learning Points
clearance. As for all the methods that are based on whole kid- • Dynamic renal scintigraphy is a noninvasive func-
ney counts, other factors of uncertainty can arise from variabil- tional imaging modality that allows assessment of
ity of renal and backgrounds ROIs and from correction of kidney function, GFR, ERPF, or TER according to
kidney depth (not actually measured in each individual patient). the specific radiopharmaceutical used.
According to the Tønnesen formula, the right/left kidney func- • The main diagnostic information provided by
tion contribution ranges from 50:50 to 43:57. However, in dynamic renal scintigraphy is estimation of the
patients with kyphoscoliosis and with ptotic, ectopic, or mal- global and separate kidney function, to be used for
formed kidneys, it would be necessary to actually measure characterizing bilateral and monolateral
renal depth or to adopt other approaches, not making blunders nephropathies.
with the estimate of global and separate renal function.
30.3 E
valuation of GFR with Blood 1,000,000
Sampling
Plasmatic radioactivity, cpm/mL

A parameter currently used in the clinical routine for estimat- C1/V1 C2/V2
ing renal function is the serum creatinine level. However, serum
creatinine rises significantly only when GFR is reduced by at
B
least 50%. On the other hand, also creatinine clearance as a A
measure of GFR is often inaccurate, especially when the renal 1,00,000
β
function is reduced, due to a compensatory increase in tubular α
secretion, which limits its validity as a measure of GFR.
Although the inulin clearance remains the “gold stan-
dard” for estimating GFR, its use is limited by the high cost
and overall complexity of the method. When relying on
agents that are cleared from plasma solely through glomeru-
lar filtration, GFR measurement can be performed using the 10,000
classical formula: 0 30 60 90 120 150 180
Time after bolus injection, min
U ×V
GFR =
P Fig. 30.8 Representative plasma curve obtained with sampling up to
180 min after bolus i.v. injection of 99mTc-DTPA in an individual with
where U is inulin urinary concentration, V is urine volume, normal renal function. In this model V1 represents volume of the central
and P is inulin plasma concentration. compartment where the tracer is introduced (the plasma volume) and
For research purposes, the procedures based on continu- rapidly and homogeneously. The second compartment (V2) corresponds
ous infusion of adequate radiopharmaceuticals and timed col- to the extravascular space in which the tracer diffuses leaving compart-
ment V1. At the same time, some tracers diffuse back from compartment
lection of blood samples and urine remain the best method for V2 to compartment V1 as a function of the concentration gradient
measuring renal clearance. However, in the clinical routine, between the two compartments (|C1 − C2|) (modified from: Volterrani
methods based on single-bolus injection offer greater sim- D, Erba PA, Mariani G, Eds. Fondamenti di Medicina Nucleare –
plicity and a sufficient accuracy to meet the clinical demand. Tecniche e Applicazioni. Milan: Springer Italy; 2010)
These techniques are much more accurate for estimating the
GFR than techniques based on dynamic gamma camera space in which the tracer diffuses but from which also redistrib-
imaging, such as the Gates technique described above. utes back into the plasma compartment, as a function of the
The radionuclide of choice for estimating GFR is concentration gradient between the two compartments.
51
Cr-EDTA because its renal clearance is the closest to inulin According to this model, the clearance of 99mTc-DTPA
clearance among all known agents. On the other hand, the (which approximates GFR) can be calculated according the
clearance of 99mTc-DTPA is close with that of 15Cr-EDTA; Stewart/Hamilton principle, as follows:
therefore, both tracers allow an accurate measurement of
Injected _ Activity
GFR, especially for values greater than 30 mL/min. 99mTc- Cl = ∞
DTPA has some advantages over 51Cr-EDTA, such as wide
∫ 0
Ae −α t + Be − β t
availability, low cost, low radiation dose to the patient, and
γ-ray emission suitable for gamma camera imaging. The accuracy of this approach depends on the number and
timing of blood sampling after tracer injection: for an accu-
rate estimate of GFR, plasma sampling is required over at
30.3.1 Single-Injection “Bolus” Techniques least 3 h, the first sample should be drawn between 5 and
10 min after the injection, and the number of blood samples
30.3.1.1 Two-Compartment Approach should be at least six [9]. The results obtained with the bi-
The disappearance curve, or plasma clearance, of 99m
Tc- compartmental model are frequently used as the reference
DTPA can be approximated to a bi-exponential function that method for other procedures.
can be described as follows:
30.3.1.2 Single-Compartment Approach
y = Ae −α t + Be − β t
The growing need for simplification has led to develop a
where A and B correspond to the intercepts and α and β are the single-compartment model, which requires only two or
two corresponding exponential constants (Fig. 30.8). In this three blood samples. Blood samples are drawn when redis-
model, plasma represents the volume of the central compart- tribution of the tracer from the extravascular to the intravas-
ment in which the tracer is injected and where it rapidly distrib- cular space essentially depends on renal clearance. In this
utes. The second compartment corresponds to the extravascular way, it is assumed that the tracer is distributed in a single
compartment from which it is exclusively eliminated by the 7. The blood samples are collected in tubes containing
kidneys. However, by ignoring the first exponential, the EDTA and at the end of the study are centrifuged (10 min
GFR will be somewhat overestimated; the extent of this at 2000 g) to separate red blood cells from the plasma.
overestimation depends on the relationship between the 8. Using the Eppendorf pipette, 1-mL samples of plasma
first and the second exponential. Since the area of the expo- are prepared from each tube.
nential tends to be larger (slower clearance) in patients with 9. Plasma samples and diluted B standards drawn from the
impaired renal function, the error introduced by ignoring flask are measured in a gamma counter. Mean B_dil is
the first rapid exponential will be relatively smaller. calculated by averaging the counts from the three sam-
Over the years, other simplified techniques, requiring ples drawn from the flask.
only one or two blood samples, have been developed such as 10. The counts of plasma samples, corrected for the decay,
the Russel, Christensen, and Groth methods [9]. The level of and the actual times (hh:mm) at which they were drawn
renal function has a certain impact on the accuracy of each are fitted by a bi-exponential function.
method, particularly those with a single sample. It is impor- 11. The activity (A) actually injected into the patient is cal-
tant to notice that, in case of severe kidney failure, the extra- culated according to the following proportion:
renal clearance of 99mTc-DTPA will significantly overestimate
the actual GFR. For this reason, in the presence of severe
B _ dil × 1000 × pB
renal failure, some authors suggest to employ methods that A=
rely on urine collection that are similarly indicated in patients pA

with ascites and edema or with other causes of expanded dis-
tribution volume [9]. 12. GFR (expressed in mL/min) is calculated as the ratio
between the injected activity (A) and the integral from 0
to ∞ of the bi-exponential curve obtained from fitting.
30.3.2 Practical Protocol to Estimate GFR
with 99mTc-DTPA (Two-Compartment
Model) Key Learning Points
• Procedures assessing 99mTc-DTPA plasma clearance
The following steps must be undertaken: provide estimations of GFR more accurate than
gamma camera methods without plasma sampling.
1. A needle is placed in an antecubital vein for drawing • Fitting of the plasma clearance of 99mTc-DTPA can
blood samples; 99mTc-DTPA is injected in the contralat- be considered a gold standard, but it is time-
eral arm. consuming and requires multiple plasma samples
2. Two syringes (A and B) with equal activity must be pre- for at least 180 min.
pared (e.g., 15–20 MBq of 99mTc-DTPA or more if also • The growing need for simplification has led to the
imaging is to be performed): use of the single-compartment model or other sim-
A—activity to be administered to the patient. plified techniques requiring only one or two blood
B—standard. samples.
3. Activities in syringes A and B must be measured using a
γ-counter and, then, recorded separately; alternatively,
the syringes can be weighed using an electronic preci-
sion scale. 30.4 D
ynamic Renal Scintigraphy
4. Syringes A and B will be recounted or reweighed later, with Frusemide
after emptying: syringe A after injecting the radiophar-
maceutical into the patient and syringe B after diluting The test consists of a diuretic stimulation (obtained through
its content into a graduated flask containing 1000 mL of i.v. administration of Frusemide) added to the standard
water or physiological saline. In this way the amount (in dynamic renal scintigraphy. The test is indicated for distin-
terms of activity or grams) of 99mTc-DTPA actually guishing urinary stasis caused by a simple dilation of the uri-
injected into the patient (pA) and diluted (pB) in the nary tract (e.g., congenital dilatation, pelvis atony) from
flask is accurately measured. dilation caused by an obstructive disease. This differential
5. Three 1-mL samples (each corresponding to 1:1000 B) diagnosis is especially important in the patient presenting
are drawn from the flask with a precision Eppendorf with dilation of the urinary tract (hydronephrosis or hydro-
pipette. ureteronephrosis); in fact, in the presence of an obstructive
6. Blood samples (3–5 each) are drawn at 5, 10, 15, 20, 30, pathology, if not properly corrected, the kidney tends to
40, 60, 90, 120, 150, and 180 min after administration of gradually lose its function developing an obstructive
99m
Tc-DTPA into the patient. nephropathy [10].
In a dilated pelvis with evident urinary stasis, the admin- • Absent, incomplete, or slow washout if there is a condi-
istration of frusemide determines a significant increase in tion of obstructive uropathy.
urinary flow already after 3–6 min that reaches its maximum
after about 15 min (from 1–2 to 15–18 mL/min), which can
induce two different responses: 30.4.1 Test Design
• Rapid washout of the tracer from the dilated pelvis if Several protocols have been described for carrying out the
obstruction is not present. test. The most frequently used protocol is defined as
Fig. 30.9 Frusemide test in a patient with hydronephrosis of the right permission from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti
kidney. Urinary stasis resolves promptly (with T½ about 3 min) after di Medicina Nucleare – Tecniche e Applicazioni. Milan: Springer Italy;
administration of the diuretic drug, thus indicating the absence of a sig- 2010)
nificant urinary stenosis (dilation without obstruction) (reproduced with
“F + 20,” whereby furosemide is injected i.v. (0.5 mg/kg 30.4.2 Interpretation Criteria

for the adult) at the end of the baseline acquisition of
sequential renal scintigraphy, therefore about 20 min after The urinary washout curve is evaluated after subtracting
administration of the radiopharmaceutical. The acquisi- background from the curve obtained by drawing an ROI over
tion of images (at least 120 10-s frames, same as the the stasis. Although there is no unanimous consensus, the
matrix scintigraphy) should start at about 20 s before i.v test suggests the absence of urinary obstruction when:
administration of the diuretic drug. A common practice in
nuclear medicine centers is to acquire the dynamic renal • Shape of the washout curve under the diuretic stimulus
scintigraphy and then the static post-void static acquisi- has an upward concavity.
tion and, if urinary stasis persists, to proceed with the • Half-time of the washout curve is <10–15 min (Figs. 30.9
diuretic test. and 30.10).
400.0
cts/sec
300.0
200.0
100.0
0.0
0.0 0.5 1.0 1.5
ksec
Fig. 30.10 Frusemide test in a patient with hydronephrosis of the left stenosis (dilation with obstruction) (reproduced with permission from:
kidney. Urinary stasis persists after administration of the diuretic drug Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina
(with T½ > 25 min), thus indicating the presence of a significant urinary Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010)
In cases with an uncertain/equivocal response (half-time angiotensin II). Therefore, ACE inhibition causes a reduction
15–20 min), some authors suggest to repeat the test with the in GFR as well as a slowing down of tubular transit times.
“F − 15” (whereby frusemide is injected 15 min before This effect, which is specific and not due to the systemic
injecting the radiopharmaceutical) or to apply the calculation hypotensive effect of the ACE inhibitor, can be visualized
of mean transit time. during dynamic renal scintigraphy with both glomerular and
Uncertain/equivocal response to the diuretic test can be tubular tracers.
caused by poorly functioning kidneys (therefore with
reduced sensitivity to the diuretic drug and unable to suffi-
ciently increase urinary flow) or by particularly voluminous 30.5.1 Preparation and Test Design
hydronephrosis that could produce a “reservoir” effect of
the pelvis which is able to continue to accumulate urine Before performing out the scan, therapy with ACE inhibitors
under the diuretic stimulus without evidence of significant must be discontinued for a few days (from 2 to 7, according
washout [11]. to biological half-life of the particular drug used), and
diuretic drugs should be discontinued for at least 1 week.
After adequate hydration (7 mL/kg over the hour preceding
Key Learning Points the test), the patient undergoes a baseline dynamic renal
• The frusemide test is indicated for distinguishing a scintigraphy. Then, on the same day or on a subsequent day,
dilation due to obstructive disease from simple dila- captopril is administered (50 mg orally), and a second
tion of the urinary tract without obstruction. dynamic renal scintigraphy is performed 60 min after capto-
• The most frequently used protocol consists in the pril administration, with the patient in the same hydration
i.v. administration of frusemide at the end of a conditions and under monitoring of blood pressure [13].
20 min dynamic renal study. Both glomerular and tubular radiotracers have been shown to
be suitable for the technique, although the majority of the
studies have been performed with 99mTc-DTPA [14].
30.5 D
ynamic Renal Scintigraphy
with Captopril Test 30.5.2 Interpretation Criteria
Renovascular hypertension is caused by renal hypoperfusion Visual assessment of scintigraphic and renographic changes
due to the presence of anatomic stenosis of the renal artery induced by captopril is the most common criterion adopted
(or its main intrarenal branches) and activation of the renin- to interpret results of the test (Figs. 30.11 and 30.12).
angiotensin system. The medical decision to treat a renal The renographic appearance varies from normal-
artery stenosis depends on the assessment of the causal rela- borderline (curve type 0–1) to parenchymal retention with
tionship between the stenosis and arterial hypertension and delayed excretion and late persistence of parenchymal activ-
on the likelihood of clinical benefit expected from the revas- ity (curve type 2–3) and to moderate and severe decrease of
cularization procedure [12]. the parenchymal uptake phase (especially with 99mTc-DTPA)
The scintigraphic captopril test explores a functional and as expression of significant reduction in renal function (curve
specific aspect of renovascular hypertension associated with type 4–5) (Fig. 30.13).
renal hypoperfusion, wherein angiotensin II contributes to False-positive results indicating renal artery stenosis may
the regulation of GFR in case of low levels of renal perfusion occur in the case of multiple renal arteries and in states of
pressure due to stenosis of the renal artery. Based on this dehydration or sodium depletion with activation of the renin
pathophysiological rationale, the captopril test was devel- system; for this reason diuretics should be stopped several
oped for the diagnosis of renovascular hypertension [13]. In days before the test. False-negative results can occur in cases
particular, administration of an inhibitor of the angiotensin of hemodynamically significant stenosis or with poorly func-
converting enzyme (ACE) abolishes the adaptation tioning kidneys already at baseline. However, the results are
mechanism; thereby a sufficient level of GFR is maintained optimal when analyzed in relation to the positive prediction
even in the presence of renal artery stenosis (mediated by the of the success of revascularization for treatment of hyperten-
vasoconstriction of the efferent arteriole operated by
sion [15].
Fig. 30.11 Renal captopril BRPA COUNT/TIME(”) BRPA

test in a patient with left renal
artery stenosis: baseline (top) RE RD
6000
and post-captopril (bottom)
99m
Tc-DTPA renograms. The
baseline study shows delayed
excretion on the left kidney.
After captopril, the left
renographic curve displays
markedly reduced amplitude,
with slow parenchymal 3000 LK
accumulation phase. Also the LK RK
condensed image shown in
right panels (corresponding to
the parenchymal phase) RK
displays reduced
accumulation of the PRE-CAPTOPRIL
BU230726
radiopharmaceutical, thus
0 0003 PRE–CPT 2.00’ – 3.00’
clearly indicating GFR
0.0 450.0 900.0 (ORIGINAL)
reduction (reproduced with
permission from: Volterrani
D, Erba PA, Mariani G, Eds. COUNT/TIME(”)
BRPA BRPA
Fondamenti di Medicina RE RD
Nucleare – Tecniche e 6000
Italy; 2010)
LK
3000 LK RK
RK
PRE-CAPTOPRIL
BU230726
0 0003 POS–CPT 2.00’ – 3.00’
0.0 450.0 900.0 (ORIGINAL)
Key Learning Points • The renal scintigraphic captopril test explores a

• The medical decision to treat renal artery stenosis in functional and specific aspect of renovascular
a patient with hypertension depends on verification hypertension associated with renal hypoperfusion.
of the causal relationship between the stenosis and • The renal scintigraphic captopril test positively pre-
arterial hypertension and on the likelihood of clini- dicts the success of revascularization for treatment
cal benefit expected from revascularization. of hypertension.
Fig. 30.12 Renal captopril 800.0

test in a patient with left renal cts/sec
artery stenosis: baseline (top)
and post-captopril (bottom)
99m
Tc-DTPA renograms. After 600.0
captopril the right renographic
curve is slightly reduced in
amplitude, and the
parenchymal accumulation 400.0
1min 3min
phase is markedly delayed.
The post-captopril study
(bottom left images) shows
mainly a delayed 200.0
parenchymal transit of the
radiopharmaceutical within
the left kidney, displaying no BASAL
significant excretion phase 0.0
even after 20 min (renal 0.0 1.0 2.0
10min 20min ksec
parenchyma accumulates 800.0
progressively the
radiopharmaceutical) cts/sec
from: Volterrani D, Erba PA, 600.0
1min 3min 400.0
200.0
POST CAPT.
0.0
0.0 1.0 2.0
10min 20min ksec
30.6 Static Renal Scintigraphy
Static renal scintigraphy with 99mTc-DMSA can be consid-

ered as the reference method for the detection of focal paren-
chymal damage. The main indications for static renal
scintigraphy can be summarized as follows:
• Diagnosis of acute pyelonephritis.

• Quantification of the resulting parenchymal damage.
• Follow-up and assessment of cortical scars (chronic
pyelonephritis).
It is used especially in the pediatric age [16], both for the

presence of afebrile forms and for the frequent lack of agree-
ment between clinical and other imaging findings (40% of
cases). Other clinical conditions in which a static renal scin-
Fig. 30.13 Types of renographic curves induced by captopril: 0 = nor- tigraphy can be performed are:
mal; 1 = mild; 2 = moderate; 3 = severe delayed parenchymal excretory
phase; 4 = mild/moderate parenchymal uptake failure; 5 = severe paren-
chymal uptake failure (washout-like curve) (reproduced with permission
• Renal agenesis.
from: Volterrani D, Erba PA, Mariani G, Eds. Fondamenti di Medicina • Ectopic kidney.
Nucleare – Tecniche e Applicazioni. Milan: Springer Italy; 2010) • Horseshoe kidney.
• Differential diagnosis between neoplastic renal masses or absent radiopharmaceutical uptake. Generally in acute
and lobular compensatory hypertrophy or prominent col- pyelonephritis, the cortical defects appear to be larger than in
umn of Bertin. the chronic phase, due to edema and ischemia associated
with the foci of acute infection. However, only follow-up
enables to distinguish between acute and chronic lesions; a
30.6.1 Technical Protocol defect that persists for >6 months is classified as a chronic
lesion, or scar. Subsequent episodes of acute pyelonephritis,
99m
Tc-DMSA is cleared from the blood by glomerular fil- when not diagnosed or properly treated, cause the develop-
tration and tubular secretion; however, a large fraction of ment of several scars in the renal parenchyma leading to an
the tracer filtered is reabsorbed and retained in the proxi- overall reduction in size of the kidney and parallel reduction
mal tubular cells (therefore at the level of the renal cor- of the functional renal mass (Fig. 30.14).
tex), thus enabling to obtain high-resolution and Image interpretation aims at quantifying the parenchymal
high-contrast imaging. No specific patient preparation is functional mass and at assessing the degree of relative or
required, and 99mTc- DMSA is administered i.v. in the absolute functional impairment, through ROI analysis.
amount of 37–110 MBq for an adult individual. Images Particularly the posterior view (along with the anterior
are acquired 2–4 h postinjection, this time interval being image) is useful for the semiquantitative evaluation of the
required to obtain a sufficient cortical accumulation of the functional mass of the two kidneys. Possible differences in
radiopharmaceutical and a low background activity. Good counting efficiency due to different depths (or attenuation) of
hydration during this time interval facilitates excretion of the two kidneys can be minimized by calculating the geo-
the fraction of radiopharmaceutical that does not accumu- metric mean of counts in the anterior and posterior views
late in the renal cortex and minimizes possible radiotracer (i.e., the square root of the product of the counts within each
stasis in the urinary tract. kidney ROI in the two views) (Fig. 30.15).
A large FOV gamma camera can be used, equipped with The geometric mean can also be calculated directly on the
LEHR or LEUHR collimators. A pinhole collimator can also images acquired in the two views with a dual-head gamma
be used to obtain high-quality images. With the patient in camera; the ROI matrix of the anterior image can be flipped,
supine position, planar images are acquired in posterior, and the geometric mean can be calculated with the same for-
anterior, and (right and left) posterior oblique views. Lateral mula at pixel level using the matrix of the posterior image. A
views are of little use as the activity of each kidney overlaps new ROI can be drawn on the image so obtained, which is
with that of the contralateral kidney (“shine-through” effect). called “conjugate image.”
In principle, SPECT acquisition is expected to improve the The real clinical impact of SPECT imaging over planar
diagnostic performance of static renal scintigraphy. imaging remains controversial. According to literature data,
the diagnostic accuracy of planar imaging is already very
high and not significantly further enhanced by SPECT acqui-
30.6.2 Image Analysis and Interpretation sitions [17]. Nonetheless, SPECT and SPECT/CT allow a
3D evaluation of both kidneys; thus, by using segmentation
In case of acute or chronic pyelonephritis, the parenchymal algorithms identifying the edges of the kidney surface, with
abnormalities are characterized as focal areas with reduced SPECT it is possible to assess the actual functional mass of
Fig. 30.14 Static renal

scintigraphy with 99mTc-
DMSA in a patient with
chronic pyelonephritis. The
left kidney is reduced in size
and cortical uptake (relative
function 31%). Focal uptake
defects correspond to scars
ANT POST
Fig. 30.15 Static renal scintigraphy

with 99mTc-DMSA in a patient with
pelvic left kidney that is better
displayed in the anterior view. By
calculating the geometric mean of
counts (in the anterior and posterior
views), in these cases it is possible to
correct for the different renal depth
of the two kidneys, thus improving
the assessment of split renal function
(reproduced with permission from:
Volterrani D, Erba PA, Mariani G,
Eds. Fondamenti di Medicina
Nucleare – Tecniche e Applicazioni.
Milan: Springer Italy; 2010) ANT POST
each kidney and more importantly volume changes that can diverticulum. Besides anatomical obstruction, there are neu-
occur during follow-up. rofunctional causes of high intravesical pressure that simi-
larly predispose to VUR, in particular a neurogenic bladder
associated with spina bifida.
Key Learning Points Treatment of VUR is aimed at preventing the pyelone-
• The main applications of static cortical scintigraphy phritis sequelae, damage to the renal parenchyma, arterial
consist in the diagnosis of acute and chronic hypertension, and chronic renal failure. In fact, VUR is an
pyelonephritis. important factor predisposing to acute pyelonephritis in
• Image analysis of a static renal study with 99mTc- patients with urinary tract infections (UTI). VUR is found in
DMSA is aimed at visualizing parenchymal abnor- 20–50% of children with UTI. It is also known that VUR can
malities and quantifying the parenchymal functional be found in about 40% of asymptomatic children whose sib-
mass, thus also assessing the degree of functional lings have VUR.
impairment.
30.7.1 Clinical Indication
30.7 Radionuclide Cystography Fluoroscopic voiding cystourethrography is used as a first-

for Vesicoureteral Reflux line diagnostic imaging procedure in male children, because
it can evaluate both the degree of VUR and the presence of
Vesicoureteral reflux (VUR) is the retrograde flow of urine posterior urethral valves. On the other hand, in female chil-
from the bladder into the ureter or into the kidney-calyceal dren, radionuclide cystography may be used as the first-
renal system. It can be observed in about 1–2% of the gen- choice imaging examination, because congenital anomalies
eral population, even in the absence of typical symptoms. associated with VUR are very rare in females. However, gen-
In children, primary VUR is due to a congenital defect of erally radionuclide cystography constitutes an important tool
the structure but also of function of the bladder-ureteral junc- for follow-up of patients with VUR (e.g., after surgical cor-
tion mechanism which can be associated with short or com- rection of VUR), since it is characterized both by high diag-
pletely absent intravesical ureter, absence of adequate nostic accuracy and by very low absorbed radiation dose to
detrusor support, lateral displacement of the ureteral orifice, the gonads. Moreover, radionuclide cystography is used for
and abnormal ureteral orifice. Primary VUR occurs in spite assessing changes in the VUR pattern in patients with neuro-
of a normal profile of the bladder filling pressure. Instead, genic bladder.
secondary VUR is caused by any of the obstructive bladder
diseases that can create excessively high filling and empty-
ing pressures. 30.7.2 Patient Preparation
In the pediatric population, the most common anatomical
obstruction to the bladder outlet is constituted by valves in Full collaboration from the patient is crucial to obtain a
the posterior urethra (which is observed almost exclusively good-quality examination. The environment should be
in male children). Another cause of reflux in girls may be comfortable and relaxing, so to put the patient at ease.
obstruction of the bladder neck or a ureterocele that pro- Direct radionuclide cystography requires infusion of the
lapses in the bladder neck. Other anatomical causes may be radiopharmaceutical into the urinary bladder through a trans-
the complete double renal district or the Hutch paraureteral urethral catheter [18].
Either 99mTc-DTPA, 99mTc-MAG3, or 99mTc-colloid (activ- too rapid distension of the bladder wall possibly eliciting the
ity of 37 MBq) are commonly used, all such radiopharma- urination reflex before adequate bladder filling is reached.
ceuticals being nonabsorbable through the bladder mucosa. During the bladder filling phase, a dynamic acquisition is
recorded, usually with 5-s frames.
The infusion ends when the patient expresses the need to
30.7.3 Technique urinate, when the volume of theoretical bladder filling is
reached, or when extravasation of urine around the catheter
A transurethral bladder catheter of appropriate caliber is appears [19]. At this point the transurethral catheter is removed
inserted, connected via a tubing to a bottle of saline (250 or (to obtain a “physiological” bladder emptying), and the void-
500 mL, possibly warmed up at 37 °C), which is placed on a ing phase of the study begins. Dynamic acquisition of the void-
drip shaft at a height of 70–100 cm from the level of the blad- ing phase can be performed with higher temporal sampling
der, in order to obtain bladder filling by gravity. (frames of 1–2 s) and is often accomplished in a few minutes.
The patient is positioned supine on the imaging table of a
LFOV gamma camera equipped with LEGP or LEHR colli-
mators, including in the FOV the symphysis pubis and the 30.7.4 Image Interpretation
xiphoid. Some centers prefer to have the patient sitting on a
large plastic container, with the head of the detector rotated The dynamic sequence acquired during bladder filling is
outwards on which the back of the patient is positioned (i.e., evaluated in “cine” mode display, looking for the presence of
in the same position as during the final bladder emptying VUR that appears as a streak of activity that goes up along
phase). the course of one or both ureters (Fig. 30.16).
The radiopharmaceutical can be mixed with saline within According to the results of radionuclide cystography,
the bottle or (more frequently) directly injected into the cath- VUR is classified with a 3-degree level:
eter when the infusion of saline starts. Bladder filling should
be not too fast (usually over about 15 min) in order to avoid • Grade 1 or mild, limited to the ureter (Fig. 30.16).
filling phase
emptying phase
Fig. 30.16 Images corresponding to the filling and emptying phases detected during the emptying phase (modified from: Volterrani D, Erba
recorded during radionuclide cystography. During the filling phase, a PA, Mariani G, Eds. Fondamenti di Medicina Nucleare – Tecniche e
mild and intermittent right VUR (grade 1) can be detected, which is not Applicazioni. Milan: Springer Italy; 2010)
Ph:1 Fr:1-2 643cts 0sec Ph:1 Fr:3-4 1338cts Ph:1 Fr:5-6 2074cts Ph:1 Fr:7-8 2962cts Ph:1 Fr:9-10 3811cts Ph:1 Fr:11-12 4660cts Ph:1 Fr:13-14 5433cts Ph:1 Fr:15-16 5996cts
Duration:10sec 64×64 10sec Duration:10sec 20sec Duration:10sec 30sec Duration:10sec 40sec Duration:10sec 50sec Duration:10sec 60sec Duration:10sec 70sec Duration:10sec
Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m
Technetium Technetium Technetium Technetium Technetium Technetium Technetium Technetium
Ph:1 Fr:17-18 6645cts Ph:1 Fr:19-20 6945cts Ph:1 Fr:21-22 7701cts Ph:1 Fr:23-24 8337cts Ph:1 Fr:25-26 9043cts Ph:1 Fr:27-28 9624cts Ph:1 Fr:29-30 9790cts Ph:1 Fr:31-32 9908cts
80sec Duration:10sec 90sec Duration:10sec 100sec Duration:10sec 110sec Duration:10sec 120sec Duration:10sec 130sec Duration:10sec 140sec Duration:10sec 150sec Duration:10sec
64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m 64×64 Pix:4.8mm 99m
Ph:1 Fr:33-34 11K Ph:1 Fr:35-36 11K Ph:1 Fr:37-38 12K Ph:1 Fr:39-40 10K Ph:1 Fr:41-42 9329cts Ph:1 Fr:43-44 7926cts Ph:1 Fr:45-46 6606cts Ph:1 Fr:47-48 5678cts
160sec Duration:10sec 170sec Duration:10sec 180sec Duration:10sec 190sec Duration:10sec 200sec Duration:10sec 210sec Duration:10sec 220sec Duration:10sec 230sec Duration:10sec
Ph:1 Fr:49-50 4677cts Ph:1 Fr:51-52 4114cts Ph:1 Fr:53-54 4031cts Ph:1 Fr:55-56 3955cts Ph:1 Fr:57-58 4038cts Ph:1 Fr:59-30 4010cts Ph:1 Fr:61-62 4079cts Ph:1 Fr:63-64 4048cts
240sec Duration:10sec 250sec Duration:10sec 260sec Duration:10sec 270sec Duration:10sec 280sec Duration:10sec 290sec Duration:10sec 5Min Duration:10sec 5.2Min Duration:10sec
Fig. 30.17 Radionuclide cystography in a patient with bilateral grade 3 VUR, detected both during the filling phase and during the emptying
phase
• Grade 2 or moderate, involving the ureter and pelvis, but 30.8 [18F]FDG PET/CT for Tumors
without dilation of the pelvis. of the Nephro-Urinary Tract
• Grade 3 or severe, when the pelvis and/or ureter(s) appear
dilated (Fig. 30.17). The role of [18F]FDG PET/CT in the evaluation of primary
renal tumors is limited. In fact, the physiologic renal excre-
The report should describe the presence, number, side tion of [18F]FDG into the urinary makes it difficult to reliably
and degree of the episodes of reflux, and should specify if visualize renal cancers that accumulate [18F]FDG [20].
VUR occurs in the filling or in the voiding phase of the Moreover, most renal tumors display an intrinsically low
study [19]. [18F]FDG avidity, which is responsible for the low sensitivity
of [18F]FDG PET/CT for diagnosing renal cancers. Some
[18F]FDG uptake is usually observed in type II papillary
Key Learning Points renal cancers, especially in the case of the hereditary leio-
• VUR is an important factor predisposing to acute myomatosis and renal cell cancer syndrome. It should also
pyelonephritis in patients with urinary tract infections. be considered that, while false-negative results can be due to
• Direct radionuclide cystography requires infusion urinary activity obscuring the tumor lesion, false-positive
of a radiopharmaceutical into the bladder. results can occur as well—due to focal accumulation of
• Direct radionuclide cystography is important for the radioactive urine that can mimic a renal lesion. On the other
characterization and follow-up of patients with VUR, hand, the use of forced diuresis coupled with parenteral
showing a high diagnostic accuracy and implying a hydration in order to enhance urinary excretion of [18F]FDG
very low absorbed radiation dose to the gonads. has not resulted in improved diagnostic accuracy of PET/CT
for the evaluation of primary renal tumors.
Notwithstanding the heavy limitations of [18F]FDG comparison with [18F]FDG uptake and correlated with
PET/CT in the identification of primary renal cancers, this immunohistochemical analysis of cell proliferation based
imaging procedure is very useful for staging or restaging. on the Ki-67 tissue marker [22]. A strong correlation was
In particular, the identification of distant metastases in found between both 18F-FLT and [18F]FDG uptake, both
high-risk tumors, the characterization of indeterminate being closely correlated with Ki-67 expression; these find-
lesions on other diagnostic imaging techniques, and the ings suggest that PET with 18F-FLT may allow for predic-
staging/restaging of patient with solitary metastasis, in tion of more aggressive phenotypes of RCC.
whom surgical resection could be considered, represent the
main clinical indications of [18F]FDG PET/CT in patients
with renal cancers. The role of [18F]FDG PET/CT is increas- 30.9.2 18F-FMISO
ing also for the evaluation of response to therapy in patients
with metastatic disease treated with novel targeted thera- Radiolabeled nitromidazoles, such as 118F-fluoromisonidazole
pies [21]. (18F-MISO), are emerging as attractive probes for noninva-
Due to the same considerations as described above for sive in vivo assessment of tissue hypoxia, as they are trapped
renal cancers, the value of [18F]FDG PET/CT for detecting in hypoxic, yet metabolically active cell environments. The
localized tumors in the bladder (and in the whole urinary use of 18F-MISO has been explored also for PET imaging of
tract) is limited due to urinary excretion of radiotracer that RCC, considering that this tumor is relatively resistant to
masks tumor activity. However, PET/CT has a definite role radiation and chemotherapy, thus behaving as a tumor con-
in detecting lymph node metastases, since [18F]FDG may taining an important hypoxic cell population. However,
detect enhanced metabolic activity in normal-sized lymph unlike what observed in other solid tumors, early reports
nodes. Although few data are available on the utility of [18F] show only modest 18F-MISO uptake in RCCs, not predictive
FDG PET/CT in assessing recurrences and metastatic dis- of response to therapy [23].
ease in patients previously submitted to treatment for pri-
mary bladder cancer, PET/CT can detect distant metastases
and can help in identifying pelvic recurrences as well. 30.9.3 124I-cG250
Over 95% of clear cell renal cancers express carbonic anhy-

Key Learning Points drase IX (CAIX) on their cell membrane. High CAIX
• The role of [18F]FDG PET/CT in the evaluation of expression has been demonstrated also in other malignan-
primary renal and bladder tumors is limited. cies, such as ovarian, colorectal, lung, brain, and bladder
• [18F]FDG PET/CT is primarily useful for diagnos- cancers. CAIX expression is associated with hypoxia and
ing extrarenal sites of disease such as lymph node appears to predict poor prognosis [24, 25]. The cG250
or distant metastases. monoclonal antibody was developed as a potential therapeu-
• The role of [18F]FDG PET/CT for evaluation of tic agent, but results of early trials showed only modest or
therapy response in patients with metastatic disease no benefit at all [26].
treated with novel targeted therapies is promising. Immuno-PET is based on radiolabeling with positron
• Similar considerations apply to tumors of the blad- emitters of monoclonal antibodies that bind to an antigen
der and urinary tract. expressed on tumor cells. The physical half-life of common
PET radionuclides, such as 11C (20.4 min) and 18F (109.7 min),
is in general too short for the in vivo pharmacokinetics of
monoclonal agents. The 4.2-day half-life of 124I is better
30.9 P
ET Tracers Other Than [18F]FDG suited to the slow pharmacokinetics of monoclonal tracers,
for Renal Cancers since the peak of uptake of such antibody conjugates in
tumors is normally several days, at least for intact immuno-
The diagnostic potential of number of non-[18F]FDG PET globulins [27]. Recently, the 124I-labeled antibody chimeric
tracers has been explored in patients with renal cancer, G250 (124I-cG250) has been used to evaluate primary RCC
although generally with studies in small series of patients. [28]. 124I-cG250 PET is promising in distinguishing clear cell
RCC from other subtypes, a clinically relevant finding con-
sidering that preoperative identification of tumor type has
30.9.1 18F-Fluorothymidine important implications for treatment planning of renal can-
cers. Thus, 124I-cG250 PET may be a useful diagnostic imag-
The diagnostic role of 18F-fluorothymidine (18F-FLT), a ing tool in the clinical management of renal masses. Imaging
PET radiotracer employed as a cell proliferation marker, with 124I-cG250 PET has the potential to change clinical
has been explored in a group of 27 patients with RCC in management in patients in whom a conservative approach
(because of renal function impairment or comorbid disease) 30.10 P

ET Tracers Other Than [18F]FDG
may be appropriate in G250-negative tumors, as these for Bladder Cancer
patients can be selected for nephron-sparing surgery or for
follow-up observation. In this regard, the 124I-cG250 PET [11C]Choline, [11C]methionine, and [11C]acetate have been
scan may be useful as an alternative to biopsy. The study proposed for improved PET imaging in patients with bladder
with 124I-cG250 PET requires 3–7 days because of the long cancer considering that, unlike [18F]FDG, these tracers are
clearance time of the antibody, thus making this approach not excreted in the urine.
impractical. In this regard, replacement of 124I with 89Zr
might result in better imaging, although the actual clinical
role of this agent has not been clarified yet [29, 30]. 30.10.1 [11C]Choline
[11C]Choline, whose uptake is highly correlated with the

30.9.4 F-Fluoride
18
rate of tumor cell proliferation, is cleared very rapidly from
the blood, and optimal tumor-to-background ratios are
One further PET tracer investigated in patients with RCC is reached within 5–7 min. De Jong et al. investigated the
18
F-fluoride, a much more sensitive tracer that the 99mTc- potential of [11C]choline PET in 18 patients with bladder
diphosphonates used for conventional bone scintigraphy. cancer compared to five healthy volunteers [32]. In the nor-
This assumption has been proved to be true by recent reports mal bladder wall, there was low uptake of [11C]choline, and
indicating that 18F-fluoride PET/CT is more sensitive than the bladder margin was only outlined by minimal urinary
conventional bone scintigraphy even for predominantly lytic radioactivity, if present. The primary tumor was correctly
bone metastases from renal cancers [31]. detected by [11C]choline PET in 10/18 patients (mean SUV
4.7 ± 3.6, range 1.5–13.0). No [11C]choline uptake was
seen in premalignant lesions or in small noninvasive
tumors.
Also two recent studies investigated the role of [11C]cho-
Key Learning Points line PET/CT in the preoperative staging of patients’ bladder
• Different PET tracers other than [18F]FDG have cancer. Gofrit et al. evaluated prospectively 18 patients with
been investigated in patients with urological can- advanced TCC (17 bladder tumors and two upper tract tran-
cers, mostly with the purpose of overcoming the sitional carcinomas) [33]. [11C]Choline uptake was found in
limitations associated with physiologic [ F]FDG 18 all primary tumors with a maximum SUV of 7.3 ± 3.2
excretion through the kidneys. (mean ± SD). The series included three patients with
• The prognostic potential of 18F-fluorothymidine, a refractory bladder carcinoma in situ (CIS), which was visu-
cell proliferation marker, has been explored in alized in all three patients on [11C]choline PET (SUV
patients with RCC, suggesting that PET with this 6.9 ± 5.6). In six patients, uptake of [11C]choline in lymph
tracer may predict biological aggressiveness. nodes as small as 5 mm was visualized (SUV 3.8 ± 14), and
• Disappointing results have been reported in a pilot the PET scan detected bone metastases in four patients.
study exploring the potential of PET with the Thus, [11C]choline PET was highly sensitive for both pri-
hypoxia agent F-fluoromisonidazole to predict
18 mary and metastatic transitional carcinomas. [11C]Choline
response to therapy in patients with RCC, a tumor PET was used also by Picchio et al. in patients referred for
that is relatively resistant to radiation and radical cystectomy and pelvic lymph node dissection [34].
chemotherapy. The presence of residual bladder cancer was correctly
• The most promising results concerning an alterna- detected in 21/25 patients by CT and in 24/25 patients by
tive to [18F]FDG for distinguishing clear cell RCC [11C]choline PET. Lymph node metastasis was correctly
from other subtypes have been obtained with detected in 4/8 patients by CT and in 5/8 patients by [11C]
immuno-PET based on the use of I-cG250, a 124 choline PET. Thus, [11C]choline PET was comparable to CT
monoclonal tracer against carbonic anhydrase IX, for detecting residual bladder cancer after TURB but was
which is highly expressed on renal cancer cells. superior to CT for the evaluation of potential additional
• Similarly as with other solid cancers, PET with lymph node metastases. However, larger prospective studies
18
F-fluoride is particularly attractive for its ability to are needed to define the actual role of [11C]choline in preop-
detect skeletal metastases with a predominant lytic erative staging of bladder cancer. Furthermore, yet no studies
component, as it occurs in RCC patients. have compared the performance of [11C]choline to that of
[18F]FDG for staging bladder cancer patients.
30.10.2 [11C]Methionine
Key Learning Points
The uptake of [11C]methionine in tissues is proportional to • There is limited experience with the use of non-
the amino acid transport and, to some extent, to protein syn- [18F]FDG PET tracers (such as [11C]choline, [11C]
thesis. Tumor uptake of [11C]methionine, which is correlated methionine, and [11C]acetate) in patients with blad-
with the amount of viable cancer tissue, occurs rapidly (with der cancer; further studies are necessary for validat-
a peak at around 10 min, followed by a plateau) [35]. [11C] ing the use of these tracers.
Methionine undergoes rapid clearance from the blood and is • In a prospective study, PET/CT with 18F-fluoride
metabolized primarily in the liver and pancreas, with no sig- performed better than bone scintigraphy with 99mTc-
nificant renal excretion MDP (both planar and SPECT/CT) and changed
PET with [11C]methionine had better diagnostic perfor- clinical management in about one out of three
mance than [18F]FDG PET, the tumor being identified in patients with bladder cancer.
18/23 patients only with [11C]methionine PET [36]. Although
the intensity of [11C]methionine uptake was correlated to
tumor grade, [11C]methionine PET did not improve staging
of bladder cancer. In another study, methionine PET was References
used for the evaluation of response to therapy in 44 patients
with different stages of bladder cancer treated with chemo- 1. Blaufox MD. The current status of renal radiopharmaceuticals. A
review. Contrib Nephrol. 1987;56:31–7.
therapy [37]. In this study the diagnostic accuracy of [11C] 2. Durand E, Prigent A. The basics of renal imaging and function
methionine PET turned out to be rather poor, and the tech- studies. Q J Nucl Med. 2002;46:249–67.
nique could not reliably assess the therapeutic efficacy of 3. Wedeen RP, Blaufox MD. The normal renogram. In: Blaufox MD,
neoadjuvant chemotherapy. editor. Evaluation of renal function and disease with radionuclides:
the upper urinary tract. 2nd ed. Basel: Karger; 1989. p. 116–29.
4. Zuckier LS, Freeman LM. Principles of renal imaging with radio-
nuclides. In: Blaufox MD, editor. Evaluation of renal function and
30.10.3 [11C]Acetate disease with radionuclides: the upper urinary tract. 2nd ed. Basel:
Karger; 1989. p. 130–48.
5. Peters AM, George P, Ballardie F, Gordon I, Todd-Pokropek
[11C]Acetate is another PET tracer with little or no urinary A. Appropriate selection of background for 99mTc-DTPA renogra-
excretion. A study in 16 patients prospectively evaluated phy. Nucl Med Commun. 1988;9:973–85.
MRI, [11C]acetate PET/CT, and CT for staging of bladder 6. Hilson AJ, Maisey MN, Brown CB, Ogg CS, Bewick MS. Dynamic
cancer [38]. MRI, [11C]acetate PET/CT, and CT exhibited renal transplant imaging with 99mTc-DTPA (Sn) supplemented by a
transplant perfusion index in the management of renal transplants.
similar levels of accuracy. For all modalities, a history of J Nucl Med. 1978;19:994–1000.
intravesical and/or systemic chemotherapy affected the accu- 7. Russell CD, Japanwalla M, Khan S, Scott JW, Dubovsky
racy of staging. In another small study in 14 patients, [11C] EV. Techniques for measuring renal transit time. Eur J Nucl Med.
acetate PET/CT was compared with [11C]choline PET/CT in 1995;22:1372–8.
8. Gates GF. Computation of glomerular filtration rate with 99mTc-
patients with bladder cancer [39]. The two PET tracers dem- DTPA: an in-house computer program. J Nucl Med. 1984;25:613–8.
onstrated equivalent diagnostic performance. 9. Blaufox MD, Aurell M, Bubeck B, Fommei E, Piepsz A, Russell C,
et al. Report of the Radionuclides in Nephrourology Committee on
renal clearance. J Nucl Med. 1996;37:1883–90.
10. Brown SCW. Nuclear medicine in the clinical diagnosis and

30.10.4 18
F-Fluoride treatment of obstructive uropathy. In: Murray IPC, Ell PJ, edi-
tors. Nuclear medicine in clinical diagnosis and treatment, vol. 1.
It can reasonably be assumed that lytic bone metastases Edimburgh: Churchill Livingstone; 1994. p. 271–93.
from BC may be better detected using 18F-fluoride PET 11. O'Reilly P, Aurell M, Britton K, Kletter K, Rosenthal L, Testa
T. Consensus on diuresis renography for investigating the dilated
than conventional bone scintigraphy with 99mTc-diphospho- upper urinary tract. Radionuclides in Nephrourology Group.
nates. Recently, the diagnostic performance of 18F-fluoride Consensus Committee on Diuresis Renography. J Nucl Med.
PET was compared with 99mTc-MDP bone scintigraphy for 1996;37:1872–6.
the detection of skeletal metastases in 48 bladder cancer 12. Safian RD, Textor SC. Renal artery stenosis. New Engl J Med.
2001;344:431–42.
patients [40]. In this prospective study, 18F-fluoride PET/ 13. Taylor A. Functional testing: ACEI renography. Semin Nephrol.
CT was superior to both planar and SPECT/CT bone scin- 2000;20:437–44.
tigraphy and changed clinical management in 17/48 14. Fommei E, Volterrani D. Renal nuclear medicine. Semin Nucl Med.
patients. 1995;25:183–94.
15. Fommei E, Ghione S, Hilson A, Mezzasalma L, Oei HY, Piepsz A, 29. Czarnecka AM, Kornakiewicz A, Kukwa W, Szczylik C. Frontiers
et al. Captopril test radionuclide in renovascular hypertension: a in clinical and molecular diagnostics and staging of metastatic clear
multicentre European study. Eur J Nucl Med. 1993;20:617–23. cell renal cell carcinoma. Future Oncol. 2014;10:1095–111.
16. Gordon I. Indications for 99mTechnetium dimercapto-succinic acid 30. Cheal SM, Punzalan B, Doran MG, Evans MJ, Osborne JR, Lewis
scan in children. J Urol. 1987;137:464–7. JS, et al. Pairwise comparison of 89Zr- and 124I-labeled cG250 based
17. Cao X, Xu X, Grant FD, Treves ST. Estimation of split renal on positron emission tomography imaging and nonlinear immuno-
function with 99mTc-DMSA SPECT: comparison between 3D kinetic modeling: in vivo carbonic anhydrase IX receptor binding
volumetric assessment and 2D coronal projection imaging. AJR and internalization in mouse xenografts of clear-cell renal cell car-
Am J Roentgenol. 2016;207:1324–8. cinoma. Eur J Nucl Med Mol Imaging. 2014;41:985–94.
18. Conway JJ. Radionuclide cystography. In: Tauxe WN, Dubovsky 31. Fuccio C, Spinapolice EG, Cavalli C, Palumbo R, D’Ambrosio D,
EV, editors. Nuclear medicine in clinical urology. East Norwalk: Trifirò G. 18F-Fluoride PET/CT in the detection of bone metastases
Appleton, Century and Crofts; 1985. p. 305–20. in clear cell renal cell carcinoma: discordance with bone scintigra-
19. Fettich J, Colarinha P, Fischer S, Frökier J, Gordon I, Hahn K, et al. phy. Eur J Nucl Med Mol Imaging. 2013;40:1930–1.
Guidelines for direct radionuclide cystography in children. Eur J 32. de Jong IJ, Pruim J, Elsinga PH, Jongen MM, Mensink HJ, Vaalburg
Nucl Med Mol Imaging. 2003;30:BP39–44. W. Visualisation of bladder cancer using 11C-choline PET: first clin-
20. Scott AMD. Clinical role of 18F-fluorodeoxyglucose positron emis- ical experience. Eur J Nucl Med Mol Imaging. 2002;29:1283–8.
sion tomography for detection and management of renal cell carci- 33. Gofrit ON, Mishani E, Orevi M, Klein M, Freedman N, Pode D,
noma. J Urol. 2001;166:825–30. et al. Contribution of 11C-choline positron emission tomography/
21. Bouchelouche K, Choyke PL. PET/computed tomography in renal, computerized tomography to preoperative staging of advanced
bladder, and testicular cancer. PET Clin. 2015;10:361–74. transitional cell carcinoma. J Urol. 2006;176:940–4.
22. Wong P, Bolton DM, Lee STLN, Davis ID, Scott AM. In-vivo 34. Picchio M, Treiber U, Beer AJ, Metz S, Bössner P, van Randenborgh
imaging of cellular proliferation in renal cell carcinoma using H, et al. Value of 11C-choline PET and contrast-enhanced CT for
18
F-FLT PET. J Urol. 2009;181(Suppl 4):155. staging of bladder cancer: correlation with histopathologic find-
23. Ammari S, Thiam R, Cuenod CA, Oudard S, Hernigou A, Grataloup ings. J Nucl Med. 2006;47:938–44.
C, et al. Radiological evaluation of response to treatment: applica- 35. Jana S, Blaufox MD. Nuclear medicine studies of the prostate, tes-
tion to metastatic renal cancers receiving anti-angiogenic treatment. tes, and bladder. Semin Nucl Med. 2006;36:51–72.
Diagn Interv Imaging. 2014;95:527–39. 36. Ahlstrom H, Malmstrom PU, Letocha H, Andersson J, Langstrom
24. Oosterwijk E. Carbonic anhydrase IX: historical and future per- B, Nilsson S. Positron emission tomography in the diagnosis and
spectives. BJU Int. 2008;101(Suppl 4):2–7. staging of urinary bladder cancer. Acta Radiol. 1996;37:180–5.
25. Jensen HK, Nordsmark M, Donskov F, Marcussen N, von der 37. Letocha H, Ahlström H, Malmström PU, Westlin JE, Fasth KJ,
MH. Immunohistochemical expression of carbonic anhydrase IX Nilsson S. Positron emission tomography with L-methyl-11C-
assessed over time and during treatment in renal cell carcinoma. methionine in the monitoring of therapy response in muscle-
BJU Int. 2008;101(Suppl 4):41–4. invasive transitional cell carcinoma of the urinary bladder. Br J
26. Divgi CR, Uzzo RG, Gatsonis C, Bartz R, Treutner S, Yu JQ, et al. Urol. 1994;74:767–74.
Positron emission tomography/computed tomography identifica- 38. Vargas HA, Akin O, Schoder H, Olgac S, Dalbagni G, Hricak
tion of clear cell renal cell carcinoma: results from the REDECT H, et al. Prospective evaluation of MRI, 11C-acetate PET/CT and
trial. J Clin Oncol. 2013;31:187–94. contrast-enhanced CT for staging of bladder cancer. Eur J Radiol.
27. Brouwers A, Verel I, Van Eerd J, Visser G, Steffens M, Oosterwijk 2012;81:4131–7.
E, et al. PET radioimmunoscintigraphy of renal cell cancer using 39. Orevi M, Klein M, Mishani E, Chisin R, Freedman N, Gofrit ON.
89Zr-labeled cG250 monoclonal antibody in nude rats. Cancer 11
C-acetate PET/CT in bladder urothelial carcinoma: intraindivid-
Biother Radiopharm. 2004;19:155–63. ual comparison with 11C-choline. Clin Nucl Med. 2012;37:e67–71.
28. Divgi CR, Pandit-Taskar N, Jungbluth AA, Reuter VE, Gönen M, 40. Chakraborty D, Bhattacharya A, Mete UK, Mittal BR. Comparison
Ruan S, et al. Preoperative characterisation of clear-cell renal car- of 18F-fluoride PET/CT and 99mTc-MDP bone scan in the detec-
cinoma using iodine-124-labelled antibody chimeric G250 (124I- tion of skeletal metastases in urinary bladder carcinoma. Clin Nucl
cG250) and PET in patients with renal masses: a phase I trial. Med. 2013;38:616–23.
Lancet Oncol. 2007;8:304–10.
Functional Imaging for Benign
Conditions of the Gastrointestinal Tract 31
Gayane Aghakhanyan, Elisa Fiasconaro, Elisa Tardelli,
Mariano Grosso, and Italia Paglianiti
Contents
31.1 Salivary Gland Scintigraphy 810
31.1.1 Anatomy and Physiology of Salivary Glands 810
31.1.2 Radiopharmaceuticals and Technique 810
31.1.3 Clinical Indications for Salivary Gland Scintigraphy 811
31.2 Oropharyngeal-Esophageal Transit Scintigraphy 812
31.2.1 Anatomy and Physiology 812
31.2.2 Radiopharmaceuticals, Technique and Image Processing 812
31.2.3 Clinical Indications 815
31.3 Gastric Emptying Scintigraphy 816
31.3.2 Clinical Indications and Contraindications 817
31.3.3 Radiopharmaceuticals, Image Acquisition, Data Processing, and Reporting 817
31.3.4 Future Perspectives 817
31.4 Small-Bowel and Colon Transit Scintigraphy 818
31.4.2 Clinical Indications and Contraindications 821
31.4.3 Radiopharmaceuticals, Image Acquisition, Data Processing, and Reporting 823
31.4.4 Small-Bowel Transit Scintigraphy 823
31.4.5 Colon Transit Scintigraphy 825
31.5 Gastrointestinal Bleeding Scintigraphy 826
31.5.1 Radiopharmaceuticals and Technique......... 827
31.6 Meckel’s Diverticulum Scintigraphy 827
31.6.1 Normal Anatomy 829
31.7 Hepatobiliary Scintigraphy 829
References 838
G. Aghakhanyan
E. Fiasconaro · E. Tardelli · M. Grosso (*) I. Paglianiti
Regional Center of Nuclear Medicine,
University Hospital of Pisa, Pisa, Italy
e-mail: m.grosso@med.unipi.it

810 G. Aghakhanyan et al.
Learning Objectives 31.1.1 Anatomy and Physiology of Salivary

• Understand the rationale of exploring function of the Glands
major salivary glands using radionuclide imaging.
• Describe the methodology of performing salivary gland The parotid, submandibular, and sublingual glands are the
scintigraphy. major salivary glands based on their anatomic, histologi-
• Understand how to analyze and interpret the results of cal, and functional characteristics. The parotid is the larg-
salivary gland scintigraphy. est gland overlaying the masseter muscle. The parotid ductal
• Understand the main pathophysiologic mechanisms system has a tree-like branching pattern composed of mul-
underlying progression of ingested meals from the very tiple smaller ducts emptying into the main Stensen’s duct
initial swallowing act in the oropharynx to esophageal (7 cm in length). The stylomandibular ligament separates the
transit, gastric filling and emptying, small-bowel transit, parotid from the submandibular gland, which is the second
and finally colonic transit out to evacuation. largest salivary gland and lies in the submandibular triangle.
• Summarize the main abnormalities possibly affecting The submandibular duct, or Wharton duct, is about 5 cm
progression of the alimentary bolus through each portion long [2]. The almond-shaped sublingual gland is the small-
of the gastrointestinal tract. est one that lies deep to the floor of mouth. Unlike the parotid
• List the indications for scintigraphic transit studies of the and submandibular glands, the sublingual gland has no true
different portions of the gastrointestinal tract. fascial capsule and a main duct. Instead, it is drained by
• Describe the methodology for performing scintigraphic approximately ten small ducts (the ducts of Rivinus) along
transit studies of the different portions of the gastrointes- the sublingual fold on the floor of mouth [3].
tinal tract.
• Understand how to analyze and interpret the scinti-
graphic findings obtained during scintigraphic transit 31.1.2 Radiopharmaceuticals and Technique
studies of the different portions of the gastrointestinal
tract. Since 99mTc-pertechnetate is taken up by salivary glands and
• Summarize the most common causes of gastrointestinal secreted by the ductal epithelium, it is commonly used for
bleeding and the indications to perform a gastrointestinal salivary gland scintigraphy. The usual adult administered
bleeding study. activity is 296–444 MBq intravenously. Administered activ-
• Describe the methodology for performing a gastrointesti- ity for children should be determined based on body weight
nal bleeding study. and should be as low as reasonably achievable for diagnostic
• Understand how to analyze and interpret the results of a image quality. A gamma camera with a small field of view
gastrointestinal bleeding study. equipped with a low-energy, high-resolution, parallel-hole
• Summarize the causes and clinical manifestations associ- collimator is used. The energy window is centered on the
ated with the presence of heterotopic gastric mucosa in 140 keV photo peak of 99mTc.
Meckel’s diverticulum. Patients are often told to fast and stop smoking prior to
• Describe the methodology for performing a Meckel’s examination. No thyroid-blocking agents (iodide or perchlo-
diverticulum scintigraphy. rate) should be administered within 48 h before the scan.
• Understand how to analyze and interpret the results of A bolus of 99mTc-pertechnetate is injected intravenously.
Meckel’s diverticulum scintigraphy. Dynamic short-interval sequential images of about 1–2 min
• Describe the diagnostic uses of hepatobiliary scintigraphy (1 s/frame) are first acquired as perfusion phase, followed by
and methodology for performing cholescintigraphy for 20 min of dynamic imaging for 2–3 min/frame to assess func-
various clinical indications. tion of the glands. Trapping of the 99mTc-pertechnetate should
• Understand the rationale and practice of hepatobiliary occur within 1 min of intravenous injection, with peak activ-
scintigraphy in the current clinical routine. ity within 10–21 min. Once the first dynamic study is com-
pleted, the excretory aspect of the salivary glands is assessed.
A sialagogue (i.e., lemon drops, ascorbic acid, or lemon
31.1 Salivary Gland Scintigraphy juice) is administered to stimulate salivary gland excretion,
which causes prompt salivation if the salivary ducts are pat-
Radionuclide imaging of salivary glands is a noninvasive ent. The excretion and re-accumulation phases are then fol-
functional imaging modality that can assess perfusion, con- lowed. Regions of interests (ROIs) are drawn around parotid
centration, and excretory function of major salivary glands. and submandibular glands and computer-assisted analysis of
The technique was first introduced by Boerner et al. in 1965 time-activity (T/A) curves are generated. Time of administra-
and subsequently modified by others [1]. tion of the sialogogue is marked on the curves [4].
31 Functional Imaging for Benign Conditions of the Gastrointestinal Tract 811
a b 100
c
90
80
Right roi2 Left 70
roi3
parotid parotid
gland gland 60
counts/sec
roi1
Right roi4 Left 50
submandibular submandibular
gland gland 40
30
20
10
0
0 10 20 30 40 50
min
Fig. 31.1 Patient with Sjögren syndrome complained of dry mouth. the submandibular and parotid glands show decreased accumulation of
(a) Anterior color-coded composite image of a dynamic 99mTc- the tracer in the parotid glands and complete failure of submandibular
pertechnetate salivary gland scan. (b) ROIs drawn on the major salivary glands to accumulate the tracer. Response of stimulation with lemon
glands. (c) Corresponding color-coded time/activity (T/A) curves from juice is normal
31.1.3 Clinical Indications for Salivary Gland be sufficiently sensitive to determine the severity of the dam-
Scintigraphy age [8]. Approximately 10% of patients experience a steep
decline in stimulated whole saliva flow rate after single high-
The common clinical applications of salivary gland scintig- activity radioiodine ablation therapy [9].
raphy are acute and chronic sialadenitis, sialolithiasis, dry Although salivary gland scintigraphy may be used
mouth, and Sjögren syndrome. Sialadenitis is the most com- for evaluation of suspected mass lesions of the salivary
mon pathology affecting the salivary glands [5]. In acute glands, the diagnostic approach in these patients is best
sialadenitis (of bacterial or viral origin), an increase in radio- done by using computed tomography (CT) and magnetic
nuclide uptake and retention is seen, due to hyperemia and resonance imaging (MRI). Most mass lesions, including
edema compressing the intralobular ducts. In chronic sial- primary tumors and metastases, usually show increased
adenitis, a variable radiotracer uptake may be seen depend- perfusion and decreased uptake that varies according to
ing on the disease stage, with little or no uptake in the late cellular content of the tumor. The exception to this are
stage of the disease [5]. Sialolithiasis with recurrent and functioning salivary gland tumors (e.g., Warthin’s tumor,
chronic obstruction causes stasis, inflammation, and persis- oncocytomas, and oxyphilic adenomas) that generally
tent swelling of the gland. About 80–90% of salivary calculi tend to contain large number of oncocytes and are always
occur in the submandibular gland [5]. The scintigraphic pat- benign [10].
tern is characterized by poor or absent radiotracer excretion
after lemon juice administration; dilated ducts may also be
seen.
Salivary gland scintigraphy has been used to assess the Key Learning Points
extent of oral involvement in Sjögren syndrome (Fig. 31.1). • Salivary gland scintigraphy is a noninvasive func-
Semiquantitative results correlate with clinical and histo- tional imaging modality that allows assessment of
pathological grading of salivary glands [6], and some authors the perfusion, concentration, and excretion function
indicate this examination as a first-choice investigation in of major salivary glands.
patients with suspected primary Sjögren syndrome [7]. • The diagnostic information provided by salivary
Functional impairment of salivary glands is often seen in gland scintigraphy is especially useful in patients
patients with head and neck cancer following radiotherapy with acute and chronic sialadenitis, sialolithiasis,
or in patients treated with radioiodine therapy for differenti- dry mouth, and Sjögren syndrome, as well as after
ated thyroid cancer. Early dry mouth due to reduced salivary head and neck radiotherapy or in patients treated
gland function usually affects the salivary gland excretion with radioiodine therapy for differentiated thyroid
phase, with decreased trapping noted in the late phase of the cancer.
scan. The maximum secretion and uptake ratio were found to
31.2 Oropharyngeal-Esophageal Transit define the upper end of the esophagus. The lower end of the
Scintigraphy esophagus is defined anatomically by the junction of the
esophagus with the stomach and, functionally, by the junc-
Since Kazem et al. first introduced esophageal transit scin- tion of the inferior border of the high-pressure zone of the
tigraphy (ETS) in 1972 [11], the technique has been used LES compared to the lower pressure in the stomach.
to assess esophageal transit time and function. The ETS Under ideal physiologic conditions, the concentric mus-
was initially introduced to evaluate the swallowing func- cular configuration permits effortless unidirectional flow of
tion of the esophagus, transit through the various segments material from the top to the bottom of the esophagus. The
of the esophagus and esophageal patency [12]. As technol- UES, 4–5 cm in length, remains in a constant state of tone
ogy improved, more complex image analysis is possible. (mean, 60 mmHg), preventing a steady flow of air into the
Scintigraphic transit measures have been well validated in the esophagus, whereas the tone in the LES (mean, 24 mmHg)
esophagus, although at a lower extent in the oropharyngeal remains elevated just enough to prevent excessive material
phase of swallowing [13]. Currently, use of the ETS is lim- from refluxing back up from the stomach into the esopha-
ited despite its validation, in part due to the lack of consen- gus. Transport of a food bolus from the mouth through the
sus guidelines for performing the examination. In addition, esophagus into the stomach begins with swallowing and
esophageal manometry as performed by gastroenterologists ends with post-relaxation contraction of the LES. There are
has matured from a research tool to a more clinically avail- three phases of swallowing: oral, pharyngeal, and esopha-
able and standardized test, thus further limiting the use of geal. Rapid series of events occur during the oropharyngeal
the ETS [14]. In spite of that, the oropharyngeal-esophageal phase of swallowing, lasting about 1.5 s. Only the first phase
scintigraphy (OPES) is useful to investigate swallowing and is totally voluntary, whereas the subsequent phases, once
esophageal motility disorders. This method offers qualita- initiated, are completely reflexive. During the esophageal
tive and semiquantitative information concerning the various phase of swallowing, the UES relaxes, and the peristaltic
stages of swallowing (Fig. 31.2) and consequently a better contractions of the posterior pharyngeal constrictors propel
clinical and diagnostic definition of the diseases and of the the food bolus into the esophagus. As pressure in the cervical
anatomical-functional disorders that cause dysphagia [15]. esophagus is positive, while it is negative in the intrathoracic
region, such pressure difference sucks the bolus into the tho-
racic esophagus. Within 0.5 s of the initiation of swallowing,
31.2.1 Anatomy and Physiology the UES closes (reaching up to 90 mmHg) and remains in
the post-relaxation contraction phase during 2–5 ms, which
The oral cavity and pharynx are anatomic spaces defined by initiates peristalsis, and prevents reflux of the bolus back into
hard and soft tissue structures. The oral cavity is defined by the pharynx. The UES pressure returns to resting pressure
different anatomic structures: anteriorly the lips, laterally the (60 mmHg) as the wave travels into the mid-esophagus. The
cheeks, superiorly the hard palate, and inferiorly the mucosa last step of swallowing (the esophageal phase) involves the
covering the superior surface of the tongue and the sheet of bolus traveling a much longer distance but involves a much
muscles attaching to the inner side of the mandible. The oral simpler peristaltic mechanism, aided by gravity. The pri-
cavity is continuous with the pharyngeal cavity. In the buc- mary peristaltic pump originating in the pharynx propagates
cal cavity, the initial food shredding, mixing with saliva, and through the inner (circular) and the outer (longitudinal) mus-
mastication take place, which through swallowing is directed cular layers of the esophagus at about 4 cm/s. Coordination
to the esophageal canal. Pharynx is a muscular membranous of the primary peristaltic pump with post-swallowing relax-
organ and an important anatomical and functional crossroad ation of the LES lets the bolus enter the stomach. Liquids
between the digestive and respiratory tracts. The esophagus require approximately 2 s to traverse the esophagus, whereas
is a hollow muscular canal that varies in length with patient solids can take about 9 s [16].
height, an average 25 cm with a diameter of about 2.5 cm.
It extends from the lower border of the cricoid cartilage to
the gastroesophageal junction below the diaphragm. The dis- 31.2.2 Radiopharmaceuticals, Technique
tal end of the tubular esophagus is normally in the abdomen and Image Processing
and flares into the gastric fundus. Functionally, the esopha-
gus extends from the lower border of the upper esophageal 31.2.2.1 Patient Preparation
sphincter (UES) to the lower border of the lower esopha- The OPES should be performed after a fast of at least 3 h
geal sphincter (LES). Anatomically, the UES is part of the but preferably overnight. A test procedure is performed prior
pharynx and is appropriately termed the inferior pharyngeal to the investigation, inviting the patient to swallow nonra-
sphincter. Thus, anatomically, the lower border of the cri- dioactive water; this enables the patient to become famil-
coid cartilage and, functionally, the lower border of the UES iar with the overall procedure, enhances compliance, and
3.0
a b
2.0
kcounts/sec
oral
1.0
pharyngeal
0.0
0 10 20 30 40 50 60
sec
esophageal
c
1.0
kcounts/sec
0.0
0 10 20 30 40 50 60
sec
d
kcounts/sec
1.0
0.0
0 10 20 30 40 50 60
sec
10 sec
Fig. 31.2 Oropharyngo-esophageal transit scintigraphy (OPES) in a curves for oral transit (b), pharyngeal transit (c), and esophageal transit
healthy volunteer. (a) ROIs drawn on the oral cavity, pharynx, and (d). (e) Condensed image of the dynamic acquisition
esophagus after the dynamic acquisition. Corresponding time/activity
ensures that the patient is capable of swallowing the amount with a low-energy, high-resolution, parallel-hole collimator
of liquid expected to be used during the actual scan (about with the energy window set at 140 keV ± 10%) to include
10 mL). The investigation starts after approximately 5 min, in the field of view the entire region from the mouth to the
with the patient standing up and with his/her face in an 80° epigastric area; the patient’s face is tilted about 80° to the
oblique projection in front of the gamma camera (equipped left. The patient keeps in his/her mouth a 10-mL bolus of
water labeled with 37 MBq (1 mCi) of 99mTc-colloid, which sequence over 120 s, during which the patient is asked to per-
is swallowed 2 s after starting a dynamic acquisition for 60 s form 4–5 Valsalva maneuvers. In this manner, it is possible
(0.125 s/frame, matrix 64 × 64, zoom 1). At the end of the to detect the presence of gastroesophageal reflux because of
dynamic scan, with the patient still in the same position, a the increased intra-abdominal pressure [16].
60-s static image is acquired to detect any retention and tra-
cheobronchial aspirate. After 30 min, the entire swallowing 31.2.2.2 Q ualitative and Quantitative Image
test is repeated using a semisolid bolus represented by 10 mL Processing
of a jellied drink labeled with 37 MBq (1 mCi) of 99mTc- Reviewing the recorded sequence in the cine mode depicts
colloid followed by the 60-s static acquisition. Scintigraphic the dynamics of swallowing. This procedure helps to identify
acquisitions are performed in the same manner as described aberrant patterns, such as oral or pharyngeal retention, bolus
for the liquid bolus. fragmentation, premature swallows resulting in swallowing
Due to aberrant swallows (which can occur in healthy inhibition, gastroesophageal reflux, tracheal aspiration, and
subjects) and intra-subject variability in repetitive swal- abnormal esophageal events [16].
lows, there is a poor sensitivity (44%) and low specificity The goal of quantitative analysis is to quantify retention
(71%) for esophageal scintigraphy performed with a single and measure the rate of the oropharyngeal-esophageal transit.
swallow test. To overcome this problem, the multiple swal- Usually, five regions of interests (ROIs) are drawn, encom-
lows approach can be used, which involves six independent passing the oral cavity, pharynx, and upper, middle, and lower
liquid bolus swallows at 30-s intervals. A summed image thirds of the esophagus (Fig. 31.2). The gastric fundus should
(condensed image) of all six swallows is used to calculate be identified and excluded from the lower third region, as
both time parameters (mean time, mean transit time, and oscillations linked to respiratory movements interfere with
transit time) and retention parameters at particular time the analysis. A summed image of one or all six swallows is
points. During the multi-swallow approach, it is important used to calculate both time parameters (mean time, mean
to pay attention to the interval between swallows, because transit time, and transit time) and retention parameters at
a second swallow within 4 s inhibits the peristaltic wave, particular time points. Activity/time curves are generated by
and a second swallow between 3 and 8 s may arrests the plotting the counting-rate columns assembled in each image.
swallow in the striated muscle. Regular peristaltic waves The following quantitative indices are derived:
are elicited at swallow intervals 10–15 s (Fig. 31.3).
In adults, esophageal transit scintigraphy should be com- • Oral transit time (OTT) is the time required for the bolus
pleted by acquiring a high-temporal-resolution dynamic to exit the oral cavity after the subject has been asked to
swallow (<1 s in normal controls). The oral curve shows a
rapidly decreasing pattern with <5% residual activity.
a • Pharyngeal transit time (PTT) is the time between the
entrance of the radioactive bolus into and exit from the
pharynx (<1.2 s in normal controls). The pharyngeal
curve shows a rapid increase in radioactivity (from 0% to
a b 100%), followed by a relatively slower decline with <5%
residual activity.
• Esophageal transit time (ETT) is the time between
entrance of the bolus into and exit from the (whole)
esophagus (<10 s in normal controls, mean
time = 5.6 ± 4 s). The esophageal curve shows a rapid
increase in radioactivity, followed by a slower decrease,
with less than 20% residual activity.
• Retention index (RI) is the fraction of the radioactive
bolus remaining in each tract 10 s after swallowing. This
value is less than 5% in normal subjects for the oropha-
ryngeal area, while it is less than 20% in the esophagus.
• Esophageal emptying rate (EER) is the fraction of
radioactivity cleared from the esophagus 10 s after
Fig. 31.3 OPES in a patient with dermatomyositis and dysphagia. (a) maximum peak. This is achieved through the formula:
Condensed image of the dynamic acquisition using multiple liquid [(Emax − E10s)/Emax] × 100, where Emax is the maximum
boluses that depicted the presence of multiple swallows for each deglu-
radioactivity counts detected in the esophagus and E10s is
tition. (b) The static image acquired after swallowing liquid boluses
reveals persistent pharyngeal retention associated with right tracheo- the activity present in the esophagus 10 s after maximum
bronchial inhalation peak. This value is greater than 80% in normal controls.
If there is radioactivity in the larynx or in the tracheobron- and those with incomplete LES relaxation. Although the first
chial tract, after the administration of approximately 100 mL two conditions can even be asymptomatic, incomplete LES
of water to clear all residual traces from the pharynx, the relaxation usually interferes with esophageal emptying; this
aspiration percentage (PA) can be calculated [17] by divid- disorder is thus better termed atypical disorder of LES relax-
ing the counts obtained in the aspiration area (RA) by the ation (with impaired esophageal clearance) rather than iso-
initial activity detected in the oral cavity before swallowing lated hypertensive LES [16].
(activity time 1; AT), subtracted from the residual activity
after swallowing (activity time 2; AT2), applying the follow- 31.2.3.2 Diffuse Esophageal Spasm
ing formula: [PA = RA × 100/(AT − AT2)]. Diffuse esophageal spasm is characterized by uncoordinated
esophageal contractions that can interfere with esophageal
clearance. The key diagnostic criterion is the presence of
31.2.3 Clinical Indications simultaneous contractions induced by liquid swallows.
Besides the discrepancies in manometric criteria, diffuse
The potential clinical role of OPES in patient’s manage- esophageal spasm is an intermittent phenomenon that only in
ment is better appreciated in light of the pathophysiologic 10% may be depicted by manometry, while nearly in 100%
changes underlying esophageal motility disorders. The deci- during the multiple liquid OPES.
sion on which diagnostic study to use for esophageal dys-
motility depends on the symptoms. If dysphagia is present, 31.2.3.3 Achalasia
a barium swallow or endoscopy is usually performed first Classic achalasia is caused by degeneration of neurons in
to exclude an anatomic lesion. If these anatomic studies are the wall of the esophagus. These neurons are involved in the
not diagnostic, manometry will likely be performed to look production of nitric oxide, which relaxes esophageal smooth
for esophageal dysmotility. Manometry is considered the muscle, therefore causing the basal LES pressure to rise. The
gold standard for diagnosis of the primary esophageal motil- result is impaired esophageal clearance and delayed transit
ity disorders, which include achalasia, scleroderma, diffuse times. The OPES easily detects the delay in transit with a
esophageal spasm, hypertensive lower esophageal sphincter, sensitivity close to 100%. Distal esophageal retention does
and nonspecific esophageal motility disorders. Manometry, not clear after the patient assumes an upright position or
however, has limitations; it provides only an indirect mea- drinks a glass of water. Scintigraphic evaluation of esopha-
sure of peristalsis, as the pressure waves recorded do not geal clearance is widely accepted as the method of choice
always correlate with the aboral forces applied to a solid or in the follow-up of patients with achalasia, for monitoring
liquid bolus in the esophagus. In addition, the presence of a the efficacy of treatment (myotomy, pneumatic dilation, or
manometric tube itself may affect normal physiology, and botulinum toxin injections).
quantification of the volume of retained solids or liquids in
the esophagus is not possible [14]. 31.2.3.4 Scleroderma
The OPES is particularly useful for to evaluate dyspha- Fibrosis and vascular obliteration of the esophageal muscle
gia following surgery, due to mechanical obstruction (neo- and its innervation are the underlying causes of ineffec-
plasia, scar), after radiotherapy, as well as when esophageal tive esophageal motility in patients with scleroderma. The
manometry is not available or not tolerated by the patient, OPES can detect esophageal involvement in patients with
or the esophageal manometry results are unequivocal or asymptomatic disease, showing a typical pattern of reten-
nondiagnostic. In addition, the ability of OPES to quantitate tion of radioactivity in the lower esophagus, with clearing
total esophageal emptying is useful for assessing response to after the patient is upright or drinks a glass of water. As an
therapy [15]. indicator of dysmotility either early or in advanced disease,
the OPES has a higher sensitivity than do manometry and
31.2.3.1 Nutcracker Esophagus barium swallows.
The term nutcracker esophagus was coined by Dalton et al.
to describe the conditions under which patients with non- 31.2.3.5 Neurogenic Dysphagia
cardiac chest pain or dysphagia exhibit peristaltic waves in Neurogenic dysphagia may be caused by a disruption in
the distal esophagus. These high-amplitude waves may not different parts of the central nervous system involved in
interfere with esophageal clearance, or strictly correlate with the swallowing process—from supranuclear, motor, and
episodes of dysphagia or chest pain, or may result in mild sensory nuclei and peripheral nerve levels to abnormalities
distal esophageal retention with reflux. The clinical conse- of muscle cells and spindles (Fig. 31.4). Neuromuscular
quences of an isolated hypertensive LES remain unclear. causes of dysphagia may include stroke, brain tumors,
Three groups of patients fall under the term “isolated hyper- brain injury, bulbar and pseudobulbar paralysis, and neu-
tensive LES”: those with abnormally elevated resting LES, rodegenerative diseases such as amyotrophic lateral scle-
those with exaggerated contraction of LES after relaxation, rosis [18], multiple sclerosis, tabes dorsalis, multisystem
8 800
7
a 700
b
6 600
kcounts/sec
counts/sec
5 500
4 400
3 300
2 200
1 100
0 0
0 10 20 30 40 50 0 10 20 30 40 50
sec sec
c
10 sec
Fig. 31.4 OPES with the liquid bolus (a) and the semisolid bolus (b) residual activity is about 12% with the liquid bolus and 50% with the
in a patient with amyotrophic lateral sclerosis (ALS): delay of pharyn- semisolid bolus. In the same patient, the condensed image (c) reveals
geal transit of the liquid bolus, which becomes even more obvious with multiple swallows and marked retention of the radiopharmaceutical at
the semisolid bolus. This confirms the neurogenic component of dys- the oropharyngeal level
phagia in ALS, which affects in particular the oropharyngeal phase. The
degenerations, Parkinson’s disease, Huntington’s disease,

myasthenia and myasthenic syndromes, myopathies, and transit time (OTT), pharyngeal transit time (PTT),
peripheral neuropathies [19]. Because of the lack of inva- esophageal transit time (ETT), retention index (RI),
siveness and excellent tolerability, OPES has confirmed and esophageal emptying rate (EER).
its clinical validity in the functional, objective, and qual- • OPES may reveal the presence of tracheobronchial
itative-quantitative study of swallowing in patients suffer- inhalation; in these patients the aspiration percent-
ing from neurogenic dysphagia. Moreover, in the scenario age can be calculated by dividing the total counts
of oropharyngeal dysphagia, it has been shown that either obtained in the aspiration area by the subtracted
videoendoscopy or the OPES have a good sensitivity and activity detected in the oral cavity before and after
overall validity in comparison to the gold standard video- swallowing.
fluoroscopy and that sensitivity and overall validity values
were high (97.9% and 86.7%, respectively), demonstrating
that these two diagnostic tools (OPES and FEES) are essen-
tially superimposable in the detection of dysphagia [20]. 31.3 Gastric Emptying Scintigraphy
31.3.1 Anatomy and Physiology
Key Learning Points In term of functional gastric emptying, the stomach is divided
• OPES is increasingly gaining attention as a valid into two anatomic regions: the proximal part (fundus), which
functional imaging modality depicting swallowing controls liquid emptying, and the distal region (antrum),
and esophageal motility disorders. which controls the rate of emptying for solids. The control of
• In addition to qualitative evaluation, OPES offers a gastric emptying of food is dependent on the volume of the
series of semiquantitative information, such as oral meal and the energy density (amount of carbohydrate, fats)
of the meal; for this reason the content of the test meal is
one of the most important variables needing standardization. detection rate of delayed emptying; in fact, short-duration
Hormones such as cholecystokinin and gastrin are known to studies may be affected by pitfalls when extrapolating the
inhibit gastric emptying. T½ value using a power exponential analysis. On the other
hand, more frequent acquisitions in the first hour (e.g.,
every 15 min) permit to study the lag phase of gastric emp-
31.3.2 Clinical Indications tying [22]. The photo-peak setting is ±20% at the 140 keV
and Contraindications peak for 99mTc.
Anterior and posterior planar images should be obtained
Gastric emptying studies are employed when standard for 1 min. Immediately after ingestion of the meal. These
radiographic or endoscopic examinations cannot ade- planar images are used to calculate the geometric mean, as
quately explain clinical signs and symptoms (nausea, the square root of the product of counts in the two views.
vomiting, abdominal fullness, distention, and weight For image post-processing, a region of interest (ROI) is
loss) [14]. Currently, it is a gold standard for the assess- drawn around the activity in the entire stomach in the ante-
ing gastric emptying with physiological meal (solids, rior and posterior views (Fig. 31.5). The ROI should include
with or without liquids). Moreover, applying quantitative any visualized activity in the fundic and antral regions of
standardized assessment, it can be used to evaluate the the stomach. Care should be taken to adjust the ROI so as to
therapeutic efficacy. For patient preparation, fasting (at avoid any activity from adjacent small bowel. All data must
least for 4 h) and cessation of medications that may delay be correct for radioactivity decay.
or affect the rate of gastric emptying (prokinetics, oppi- The final measurement of gastric emptying is based
ates, antispasmodic agents, atropine, nifedipine, proges- on the percentage of gastric retention at specific times,
terone, octreotide, and benzodiazepine) at least for 2 days so as to obtain a time/activity curve calculated from
are necessary. For insulin-dependent diabetic patients, the the geometric mean of gastric counts at all time points
glucose level should be under 270 mg/dL, while tobacco (Fig. 31.6). Half-times (T½ values) of emptying may be
smoking should be prohibited during the examination. potentially less accurate than percentages of retention
Contraindications are food allergy to the radiolabeled measured at fixed time points, particularly in individuals
meal and hypoglycemia in the insulin-dependent diabetic with very prolonged gastric emptying in which extrapola-
patients, possibly induced by the long fasting period. tion is needed to calculate the half-time if it was not actu-
Sources of error can be the incomplete meal ingestion, ally reached during the test. A description of the pattern of
vomiting a portion of the ingested meal, poor labeling, emptying may also be helpful (e.g., tracer remains in the
and gastroesophageal reflux. fundus, antrum, or even in the distal esophagus through-
out the study) (Fig. 31.7).
31.3.3 Radiopharmaceuticals, Image

Acquisition, Data Processing, 31.3.4 Future Perspectives
and Reporting
Employing standardized methodology for gastric emptying
The American Neuro-gastroenterology and Motility Society scintigraphy with inter-center comparable normative val-
(ANMS) and the Society of Nuclear Medicine and Molecular ues will build uniformity and consistency in the reports of
Imaging (SNMMI) recommend two liquid eggs white, two the nuclear medicine specialists with further applications
slices of toasted white bread, 30 g of jam or jelly, and 120 mL for multicenter studies. In addition, it will help to develop
of water as the components of the meal [21]. The meal is disease-specific scales for assessing the severity of delayed
prepared by mixing 0.5–1 mCi of 99mTc-colloid into the liq- gastric emptying [23].
uid egg whites and by cooking the eggs in a microwave or Another future consolidation might be to analyze differ-
on a hot nonstick skillet. The radiolabeled test meal should ent aspects of gastric motility (such as fundal-antral coor-
be ingested as quickly as possible, optimally within 10 min dination, antropyloric coordination, gastric accommodation,
(note to record the timing to ingest the meal and whether any regional muscular contraction patterns within the stomach,
portion of the meal was not eaten). and separate fundal and antral emptying curves) and the
Images are obtained in a 128 × 128 matrix using a effect of varying meal composition on gastric emptying.
general-purpose collimator or a low-energy high-resolution Furthermore, establishing normative values for the post-
collimator, preferably with a dual-head gamma camera gastric surgery patients should be an important target for
for 1-min acquisitions at 0, 1, 2, and 3 h as the standard nuclear medicine specialists, with the aim of enhancing even
procedure. Nevertheless, it has been shown that extending further the usefulness of the gastric emptying test for the
the measurements out to 4 h (or even longer) increases the clinical management of these patients [24].
31.4 Small-Bowel and Colon Transit

Key Learning Points Scintigraphy
• In gastric emptying scintigraphy, fasting at least for
4 h and cessation of the medications that may delay Small-bowel and colon transit scintigraphy has been in use
or affect the rate of gastric emptying at least for for more than 20 years. The methods are safe and noninva-
2 days are necessary. sive, but a certain lack of standardization limits their wide-
• A carefully standardized composition of the radio- spread use. Non-radionuclide methods of measuring bowel
labeled test meal is important for inter-center com- transit include hydrogen breath tests for small-bowel transit
parisons and for assessing the efficacy of and radiopaque markers for colon transit. Bowel transit is
treatments. difficult to quantify, and, thus far, scintigraphy has yielded
• The use of short-duration studies may lead to pit- the most reliable result.
falls in estimating the gastric emptying half-value. Measurement of small-bowel transit is complex because
• Extending the measurements out to 4 h improves entry of a radiolabeled meal into the small intestine depends
the evaluation of patients with markedly delayed on gastric emptying and because small-bowel chyme spreads
gastric emptying. over a large distance while it progresses toward the colon.
Fig. 31.5 Gastric emptying study in a 1 Ant1 2 Ant2 3 Ant3

patient with dyspepsia. Right panels:
anterior and posterior planar images of
the gastric contents up to 3 h after
ingestion of the radiolabeled meal. Right
panel: activity/time curve of the gastric
contents shows a gastric emptying pattern
within normal limits (T½ < 120 min)
4 Ant4 5 Ant5 6 Ant6
Anterior
Post1 Post2 Post3
1 2 3
6
Post4 Post5
Post6
4 5 6
Post7 Post8 Post9

7 8 9
Posterior
Raw Gastric Emptying Linear Fit 50% Emptying
30
28
26
24
22
20
18
kcounts/min
16
14
12
10
4 Linear_Fit T1/2
2
0
0 20 40 60 80 100 120 140 160 180
Minutes
Decay Corrected by Xeleris: Yes
Linear Fit T 1/2 (min) = 100.18

Linear Fit Slope (%/min) = 0.50 Tc_99m
Raw Data T 1/2 (min) = 98.22 ** Geometric Mean
Frame/Time Fit / Raw % Empty KCounts/min** Anterior / Post
1 0.0 0 0 28.87 63.51 13.13
2 17.0 8 8 26.57 63.28 11.16
3 33.0 16 17 23.91 58.54 9.77
4 43.0 21 21 22.74 53.82 9.61
5 56.0 28 22 22.40 52.43 9.57
6 91.0 45 45 15.86 38.46 6.54
7 116.0 58 62 10.92 24.00 4.97
8 143.0 71 75 7.32 15.32 3.50
9 170.0 85 80 5.68 11.74 2.75
Furthermore, transit through the small bowel can be impeded 31.4.1 Anatomy and Physiology
by slow colonic transit. The breath hydrogen test may reflect
bacterial contamination of the small bowel, and cautious Small and large bowel motility is the result of complex gas-
interpretation is required in conditions that may be associ- trointestinal contractions that promote the aboral movement
ated with bacterial overgrowth. of intestinal chyme and indigestible solids. The function
Colon transit scintigraphy has been shown to correlate of the small bowel is to transport food as is empties from
with radiopaque markers, and it provides information on the stomach and to mix it with bile and with pancreatic and
regional and overall colon transit. It is recommended that intestinal secretions to facilitate absorption over the bowel
gastrointestinal transit studies are used to localize the poten- mucosal surface. There is no simple small-bowel peristaltic
tial site of disease and guide therapy. The ANMS task force pattern. Anterograde and retrograde movements of intestinal
on gastroenterology transit stated that “the scintigraphic chyme occur in the jejunum and ileum, with some areas pro-
method is the only one that reliably allows determination of gressing rapidly and others slowly. Jejunal peristaltic activity
both total and regional transit times” for gastrointestinal and is typically more rapid and intense, with slowing of peristal-
colon transit [25]. sis seen in the ileum. This process, although irregular, results
Fig. 31.6 Gastric emptying

study in a patient with diabetic 1 2 Ant2 3 Ant3
gastroparesis. Upper panel:
anterior and posterior planar
images of the gastric contents up
to 3 h after ingestion of the
radiolabeled meal. Lower panel:
the activity/time curve of the
gastric contents shows delayed Ant4 Ant5 Ant6
gastric emptying (T½ > 120 min) 4 5 6
Anterior
Post2 Post3
1 2 3
Post4 Post5
Post6
4 5 6
Post7 Post8 Post9

7 8 9
Posterior
Raw Gastric Emptying Linear Fit 50% Emptying
40
30
kcounts/min
20
10
0
0 20 40 60 80 100 120 140 160 180
Minutes
Decay Corrected by Xeleris: Yes
Linear Fit T 1/2 (min) = 264.45

Linear Fit Slope (%/min) = 0.19 Tc_99m
Raw Data T 1/2 (min) = None ** Geometric Mean
Frame/Time Fit / Raw % Empty KCounts/min** Anterior / Post
1 0.0 0 0 34.88 80.87 15.05
2 12.0 2 -1 35.35 73.61 16.98
3 29.0 5 4 33.60 67.26 16.78
4 44.0 8 14 30.14 61.85 14.68
5 57.0 11 22 27.19 61.16 12.08
6 86.0 16 23 26.97 61.31 11.87
7 116.0 22 22 27.09 60.59 12.11
8 141.0 27 27 25.54 61.54 10.60
9 173.0 33 31 24.12 58.28 9.99
in a net overall progression of chyme toward the colon. The pseudo-obstruction, scleroderma, celiac disease and mal-
ileocolonic junction controls how the liquid small-bowel absorption, irritable bowel syndrome with diarrhea (IBS-
contents enter the colon. The motility of the colon controls D), dolichocolon (Fig. 31.8), and/or anismus (Fig. 31.9).
slow mixing and the movement of its contents so the colon Approximately 35% of patients with IBS have abnormal
can absorb water and liquids can be transformed to semisol- overall colonic transit, including 48% of those with IBS-
ids or solids in the sigmoid colon. D. Symptoms attributable to disorders of the small bowel
may mimic those of abnormalities in gastric emptying, thus
delaying diagnostic work-up of small-bowel transit.
31.4.2 Clinical Indications Patients are required to fast overnight or at least 8 h before
and Contraindications the study and to discontinue medications that affect motility
(opiate analgesics, anticholinergics, and prokinetic agents)
Intestinal transit scintigraphy is indicated for patients with at least 48–72 h before the start of the procedure. Insulin-
either of the following conditions: dyspepsia, irritable bowel dependent diabetic patients should be instructed to monitor
syndrome, chronic diarrhea, chronic idiopathic intestinal and regulate the insulin levels during the test meal; if the
1 2 3
4 5 6
7 8 9
Anterior
1 2 3
4 5 6
7 8 9
Posterior
Fig. 31.7 Sequential images of a gastric emptying study: anterior and ease. After the ingestion of the radiolabeled meal, there is evidence of
posterior planar images of the gastric contents up to 3 h after ingestion activity at the distal esophagus
of the radiolabeled meal in a patient with gastroesophageal reflux dis-
Fig. 31.8 Colon transit scintigraphy in a patient with dolichocolon and slow colonic transit (Images provided by courtesy of Venanzio Valenza,
MD, Nuclear Medicine Department, Catholic University and “Agostino Gemelli” University Hospital, Rome, Italy)
blood glucose levels are >250 mg/dL, the test should not be of a gastric emptying study. The most common method uses a
performed, because hyperglycemia can delay gastric empty- mixed solid-liquid gastric emptying meal, with the liquid phase
ing and affect small-bowel transit measurements. For colon being radiolabeled with 111In-diethylenetriaminepentaacetic
transit studies, medications that can affect gastrointestinal acid (111In-DTPA) [26]. As the liquid mixes with intestinal
transit are typically withheld before and during the entire chyme, transit of the radiopharmaceutical is viewed to assess
test (laxatives). No other diagnostic studies, e.g., with bowel small-bowel and colon transit. If a solid-phase gastric emp-
radiologic contrast materials, should be scheduled 4 days tying study is not requested, a solid meal is administered,
prior to the scintigraphic examination. but without radiolabeling to maintain the standardized meal
content.
31.4.3 Radiopharmaceuticals, Image

Acquisition, Data Processing, 31.4.4 Small-Bowel Transit Scintigraphy
and Reporting
The simplest conceptual approach to scintigraphic measure-
Small-bowel and colon transit scintigraphy are typically per- ment of small-bowel transit is to measure the oro-cecal tran-
formed alone or, with minor modifications, as a continuation sit time by imaging the leading edge of radiotracer transit
Fig. 31.9 Colon transit scintigraphy in a patient with anismus, a medi- colon, with full and rapid concentration in the rectum at 48 h, followed
cal condition also known pelvic floor dis-synergism or paradoxical by a prolonged stasis up to 96 h and subsequent complete evacuation at
puborectalis contraction. In this condition, the external anal sphincter 120 h. The time-activity curves (T/Ac) corresponding to the “entire
and the puborectalis muscle contract rather than relaxing during an colon” (bottom left) and to the rectum only (bottom right) show pro-
attempted bowel motion. It is one of the causes of obstructed defaeca- longed retention of the administered bolus up to 96 h, with complete
tion syndrome (ODS). Scans acquired starting 6 and up to 120 h post- evacuation at 120 h (Images provided by courtesy of Venanzio Valenza,
administration of 100 μCi of 67Ga-citrate. The anterior/posterior images MD, Nuclear Medicine Department, Catholic University and “Agostino
(right to left; top to bottom) show regular bolus transit throughout the Gemelli” University Hospital, Rome, Italy)
through the bowel. However, accurately defining the lead- The recent SNMMI/EANM guidelines recommend using
ing edge (the first visualized arrival of activity in the cecum) the percentage of administered liquid meal that is accumu-
requires frequent (every 10–15 min) and prolonged imaging lated in the terminal ileum at 6 h after meal ingestion, as a
because of the stasis in the terminal ileum. Hydrogen breath simple index of small-bowel transit [27]. Small-bowel tran-
testing, which measures leading-edge transit, correlates well sit is considered normal if more than 40% of administered
with scintigraphy. The small-bowel oro-cecal transit time activity has progressed into the terminal ileum or passed into
will vary with the meal administered: 151–290 min when the cecum and ascending colon at 6 h. In the case of delayed
using resin pellets mixed with a meal, 72–392 min when transit, the activity persists in multiple loops of small bowel
measured using the liquid phase of a mixed solid-liquid meal at 6 h and little (<40%) to no activity arrives in the terminal
in healthy individuals. An alternative is to measure simply ileum reservoir. Another index of small-bowel transit is the
the overall bulk movement as it progresses distally into the colon filling at 6 h, with normal range for nondigestible par-
terminal ileum and use the progressive buildup of activity in ticles between 11% and 70% and between 43% and 95% for
the ileocolonic junction as an index of small-bowel motility. digestible solids.
Fig. 31.10 Colon transit scintigraphy in a healthy volunteer: anterior/ ity in the rectum after 48 and 72 h. The activity-time curves at the bot-
posterior images (from right to left and from top to bottom). Normal tom for the entire colon transit confirm qualitative analysis: complete
transit of the liquid bolus (100 μCi of 67Ga-citrate in 100 mL of water) bolus retention at 6 and 24 h, with markedly decrease of activity at 48
in the right colon can be observed (visible after 6 h), followed by trans- and 72 h, after spontaneous evacuation (Images provided by courtesy of
verse, descending, and sigmoid colon (well filled at 24 h and almost Venanzio Valenza, MD, Nuclear Medicine Department, Catholic
completely emptied after 48 h). Note the normal minimal residual activ- University and “Agostino Gemelli” University Hospital, Rome, Italy)
31.4.5 Colon Transit Scintigraphy colon. To quantitate colon transit, the concept of geometric
center (GC) has been defined, which is a weighted average of
When given orally, 111In-DTPA (2.8 days half-life) and the counts for each anatomical subdivision of the colon, the
67
Ga-citrate (3.2-days half-life) are both nonabsorbable and counts being expressed as a fraction of administered activity,
have physical decay half-lives long enough to permit several as follows: ROI-counts/(administered counts).
days of imaging (Fig. 31.10). The Temple University and Some centers use the geometric center that is based on six
Johns Hopkins Hospital protocols combine gastric emptying colonic regions according to the Mayo method (four main
and small-bowel scintigraphy by using the solid and liquid regions plus the hepatic and splenic flexures). The Temple
phase of the meal, respectively, with the 99mTc-colloid-labeled University method adds a seventh, “virtual region” which is
egg sandwich and with 111In-DTPA-labeled water. The latter represented by activity in the expelled stool. The fraction of
is used to assess small-bowel and colonic transit. Anterior and counts in each anatomic region is corrected using a weight-
posterior images are obtained at predefined times (e.g., 4, 24, ing factor that increases from 1 to 5 (Mayo Clinic protocol) or
and 48 h), while the patient stands erect; an ROI is then drawn from 1 to 7 (Temple University protocol) progressing proximal
around the images corresponding to different regions of the to distal from cecum to rectosigmoid and to evacuated stool.
4 4
2 3 2 3
Key Learning Points
• The simplest index for small-bowel transit scintig-
raphy is the percentage of administered liquid meal
that is accumulated in the terminal ileum at 6 h after
1
5
1
5 meal ingestion.
• Small-bowel transit is considered normal if more
than 40% of administered activity has progressed
into the terminal ileum or passed into the cecum
6 6 and ascending colon at 6 h.
• Colon transit scintigraphy with qualitative and
semiquantitative analysis depicts the regional pat-
7 7 tern of abnormal colon transit, both for slow (24-h
colon GC <1.7) or accelerated transit (24-h colon
GC >4).
0.40 x 1 = 0.40+ 0.30 x 3 = 0.90+
0.10 x 2 = 0.20+ 0.10 x 4 = 0.40+
0.25 x 3 = 0.75+ 0.34 x 5 = 1.70+
0.10 x 4 = 0.40+ 0.25 x 6 = 1.50+
0.15 x 5 = 0.75+ 0.16 x 7 = 1.12+ 31.5 Gastrointestinal Bleeding
Scintigraphy
GC = 2.50 GC = 5.62
Gastrointestinal bleeding scintigraphy (GIBS) is a nonin-

vasive diagnostic radionuclide imaging study that is per-
Fig. 31.11 Exemplification of the method for calculating the geomet- formed in patients with suspected gastrointestinal bleeding
ric center (GC) for colonic transit studies. In this diagram the colon is
subdivided into six regions with weighting factors increasing from 1 to (not detected through radiological investigations such as the
6 progressing from the cecum/ascending colon to the rectosigmoid esophagogastroduodenoscopy or colonoscopy). Purposes of
tract; a weighting factor of 7 is assigned to radioactivity evacuated in the investigation are (1) to determine whether the bleeding
stools (estimated as the activity unaccounted for in the images). Each is active, (2) to localize the bleeding site, and (3) to estimate
black dot represents 1% (or 0.01 fraction) of radioactivity that has
reached the colon. Examples of calculation are given in inserts for two the approximate bleeding volume for prognostic purposes.
different CG values, 2.50 on left and 5.62 on right. In healthy individu- Often GI bleeding is intermittent, resulting in long inter-
als, most of radioactivity would be seen in transverse colon at 24 h (GC vals with no extravasation of blood into the GI tract, alternat-
approximately 3), progressing to descending colon at 48 h (GC about ing with brief intervals of bleeding. As a result, a technique
5), and being mostly evacuated at 72 h (GC about 6–7). In patients with
colonic inertia, radioactivity would not significantly progress past the such as GIBS, which permits repeated evaluation for bleed-
hepatic flexure at 48 h or at 72 h. (CG values about 1.5–2 and 2–3, ing, is preferred to a study where bleeding must occur dur-
respectively). In patients with functional obstruction of the rectosig- ing the time of imaging (e.g., angiography). Gastrointestinal
moid tract, radioactivity would progress nearly normally until 48 h after bleeding scintigraphy enables continuous monitoring of the
ingestion; however, there would be little further progression at 72 h (GC
about 5–6), thus indicating obstructed defecation (reproduced from Ref. entire gastrointestinal tract for up to approximately 24 h
[33]: Mariani G, et al. J Nucl Med. 2008;49:776–787) [29], and it can detect bleeding at low flow rates (0.04 mL/
min). This investigation does not require any special patient
preparation, can be performed with standard nuclear medi-
A low GC value indicates that most of administered activity is cine instrumentation, and is well tolerated even in patients
in still the cecum, whereas a high GC value indicates that most who are acutely ill [30].
of the activity is in the rectosigmoid region or has been evacu- Gastrointestinal bleeding may be classified as upper gas-
ated. Figure 31.11 illustrates how the CG value is calculated trointestinal bleeding (above the ampulla of Vater and within
after acquiring sequential images during a colonic transit study. reach of esophagogastroduodenoscopy), mid-gastrointestinal
The average overall colonic transit, expressed as the GC bleeding (small bowel from the ampulla of Vater to the ter-
at 24 h, is 2.7 ± 1.05 in subjects with normal gastrointestinal minal ileum, which can be evaluated by capsule endoscopy
transit. Patients with a 24-h colon GC <1.7 are considered to or double-balloon enteroscopy), or lower gastrointestinal
have slow transit, and those with a GC of 4.0 and higher are bleeding in the colon, which can be evaluated by colonos-
considered to have an accelerated transit [28]. Scintigraphy copy [31]. Common causes of upper gastrointestinal bleed-
can identify regional patterns of delays in colon transit that ing include esophageal varices, gastric and duodenal ulcers,
could suggest slow transit constipation, colon inertia, or accel- gastritis, esophagitis, Mallory-Weiss tears, and neoplasms.
erated transit (e.g., in carcinoid diarrhea or irritable colon). The most common causes of mid-gastrointestinal bleeding
are angiodysplasia, neoplasms, Crohn’s disease, diverticula, chronic occult gastrointestinal bleeding because the guaiac
and heterotopic gastric mucosa in Meckel’s diverticulum. fecal occult blood test may detect bleeding at rates well
Common causes of lower gastrointestinal bleeding include below those necessary to be identified with GIBS [33].
angiodysplasia, diverticulosis, benign and malignant bowel GIBS is indicated primarily for overt mid or lower gastro-
neoplasms, adenomatous polyps, inflammatory bowel dis- intestinal bleeding, specifically when an upper gastrointestinal
ease, and infectious bowel disease. bleed has been excluded by nasogastric lavage [34]. In this
scenario, GIBS can be used as an early diagnostic study for
gastrointestinal bleeding especially for hospitalized patients or
31.5.1 Radiopharmaceuticals and Technique patients in the emergency department. GIBS can be beneficial
when other studies require lengthy patient preparation or are
Historically, two radiopharmaceuticals have been used for contraindicated. Although GIBS can also identify overt upper
GIBS: 99mTc-labeled red blood cells (RBCs) and 99mTc-sul- gastrointestinal bleeding, usually the first procedure performed
fur colloid. Radiolabeled red blood cells are the radiophar- to confirm upper gastrointestinal bleeding is nasogastric lavage,
maceutical of choice for performing GIBS because of an followed by esophagogastroduodenoscopy to identify and treat
intravascular half-life that allows continuous imaging of the suspected overt upper gastrointestinal bleeding.
gastrointestinal tract over many hours. GIBS is also indicated to help identify the source of
Three methods are available to label RBCs: in vitro, mod- obscure overt gastrointestinal bleeding [35].
ified in vivo, and in vivo. The recommended radiopharma- Among some of the other common clinical indications
ceutical is in vitro labeled autologous RBCs; administered for GIBS are stratifying risk in patients with gastrointesti-
activity of 99mTc-RBCs is 555–1110 MBq (15–30 mCi) in nal bleeding, directing timely diagnostic angiography, and
adults [29]. In children under 18 years old, the recommended assisting in plans for surgical or other interventional proce-
administered activity is based on the European Society of dures [30].
Nuclear Medicine (EANM) Pediatric Dosage Card, which
uses a baseline activity of 56 MBq (1.51 mCi) multiplied by
weight of the patient [32]. Key Learning Points
GIBS commonly uses serial images after an intravenous • Gastrointestinal bleeding scintigraphy enables to
bolus injection of the radiolabeled, washed RBCs. The exact detect bleeding throughout the GI tract in a nonin-
timing and modalities of dynamic imaging vary from center vasive manner, even if bleeding occurs at flow rates
to center, but recording the initial phase (1–2 min after injec- as low as 0.04 mL/min.
tion) as an abdominal blood-flow study (1–5 s/frame) may • By acquiring imaging over an extended period (up
help with recognizing a vascular mass in the abdomen and to 24 h), GIBS enables to detect also intermitted
with localizing the site of bleeding more precisely. After the bleeding in the GI tract.
initial radionuclide angiogram, serial images are commonly • The most important sign to confirm active gastroin-
recorded at 1 min per frame for 60–90 min. If no bleeding testinal bleeding is the presence of an area of extra-
site is identified within the first 90 min, delayed images are vascular activity that increases in intensity over
usually acquired, typically at 2–6 h or 18–24 h after injec- time or the presence of a focus of activity that
tion. Acquisition of SPECT images can be helpful for local- moves in a pattern corresponding to bowel
izing bleeding sites, especially if the imaging is performed movement.
with hybrid SPECT/CT (Fig. 31.12). • SPECT/CT imaging helps to accurately localize the
Interpretation of the images is qualitative. The key sign of site of bleeding.
gastrointestinal bleeding is the presence of an area of extra-
vascular activity that increases in intensity over time or the
presence of a focus of activity that moves in a pattern corre-
sponding to bowel. Movement of activity in the lumen might 31.6 Meckel’s Diverticulum Scintigraphy
be anterograde or retrograde because of the irritating charac-
teristics of intraluminal blood. Meckel’s diverticulum scintigraphy is a test for the identifica-
tion of heterotopic gastric mucosa in Meckel’s diverticulum,
with high specificity and positive predictive value (close to
31.5.2 Clinical Indications 100%) in both children and adults. The abnormal presence
of heterotopic gastric mucosa in Meckel’s diverticulum is the
GIBS is commonly indicated for identifying an active gas- most common cause of lower gastrointestinal hemorrhage in
trointestinal bleeding site in patients with overt gastrointesti- previously healthy infants. Over 50% of these patients pres-
nal bleeding. GIBS should not be performed in patients with ent with bleeding by the age of 2 years.
Fig. 31.12 Gastrointestinal bleeding scintigraphy in a patient with pharmaceutical at the level of the left flank corresponding to low-flow
hypocromic microcytic anemia. The images acquired at different times bleeding at the small bowel (a). SPECT/CT imaging allows to better
reveal depict the presence of a progressive accumulation of the radio- locate the site of bleeding (b)
31.6.1 Normal Anatomy

Key Learning Points
Meckel’s diverticulum is the vestigial remnant of the ompha- • Meckel’s diverticulum scintigraphy is based on the
lomesenteric duct and represents the most common congeni- property of gastric mucosa to actively accumulate
tal anomaly of the gastrointestinal tract, with an incidence of 99m
Tc-pertechnetate.
1–3% in the general population. It is normally located on the • Gastric mucosal islands with diameter of over 1 cm
antimesenteric border of the terminal ileum within 80–100 cm can be detected with almost 90% accuracy by
of the ileocecal valve and is on average 2 cm in length. Meckel’s diverticulum scintigraphy.
Approximately 57% of Meckel’s diverticula contain • Meckel’s diverticulum scintigraphy should not be
ectopic gastric mucosa, which actively secretes the hydro- used when the patient is actively bleeding.
chloric acid responsible for mucosal ulcerations within the • Active bleeding is best explored with 99mTc-RBCs
diverticulum and the unprotected wall of the adjacent ileum scintigraphy.
[34]. The most common sign of this abnormality of Meckel’s
diverticulum is gross rectal bleeding, which may or may not
be associated with abdominal symptoms. Almost all diver-
ticula of children with symptoms of lower gastrointestinal 31.7 Hepatobiliary Scintigraphy
bleeding contain ectopic gastric mucosa [36].
Hepatobiliary scintigraphy is a radionuclide diagnostic
imaging procedure that evaluates hepatocellular function
31.6.2 Radiopharmaceuticals and Technique and proper patency/functioning of the excretory biliary
route by tracing the production and flow of bile from its
Technetium-99m pertechnetate is known to accumulate in production in the liver to its passage through the biliary
the gastric mucosa, and it is used to visualize symptomatic system into the small intestine. Since the 1980s, hepato-
Meckel’s diverticula, which often contain heterotopic gastric biliary scintigraphy (also called cholescintigraphy) consti-
epithelium. Such mucosal islands with diameter of over 1 cm tutes a valuable clinical diagnostic test largely available
can be detected with almost 90% accuracy using high-resolution in nuclear medicine centers. When optimally performed,
gamma cameras [35]. The scintigraphic study is based on i.v. hepatobiliary scintigraphy is a sensitive method for detect-
administration of 370 MBq (10 mCi) of 99mTc-pertechnetate in ing various disorders of the liver and biliary system.
the adult, while the minimum pediatric activity is 20 MBq. The Additional pharmacologic testing may enhance the diag-
patient is positioned supine, and the imaging field is the abdo- nostic utility of hepatobiliary scintigraphy and provide the
men and pelvis (to include stomach and bladder). quantitative assessment that is necessary for certain spe-
Anterior abdominal dynamic images are obtained with cific applications.
a frame rate of 1 image every 30–60 s for at least 30 min.
Additional static images (anterior, anterior oblique, lateral,
and posterior projection views) are recommended at the end 31.7.1 Anatomy and Physiology
of the dynamic acquisition. SPECT imaging (and especially
hybrid SPECT/CT) may improve the detection of a small The liver is the second largest organ of the human body,
diverticulum, or a diverticulum obscured by the urinary blad- weighing 1200–1500 g or approximately 4–5% of body
der, when the clinical suspicion for a Meckel’s diverticulum weight. It is located in the right upper abdominal quadrant,
is high and the planar images have negative or equivocal or the right hypochondriac and epigastric regions, behind the
findings (Fig. 31.13) [37]. lower ribs. The falciform ligament divides the liver anatomi-
cally into two unequal lobes: right and left. Two additional
smaller lobes, the quadrate and caudate lobes, are more vis-
31.6.3 Clinical Indications ible in cross section. Physiologically though, the division is
equal, following the fossa for gallbladder and inferior vena
The indication for Meckel’s diverticulum scintigraphy is cava. There is no evidence for difference in functions among
to localize ectopic gastric mucosa in a Meckel’s diverticu- the four anatomical lobes. The gallbladder is a saccular, pear-
lum as the source of unexplained gastrointestinal bleed- shaped organ located posterior to the most anterior aspect of
ing. Bleeding Meckel’s diverticula usually occur in young the liver that functions to store bile. It has a mean capacity of
children. Meckel’s diverticulum scintigraphy should be 30–50 mL. Mucosal folds, called the spiral valves of Heister,
employed when the patient is not actively bleeding. Even maintain patency of the cystic duct to allow passage of bile.
in young children, active bleeding is best studied by 99mTc- Presence of fats in the duodenum stimulates the gallbladder
RBCs scintigraphy. to contract.
Fig. 31.13 Meckel’s diverticulum scintigraphy in a child with melena. The SPECT/CT acquisition demonstrates abnormal accumulation of
(a) Anterior abdominal dynamic images (1 frame every 60 s) demon- radioactivity corresponding to a diverticular formation that originates
strate, at the beginning of acquisition, a focal area of radioactivity accu- from the upper wall of the small loop. The findings are compatible with
mulation at the level of the right flank, in the paraumbilical region. This the presence of ectopic gastric mucus in Meckel’s diverticulum
finding persists and increases in intensity in the following minutes. (b)
Bile is synthesized and secreted by hepatocytes into the relevance, even if changing the 131I- with the 123I-label.
biliary canaliculi, and then it flows into progressively larger 99m
Tc-labeled hepatobiliary radiopharmaceuticals were
ducts forming the right and left hepatic ducts. These ducts discovered by chance. During investigations for a cardiac
ultimately drain into the common hepatic duct. The common imaging radiopharmaceutical, it was found that 99mTc-
hepatic duct then joins with the cystic duct from the gallblad- labeled lidocaine was not a good cardiac imaging agent
der to form the common bile duct, through which the bile but cleared rapidly through the hepatobiliary system. With
reaches the duodenum via the greater duodenal papilla (of some chemical modification, 99mTc-labeled hepatobiliary
Vater) (see Fig. 31.14). However, not all bile runs directly imaging agents were developed and began to be widely
into the duodenum. About 50% of the bile produced by the used by the early 1980s.
liver is first stored in the gallbladder. Then, when food is The first 99mTc-labeled hepatobiliary radiopharmaceutical,
ingested, the gallbladder contracts and releases stored bile iminodiacetic acid (HIDA), was approved for clinical use by
into the duodenum to help break down and absorb the fats. the FDA in 1982. 99mTc-dimethyl (2,6-diisopropylacetanilido
The biliary system’s main functions are (1) to drain waste iminodiacetic acid) constituted a major advance, but image
products from the liver into the intestine and (2) to help quality and diagnostic utility were suboptimal in patients
in digestion with the controlled release of bile. Bile is the with serum bilirubin levels above 5.0 mg/dL. This agent is
greenish-yellow fluid (consisting of waste products, choles- no longer available commercially.
terol, and bile salts). Bile salt is the actual component that In 1986, the FDA approved 99mTc-disofenin (diisopro-
helps break down and absorb fats. pyl iminodiacetic acid), that provides images of diagnos-
tic quality even with bilirubin levels as high as 25–30 mg/
dL. In 1993, 99mTc-mebrofenin (bromo-2,4,6-trimethylacet-
31.7.2 Radiopharmaceuticals and Technique anilido iminodiacetic acid) was approved, based on higher
liver extraction (98% versus 89%) and more rapid biliary
The first clinically useful cholescintigraphic agent was clearance than 99mTc-disofenin (17 min half-life versus
the 131I-rose bengal; however, image quality was rela- 19 min) [38].
tively poor and currently, this agent has only historical 99m
Tc-disofenin or 99mTc-mebrofenin is administered
intravenously in activities of 111–185 MBq (3–5 mCi) for
adults; a greater administered activity may be needed in
patients with hyperbilirubinemia. 99mTc-mebrofenin may be
selected instead of 99mTc-disofenin in patients with moderate
to severe hepatic dysfunction, because of its higher hepato-
cellular extraction.
The administered activity for infants and children is
1.8 MBq/kg (0.05 mCi/kg), with a minimum administered
activity of 18.5 MBq (0.5 mCi). 99mTc-mebrofenin is always
common
hepatic duct
preferred in neonates with hyperbilirubinemia, with a mini-
mum administered activity of 37 MBq (1.0 mCi), as up to
choledochus 24-h delayed images are often necessary [39].
After intravenous injection, 99mTc-labeled HIDA radio-
cystic pharmaceuticals are transported in the blood bound to
duct m serum albumin. They dissociate from albumin in the hepatic
nu
ode perisinusoidal space and are extracted by hepatocytes by
du
receptor-mediated endocytosis, similar to bile salts, free
s fatty acids, and bilirubin. They then follow the same meta-
ea
ncr bolic pathway as bilirubin, except that they are secreted into
pa
biliary canaliculi unchanged, without undergoing conjuga-
sphyncter
of Oddi tion [38]. Approximately two-thirds of the tracer enters the
gallbladder via the cystic duct, whereas the remainder exits
into the second part of duodenum via the common bile duct
and sphincter of Oddi. The distribution of tracer within the
Fig. 31.14 Schematic representation of the main intra- and extrahe- biliary system depends on the patency of the biliary ducts,
patic biliary tree, gallbladder, and surrounding structures (modified
from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di Medicina
intraluminal pressures, and sphincter of Oddi tone. Hence,
Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010) an unusual pattern of distribution alludes to one of the factors
mentioned earlier.
31.7.2.1 Patient Preparation liver and be cleared from the cardiac blood pool within about
All available clinical, laboratory, and other imaging informa- 5 min. Persistence of activity within the blood pool beyond
tion (e.g., ultrasound, MRI, as adequate) should be reviewed 20 min indicates reduced hepatocellular function, which can
before the scan. Additional information specifically related be due to causes of liver failure or intercurrent illnesses.
to hepatobiliary scintigraphy should include: Radioactivity is normally seen within the biliary ducts
between 10 and 30 min and in the bowel by 60 min postin-
• History of prior surgeries, especially biliary and jection of the 99mTc-HIDA radiopharmaceutical (Fig. 31.16).
gastrointestinal Patients pretreated with sincalide may have a longer delay
• Time of most recent meal before activity of radiotracer is seen within the bowel. This is
• Current medications with particular attention to opioids probably due to preferential accumulation of radiotracer into
• Recent bilirubin and liver enzyme levels an empty gallbladder. However, if the parenchymal clearance
• Gallbladder or abdominal ultrasound and the radiotracer transit times into the bowel are severely
prolonged, then biliary obstruction should be considered.
In order to allow for fast gallbladder visualization, the This can occur when there is a stone in the common bile duct
adult patient must have fasted for a minimum of 2 h and in case of acute calculous cholecystitis, causing obstruction.
preferably 6 h before administration of the radiopharmaceu-
tical. Children should be instructed to fast for 2–4 h, whereas
infants need to fast only 2 h before radiotracer injection. In 31.7.3 Clinical Indications
the latter group, clear liquids (no milk) are permissible, if
medically necessary. Hepatobiliary scintigraphy is appropriately used in many
However, fasting for longer than 24 h (including those on clinical scenarios; the most common clinical indications
total parenteral nutrition) can cause the gallbladder not to fill include:
with radioactivity within the normally expected time frame.
In this case the patient may be pretreated with sincalide. • Functional biliary pain syndromes in adults and in pediat-
Disregard of the above precautions may result in a false- ric patients
positive non-visualization of the gallbladder. Interference by • Acute cholecystitis
opioids can be minimized by delaying the study for a time cor- • Right upper quadrant pain variants
responding to four half-lives of the medication. In some cases • Biliary system patency
the effect can be reversed with naloxone hydrochloride [39]. • Bile leakage
• Neonatal hyperbilirubinemia (biliary atresia versus neo-
31.7.2.2 Image Acquisition natal hepatitis syndrome)
A large-field-of-view g-camera equipped with a low-energy • Assessment of biliary-enteric bypass (e.g., Kasai
all-purpose or high-resolution collimator is recommended. procedure)
Whenever possible, continuous (dynamic) computer acqui- • Assessment of liver transplant
sition (usually in the anterior or left anterior oblique view) • Afferent loop syndrome
should be performed (1 frame/min). A 128 × 128 matrix • Assessment of choledochal cysts
(with zoom factor 1) is optimal for a standard large-field-of- • Calculation of the gallbladder ejection fraction (GBEF)
view camera. In pediatric patients an appropriate electronic • Functional assessment of the liver before partial
acquisition zoom should be used. Initial images are usually hepatectomy
acquired dynamically, starting at injection and continuing • Demonstration of anomalous liver lobulation
for 60 min. When visualization of the gallbladder is the end- • Assessment of enterogastric (duodeno-gastric) reflux
point of the study, acquisition can be stopped earlier, as soon • Esophageal bile reflux after gastrectomy
as activity is seen in the gallbladder. Additional views (e.g., • Sphincter of Oddi dysfunction
right lateral, left or right anterior oblique) may be obtained
as needed, to clarify anatomy. Visualization of tracer activity 31.7.3.1 Acute Cholecystitis
in the small bowel may resolve concern about common bile Acute cholecystitis is an acute inflammation of the gall-
duct obstruction (highly unlikely in the presence of gallblad- bladder, most frequently (>90% of the cases) constituting
der visualization) [39]. a complication of cholelithiasis. The difference between
Typical activity/time curves corresponding to ROIOs acute calculous cholecystitis and acalculous cholecysti-
defined over the main structures visualized during dynamic tis is the absence of cholelithiasis in the latter. Bile stasis
recording of hepatobiliary scintigraphy in a subject without in the gallbladder is central to the pathophysiology of acute
hepatocellular or biliary dysfunction are plotted in Fig. 31.15. calculous cholecystitis, leading to chemical irritation from
In normal individuals, the radiotracer should concentrate in the toxic lysolecithin, which is produced when mucosal and
DYNAMIC_E/2009.05.14/09:05:42.0/ fegato/fegato_c1
130
110
counts/sec
90
70
50
30
10
0 10 20 30 40 50 60
min
DYNAMIC_E/2009.05.14/09:05:42.0/colecisti/colecisti DYNAMIC_E/2009.05.14/09:05:42.0/intestino/intestino
230 400
190 300
counts/sec
counts/sec
150 200
110 100
70 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
min min
DYNAMIC_E/2009.05.14/09:05:42.0/vbi/vbi_c1 DYNAMIC_E/2009.05.14/09:05:42.0/coledoco/coledoco
40 40
30 30
counts/sec
counts/sec
20 20
10 10
0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
min min
Fig. 31.15 Dynamic hepatobiliary scintigraphy in an individual with panel), duodenal lumen (yellow curve in middle right panel), main
normal hepatocellular function and biliary drainage pattern: activity/ hepatic ducts (green curve in lower left panel), and choledochus (light
time curves obtained over ROIs manually drawn over the compressed blue in lower right panel) (reproduced with permission from: Volterrani
image (upper left panel) corresponding to the liver parenchyma (red D, Erba PA, Mariani G, eds. Fondamenti di Medicina Nucleare –
curve in upper right panel), gallbladder (deep blue curve in middle left Tecniche e Applicazioni. Milan: Springer; 2010)
luminal phosphoipases hydrolyze lecithin to form lysoleci- A few patients will have an elevated level of serum bili-
thin. Lysolecithin disrupts the protective glycoprotein on the rubin caused by stones lodged in both the cystic and the
gallbladder’s mucousal layer, thus exposing the epithelium biliary ducts or, less commonly, by an impacted cystic duct
to the detergent action of the bile salts. This insult precipi- stone compressing the adjacent hepatic or common bile duct
tates release of inflammatory mediators and the secretion of (Mirizzi syndrome).
an inflammatory transudate by the gallbladder wall that, in A sequence of pathophysiologic events occurs with acute
turn, elevates the intraluminal gallbladder pressure, leading cholecystitis. Initially the cystic duct is obstructed, usually
to its distention that results in a colic pain [40]. Except for by a stone. The gallbladder mucosa becomes edematous, fol-
low-grade fever and moderately elevated leukocyte count, lowed by neutrophilic infiltration, then mucosal ulceration,
laboratory findings (including liver function studies) may be hemorrhage, and necrosis, and ultimately, if left untreated,
normal. the complications of perforation, gangrene, and abscess.
Fig. 31.16 Images extracted at

5-min intervals (left to right and
top to bottom) from a dynamic
acquisition up to 60 min during
hepatobiliary scintigraphy in an
individual with normal
hepatocellular function and
biliary drainage pattern. There is
fast visualization of the liver
parenchyma in the early phase of
uptake and concentration of the
99m
Tc-HIDA radiopharmaceutical.
The gallbladder is visualized at
about 15 min, and excreted
radioactivity starts to accumulate
in the bowel already at about
20 min. There is no stasis of
radioactivity in the main intra-
and extrahepatic biliary ducts.
Later frames show nearly
complete clearance of
radioactivity from the liver
parenchyma, the gallbladder
being fully distended with
radioactive bile (this would be an
adequate time to stimulate
gallbladder emptying induced
either by a fat meal or by
administration of the
cholecystokinin analogue
sincalide) (reproduced with
permission from: Volterrani D,
Erba PA, Mariani G, eds.
Applicazioni. Milan: Springer;
2010)
Acute acalculous cholecystitis occurs in less than 10% of Abdominal ultrasound (AUS) is currently regarded as
patients with acute cholecystitis. Most patients with acute the initial test of choice in the evaluation of suspected acute
acalculous cholecystitis have cystic duct obstruction not cholecystitis because it does not expose patients to ionizing
caused by a stone. In some cases, the obstruction is due to radiation and has high diagnostic accuracy for acute chole-
kinking, fibrosis, adhesive bands, anomalous vessels, tumor, cystitis. In addition, AUS can diagnose causes of right upper
or lymphadenopathy. In most cases the cause is inspissated quadrant abdominal pain other than acute cholecystitis (e.g.,
bile, cellular debris, and local edema. This entity typically pancreatitis). AUS is also the most expedient imaging test
occurs in patients who have had multiple traumatic injuries, and can be performed and interpreted by an ultrasound-
extensive serious burns, postoperative complications, shock, certified emergency room physician as accurately as when it
sepsis, or other serious illnesses. Acute acalculous cholecys- is supervised and interpreted by a radiologist. Nevertheless,
titis is associated with a high morbidity and mortality, at least AUS has some disadvantages, such as the operator skill
partly because the disease presents atypically and may be dependence and variability in diagnostic criteria for acute
obscured by the primary illness [38]. cholecystitis among clinical investigators and practitioners.
Anyway, AUS sets a high standard for assessing the useful- dren and respiratory depression in non-ventilated patients.
ness of all other imaging modalities employed for diagnos- Relative contraindication for morphine use is acute pancre-
ing acute cholecystitis. atitis [39].
Acute cholecystitis is the most common clinical indi-
cation for hepatobiliary scintigraphy. If the gallbladder is 31.7.3.2 B iliary Dyskinesia and Gallbladder
visualized, the study is interpreted as negative for acute cho- Ejection Fraction
lecystitis. However, such a straightforward distinction is not Biliary dyskinesia is characterized by symptoms of typical
always possible. Some patients have hepatocellular or biliary biliary pain in the absence of gallstones. The underlying
disease, such as hepatitis or ascending cholangitis, with asso- pathophysiology of pain is probably gallbladder dysmotil-
ciated decreased liver uptake and poor or absent radiotracer ity, with consequent biliary stasis promoting chemical irri-
excretion into the biliary tree. Other patients exhibit good tation and chronic inflammation of the gallbladder wall.
parenchymal uptake but no biliary excretion because of com- Many of these patients will benefit from cholecystec-
mon bile duct obstruction or severe nonobstructive (intrahe- tomy, but the diagnosis and management of this condition
patic) cholestasis. The most appropriate interpretation of remains controversial. The gallbladder ejection fraction
gallbladder non-visualization under these circumstances is (GBEF) can be derived from 99mTc-HIDA scintigraphy
to call the study indeterminate for acute cholecystitis but by inducing gallbladder emptying with either a fat meal
suggestive of either common bile duct obstruction or severe (Fig. 31.17) or a dministration of a synthetic analogue of
intrahepatic cholestasis. cholecystokinin (sincalide for injection); this procedure
Acute cholecystitis is not excluded by an indeterminate can identify patients with gallbladder dyskinesia who are
hepatobiliary scintigraphy and warrants further diagnos- likely to benefit from cholecystectomy. Sincalide stimu-
tic investigation by other means (AUS, CT, or MRI) [40]. lates gallbladder contraction and simultaneous relaxation
In most patients with acute acalculous cholecystitis, the of the sphincter of Oddi, inhibition of gastric emptying,
diagnosis can be made with hepatobiliary scintigraphy. and increase in intestinal motility. A lactose-free fatty meal
However, perhaps a quarter of these patients do not have may be used as a physiological and less expensive option
cystic duct obstruction but rather direct inflammation of the versus the use of sincalide (which may not be widely avail-
gallbladder wall secondary to sepsis, toxemia, or ischemia. able in all countries worldwide). Although there is limited
Since there is no cystic duct obstruction, the gallbladder evidence to show that this can produce a similar range of
may become visualized on cholescintigraphy, thus result- GBEF in the normal population, most guidelines accept
ing in an apparently false-negative finding for acute chole- that sincalide injection provides more standardized and
cystitis. The diagnostic sensitivity of cholescintigraphy for reproducible effects [41].
acute acalculous cholecystitis varies in the literature, and Cholecystokinin is a 33- to 58-amino-acid polypeptide
patient numbers are small; however, it is probably in the hormone. The terminal octapeptide is the physiologically
range of 70–80%, compared with more than 95% for acute active form of the hormone. Sincalide is an analogue of the
calculous cholecystitis [38]. terminal octapeptide. The only contraindications to sincalide
When acute cholecystitis is suspected and the gallblad- include a history of a previous allergic reaction (quite rare)
der is not seen by 30–60 min postinjection, morphine sulfate and pregnancy [38]. There is wide variability in sincalide
(0.04 mg/kg or a standard 2 mg dose) may be administered injection protocols, which makes direct comparison between
intravenously over 2–3 min. If the cystic duct is patent, flow the published series difficult, and also does not allow a
of bile into the gallbladder will be facilitated by morphine- univocally accepted ejection fraction to be set as a definite
induced temporary spasm of the sphincter of Oddi. The threshold for resolving abnormal cases. A few large-scale
intrahepatic biliary tree and common bile duct must contain studies that have been performed using 45–60 min sincalide
radioactive bile, and the tracer activity should be present in infusions have suggested the normal value of GBEF to be
the small bowel at the time of morphine injection. A second greater 38–40%, but the normal range can vary between 35%
injection of radiopharmaceutical (74 MBq) may be neces- and 65% using shorter injection protocols [42].
sary before morphine administration if the remaining liver An expert panel of gastroenterologists, surgeons, and
or biliary tree activity appears insufficient to permit gall- nuclear medicine physicians [42] issued the following rec-
bladder filling; on the other hand, the second radiopharma- ommendations on how to perform and interpret hepatobiliary
ceutical injection can be given as a standard part of the test. scintigraphy with the sincalide stimulation test:
Imaging is continued for another 30–60 min after morphine
administration. This time should be extended if there is poor Presincalide Procedure
hepatocyte function. Absolute contraindication to the use of • Inject a 99mTc-HIDA radiotracer i.v. with the patient
morphine includes increased intracranial pressure in chil- supine on the imaging table.
Pre-meal after fat meal
Fig. 31.17 Left panel: image extracted from hepatobiliary scintigra- gallbladder and passage of radioactive bile into the small bowel. The
phy at the end of the 60-min dynamic acquisition, demonstrating a gall- counts obtained in the ROIs drawn over the gallbladder before and after
bladder fully distended gallbladder and choledochus without obvious the fat meal, respectively, can be used for the equation yielding the gall-
passage of radioactive bile into the duodenum (possible dyskinesis of bladder ejection fraction (GBEF, see text) (reproduced with permission
the sphincter of Oddi). Right panel: static view acquired about 1 h after from: Volterrani D, Erba PA, Mariani G, eds. Fondamenti di Medicina
ingestion a fat meal, demonstrating almost complete emptying of the Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010)
• Acquire images up to 1 h to ensure visualization of the gallbladder. The GBEF should be calculated at 60 min using
gallbladder prior to sincalide infusion. the background- and decay-corrected counts in the gallblad-
• If the gallbladder has not filled by 60 min in a properly der ROI according to the equation:
prepared patient, the finding is reported as abnormal (con-
sistent with either acute or chronic cholecystitis depend- GBEF = ( Maximum - minimum ) / maximum ´ 100.
ing upon the clinical presentation).
• If the gallbladder has filled, place the gamma camera head
in a 35–40° left anterior oblique view to ensure minimal Interpretation
overlap of the gallbladder with duodenum and small bowel. A GBEF less <38% is considered abnormal. Since the GBEF
• Visualization of the small bowel is not necessary prior to upper limit of normal approaches 100%, there is no convinc-
sincalide infusion. ing evidence that a high GBEF bears clinical significance.
Sincalide Infusion Procedure 31.7.3.3 Sphincter of Oddi Dysfunction

• Sincalide at the dose of 0.02 μg/kg should be drawn into a Sphincter of Oddi dysfunction (SOD) is a poorly understood
30–50 mL syringe and diluted with normal saline to the disorder of sphincter dysmotility characterized by a com-
volume of the syringe. bination of biliary-type pain, abnormal liver function tests,
• The syringe should be placed in an infusion pump. and/or bile duct dilation. Some limited indirect evidence
• Tubing between the syringe and the patient should be suggests that SOD is involved in the etiology of recurrent
filled with the sincalide infusate prior to starting the pancreatitis. Although the condition may affect subjects
infusion. with an intact gallbladder, it is usually diagnosed in the set-
• Sincalide should be infused continuously, with the infu- ting of persistent biliary pain postcholecystectomy, with
sion being completed at 60 min. an estimated incidence between 1% and 14%. Treatment
• Dynamic imaging (1 frame/min) should start simultane- often requires endoscopic sphincterotomy, a procedure that
ously with sincalide infusion and stopped at the end of the carries a complication rate approaching 20%; therefore, it
60-min infusion. should not be undertaken in unselected patients. The refer-
ence standard test for SOD is manometry, an invasive test
Computer Processing and Quantification that is frequently complicated by pancreatitis. 99mTc-HIDA
An ROI should be drawn around the gallbladder and a back- scintigraphy can provide valuable diagnostic information
ground liver ROI about 1–2 cm superior and lateral to the in these patients, by demonstrating a delay in bile flow into
Fig. 31.19 Images extracted at 5-min intervals (left to right and top to
bottom) from a dynamic acquisition up to 60 min during hepatobiliary
scintigraphy in a patient previously submitted to cholecystectomy and
Fig. 31.18 Images extracted at 5-min intervals (left to right and top to suffering from dyspepsia. Black arrowheads point to radioactivity con-
bottom) from a dynamic acquisition up to 60 min during hepatobiliary tained in the stomach, demonstrating duodeno-gastric reflux (repro-
scintigraphy in a patient with biliary pain but without biliary stones duced with permission from: Volterrani D, Erba PA, Mariani G, eds.
detected by ultrasound. No passage of radioactivity in the duodenum is Fondamenti di Medicina Nucleare – Tecniche e Applicazioni. Milan:
seen up to 60 min postinjection of 99mTc-HIDA, both the gallbladder Springer; 2010)
and choledochus appearing to be fully distended with radioactive bile.
Considering the absence of biliary stones, this pattern suggests dyskine-
sis of the sphincter of Oddi (reproduced with permission from: ally enable to visualize reflux of the radioactive content from
Volterrani D, Erba PA, Mariani G, eds. Fondamenti di Medicina the duodenum, that has been rendered radioactive by virtue
Nucleare – Tecniche e Applicazioni. Milan: Springer; 2010) of the biliary-excreted 99mTc-HIDA, back into the stomach at
some time during the overall acquisition period (Fig. 31.19).
the duodenum. Several scintigraphic variables are used to
detect SOD, including delayed biliary clearance from the 31.7.3.5 Biliary Atresia
bile ducts into the small bowel (liver hilum to duodenum Biliary atresia presents as neonatal cholestatic jaundice. It is
transit time >10 min), delayed entry of tracer into the bowel caused by a progressive inflammatory sclerosis that obliter-
(>30–60 min) (Fig. 31.18), and failure of sincalide to accel- ates extrahepatic and intrahepatic biliary ducts. Early diag-
erate drainage to accelerate of radioactive bileation [41]. nosis is critical and must be made within the first 60 days of
Augmentation of the morphine dose (0.04 mg/kg morphine life to prevent irreversible liver failure. Treatment requires
sulfate administered intravenously over 2–5 min immedi- palliative hepato-portoenterostomy (Kasai procedure) and,
ately after 99mTc-HIDA injection) can yield >80% sensitivity in more severe cases unresponsive to the Kasai procedure,
for the detection of SOD [43]. even liver transplantation. The major differential diagnosis
includes neonatal hepatitis of various causes.
31.7.3.4 Duodeno-Gastric Reflux The negative predictive value of hepatobiliary scintigra-
Dyspepsia (a generic clinical definition that includes a mul- phy is very high. Patient preparation for cholescintigraphy
tifaceted scenario with symptoms ranging from epigrastric should include phenobarbital to activate liver excretory
pain to slow digestion, nausea, vomiting) is sometimes enzymes and increase bile flow. The scan demonstrates a
caused by altered motility of the upper gastrointestinal tract high-grade biliary obstruction, that is, a persistent hepato-
causing anomalous reflux of the small-bowel contents back gram and no biliary-to-bowel transit over 24 h. Gallbladder
into the stomach. Although such reflux can be episodic, the filling has the same importance as intestinal activity. Given
relatively long acquisition times of hepatobiliary scintig- the high back pressure of the high-grade obstruction seen
raphy (including both the 60-min dynamic acquisition and in biliary atresia, no bile secretion is seen. A short 10-min
additional static views recorded in the following hour) usu- SPECT/CT acquisition can be diagnostic in uncertain cases.
14. Maurer AH. Gastrointestinal motility, Part 1: Esophageal transit

Key Learning Points and gastric emptying. J Nucl Med Technol. 2016;44:1–11.
15. Fattori B, Grosso M, Ursino F, Matteucci F, Mancini V, Rizza E, et al.
• Despite noticeable advances in ultrasound and Clinical applications of oro-pharyngo-oesophageal scintigraphy in
cross-sectional imaging over the last 40 years, the study of dysphagia. Acta Otorhinolaryngol Ital. 2007;27:192–9.
99m
Tc-HIDA scintigraphy remains an important 16. Mariani G, Boni G, Barreca M, et al. Radionuclide gastroesopha-
noninvasive investigation for disorders of the hepa- geal motor studies. J Nucl Med. 2004;45:1004–28.
17. Grosso M, Duce V, Fattori B, et al. The value of oro-pharyngo-
tobiliary system. esophageal scintigraphy in the management of patients with aspira-
• The gallbladder ejection fraction (GBEF) derived tion into the tracheo-bronchial tree and consequent dysphagia. N
from 99mTc-HIDA scintigraphy after stimulation Am J Med Sci. 2015;7:533.
with sincalide (a synthetic analogue of cholecysto- 1
8. Fattori B, Grosso M, Bongioanni P, et al. Assessment of swallow-
ing by oropharyngoesophageal scintigraphy in patients with amyo-
kinin) can identify patients with gallbladder dyski- trophic lateral sclerosis. Dysphagia. 2007;21:280–6.
nesia who may benefit from cholecystectomy. 19. Galli J, Valenza V, D’Alatri L, Gajate Samanes AM, Reale F, La
• A normal GBEF makes symptomatic chronic calcu- Mura F. Validity of schintigraphy in the study of neurogenic dys-
lous cholecystitis unlikely. phagia. Acta Otorhinolaryngol Ital. 2000;20:250–9.
20. Fattori B, Giusti P, Mancini V, Grosso M, Barillari MR, Bastiani L,
• In patients suspected of having sphincter of Oddi et al. Comparison between videofluoroscopy, fiberoptic endoscopy
dysfunction after cholecystectomy, 99mTc-HIDA and scintigraphy for diagnosis of oro-pharyngeal dysphagia. Acta
scintigraphy with the sincalide test is a useful diag- Otorhinolaryngol Ital. 2016;36:395–402.
nostic screening test. 2
1. Abell TL, Camilleri M, Donohoe K, et al. Consensus recommen-
dations for gastric emptying scintigraphy: a joint report of the
American Neurogastroenterology and Motility Society and the
Society of Nuclear Medicine. J Nucl Med Technol. 2008;36:44–54.
22. Donohoe KJ, Maurer AH, Ziessman HA, Urbain JLC, Royal HD,
Martin-Comin J. Procedure guideline for adult solid-meal gastric-
References emptying study 3.0. J Nucl Med Technol. 2009;37:196–200.
23. Tougas G, Chen Y, Coates G, Paterson W, Dallaire C, Pare P, et al.
1. Schall GL, Di Chiro G. Clinical usefulness of salivary gland scan- Standardization of a simplified scintigraphic methodology for
ning. Semin Nucl Med. 1972;2:270–7. the assessment of gastric emptying in a multicenter setting. Am J
2. Abdullah A, Rivas FFR, Srinivasan A. Imaging of the salivary Gastroenterol. 2000;95:78–86.
glands. Semin Roentgenol. 2013;48:65–74. 24. Tougas G, Eaker EY, Abell TL, et al. Assessment of gastric emp-
3. Silvers AR, Som PM. Salivary glands. Radiol Clin N Am. tying using a low fat meal: establishment of international control
1998;36:941–66. values. Am J Gastroenterol. 2000;95:1456–62.
4. Afzelius P, Nielsen M-Y, Ewertsen C, Bloch KP. Imaging of the 25. Maurer AH, Camilleri M, Donohoe K, Knight LC, Madsen JL,
major salivary glands. Clin Physiol Funct Imaging. 2014;36:1–10. Mariani G, et al. The SNMMI and EANM practice guideline for
5. Razek AA, Mukherji S. Imaging of sialadenitis. Neuroradiol J small-bowel and colon transit 1.0. J Nucl Med. 2013;54:2004–13.
2017. https://doi.org/10.1177/1971400916682752. 26. Maurer AH. Gastrointestinal motility, Part 2: Small-bowel and

6. Aksoy T, Kiratli PO, Erbas B. Correlations between histopatho- colon transit. J Nucl Med Technol. 2016;44:12–8.
logic and scintigraphic parameters of salivary glands in patients 27. Mariani G, Paglianiti I. Joint SNMMI and EANM guideline
with Sjögren’s syndrome. Clin Rheumatol. 2012;31:1365–70. for small-bowel and colon transit: an important step towards
7. Angusti T, Pilati E, Parente A, et al. Semi-quantitative analysis of long-awaited standardization. Eur J Nucl Med Mol Imaging.
salivary gland scintigraphy in Sjögren. Clin Oral Investig 2017. 2014;41:405–7.
https://doi.org/10.1007/s00784-016-2034-6. . [Epub ahead of 28. Szarka LA, Camilleri M. Methods for the assessment of small-
print]. bowel and colonic transit. Semin Nucl Med. 2012;42:113–23.
8. Raza H, Khan AU, Hameed A, Khan A. Quantitative evaluation of 29. Winzelberg GG, McKusick KA, Strauss HW. Evaluation of gas-
salivary gland dysfunction after radioiodine therapy using salivary trointestinal bleeding by red blood cells labeled in vivo with
gland scintigraphy. Nucl Med Commun. 2006;27:495–9. technetium-99m. J Nucl Med. 1997;20:1080–6.
9. Klein Hesselink EN, Brouwers AH, de Jong JR, van der Horst- 30. Dam HQ, Brandon DC, Grantham VV, Hilson AJ, Howarth DM,
Schrivers ANA, Coppes RP, Lefrandt JD, et al. Effects of radio- Maurer AH, et al. The SNMMI procedure standard/EANM practice
iodine treatment on salivary gland function in patients with guideline for gastrointestinal bleeding scintigraphy 2.0. J Nucl Med
differentiated thyroid carcinoma: a prospective study. J Nucl Med. Technol. 2014;42:308–17.
2016;57:1685–91. 31. Ell C, May A. Mid-gastrointestinal bleeding: capsule endoscopy
10. Wu C-B, Xi H, Zhou Q, Zhang L-M. The diagnostic value of tech- and push-and-pull enteroscopy give rise to a new medical term.
netium 99m pertechnetate salivary gland scintigraphy in patients Endoscopy. 2006;38:73–5.
with certain salivary gland diseases. J Oral Maxillofac Surg. 32. Lassmann M, Treves ST, EANM/SNMMI Paediatric Dosage
2015;73:443–50. Harmonization Working Group. Paediatric radiopharmaceutical
11. Kazem I. A new scintigraphic technique for the study of the esoph- administration: harmonization of the 2007 EANM paediatric dos-
agus. Am J Roentgenol Radium Ther Nucl Med. 1972;115:681–8. age card (version 1.5.2008) and the 2010 North American consen-
12. Klein HA. Esophageal transit scintigraphy. Semin Nucl Med.
sus guidelines. Eur J Nucl Med Mol Imaging. 2014;41:1036–41.
1995;25:306–17. 33. Mariani G, Pauwels EKJ, Alsharif A, et al. Radionuclide evaluation
13. Shaw DW, Williams RBH, Cook IJ, Wallace KL, Weltman MD, of the lower gastrointestinal tract. J Nucl Med. 2008;49:776–87.
Collins PJ, et al. Oropharyngeal scintigraphy: a reliable technique 34. O’Neill BB, Gosnell JE, Lull RJ, Schecter WP. Cinematic nuclear
for the quantitative evaluation of oral-pharyngeal swallowing. scintigraphy reliably directs surgical intervention for patients with
Dysphagia. 2004;19:36–42. gastrointestinal bleeding. Arch Surg. 2000;135:1076–81.
35. Panes J, Bouhnik Y, Reinisch W, et al. Imaging techniques for 40. Tulchinsky M, Colletti PM, Allen TW. Hepatobiliary scintigraphy
assessment of inflammatory bowel disease: Joint ECCO and in acute cholecystitis. Semin Nucl Med. 2012;42:84–100.
ESGAR evidence-based consensus guidelines. J Crohns Colitis. 41. Lambie H, Cook AM, Scarsbrook AF, Lodge JPA, Robinson PJ,
2013;7:556–85. Chowdhury FU. Tc99m-hepatobiliary iminodiacetic acid (HIDA)
36. Clark JK, Paz DA, Ghahremani GG. Imaging of Meckel’s diver- scintigraphy in clinical practice. Clin Radiol. 2011;66:1094–105.
ticulum in adults: pictorial essay. Clin Imaging. 2014;38:557–64. 42. DiBaise JK, Richmond BK, Ziessman HH, Everson GT, Fanelli
37. Spottswood SE, Pfluger T, Bartold SP, et al. SNMMI and EANM RD, Maurer A, et al. Cholecystokinin-cholescintigraphy in adults:
practice guideline for Meckel diverticulum scintigraphy 2.0. J Nucl consensus recommendations of an interdisciplinary panel. Clin
Med Technol. 2014;42:163–9. Gastroenterol Hepatol. 2011;9:376–84.
38. Ziessman HA. Hepatobiliary scintigraphy in 2014. J Nucl Med 43. Thomas PD, Turner JG, Dobbs BR, Burt MJ, Chapman BA. Use
Technol. 2014;42:249–59. of 99mTc-DISIDA biliary scanning with morphine provocation for
39. Tulchinsky M, Ciak BW, Delbeke D, Hilson A, Holes-Lewis KA, the detection of elevated sphincter of Oddi basal pressure. Gut.
Stabin MG, et al. SNM practice guideline for hepatobiliary scintig- 2000;46:838–41.
raphy 4.0. J Nucl Med Technol. 2010;38:210–8.
Hybrid Imaging for Malignant
Conditions of the Gastrointestinal Tract 32
Joan Duch and Albert Flotats
Contents
32.2 Oesophageal Cancer 842
32.2.1 Indications for [18F]FDG PET/CT 843
32.2.2 [18F]FDG PET/CT Interpretation and Reporting 844
32.3 Gastric Cancer 845
32.3.1 Indications of [18F]FDG PET/CT 845
32.4 Pancreatic Cancer (Adenocarcinoma) 846
32.4.1 Indications for [18F]FDG PET/CT 846
32.5 Colorectal Cancer 847
32.6 Anal Cancer 850
32.7 Hepatobiliary Tract Cancer (Cholangiocarcinoma) 851
32.8 Hepatocellular Cancer 852
32.9 Contraindications of [18F]FDG PET/CT (and/or Possible Side Effects) 852
32.10 [18F]FDG PET/CT Acquisition 852
32.11 [18F]FDG PET/CT Process 855
32.12 [18F]FDG PET/CT Interpretation and Reporting 855
J. Duch
Nuclear Medicine Department, Hospital de la Santa Creu i Sant
A. Flotats (*)
Nuclear Medicine Department, Hospital de la Santa Creu i Sant
Universitat Autònoma de Barcelona, Barcelona, Spain
e-mail: aflotats@santpau.cat

842 J. Duch and A. Flotats
32.13 Future Perspectives 855

32.13.1 Harmonized Quantitative Interpretation 855
32.13.2 Radiopharmaceuticals Other than [18F]FDG 855
32.13.3 State-of-the-Art PET/CT and Contrast Media in PET/CT 855
32.13.4 PET/MR Imaging 855
References 856
Learning Objectives cated for treatment response assessment, radiotherapy treatment

• Briefly summarize the main issues concerning epidemiol- planning, and detection of recurrent disease. Moreover, PET/
ogy, aetiology, and the most common clinical presenta- CT may help in decision-making by detecting distant metasta-
tion for cancers of the gastrointestinal tract (oesophagus, ses especially in potentially resectable colorectal cancer (CRC),
stomach, colon, rectum, anus), for pancreatic cancer and not only when disease is limited to one site but also when plan-
for hepatobiliary cancers. ning therapy for curative treatment of limited metastatic disease.
• Briefly summarize the role of imaging other than However, some mucinous tumours cannot be well characterized
radionuclide-based methods for the cancers of the gastro- by [18F]FDG, because the lesions have a minimal increase in
intestinal tract, pancreatic cancer, and hepatobiliary can- glucose metabolism, making them difficult or impossible to
cers in the clinical course of neoplastic disease (diagnosis, identify on [18F]FDG PET/CT. Some gastric and pancreatic
staging, assessment of response to treatment, and re- cancers fall into this category. Finally, incidental focal colonic
evaluation in case of recurrence). [18F]FDG uptakes may require exploration by colonoscopy, as
• Emphasize the role of hybrid imaging with [18F]FDG they are often associated with (pre)malignant lesions.
PET/CT in key moments over the clinical course of dis- A focal region of increased [18F]FDG uptake is not spe-
ease in patients with cancers of the gastrointestinal tract, cific for a neoplasm, since physiologic [18F]FDG uptake is
pancreatic cancer, and hepatobiliary cancers. often heterogeneous and increased focal uptake occurs in
• Detail the established indications of [18F]FDG PET/CT in areas of inflammation.
patients with cancers of the gastrointestinal tract, pancre-
atic cancer, and hepatobiliary cancers.
• Indicate how to interpret and report the [18F]FDG PET/ 32.2 Oesophageal Cancer
CT findings in patients with cancers of the gastrointes-
tinal tract, pancreatic cancer, and hepatobiliary OC is the 13th most common tumour worldwide. It is more
cancers. common in males (ratio 4:1) and represents ~1% of all can-
• Indicate how to use semiquantitative and quantitative cers diagnosed in the USA, but it is much more common
parameters derived from [18F]FDG PET/CT imaging. elsewhere (Iran, Northern China, India, and Southern Africa)
• Summarize the perspectives for the use of PET imaging [1, 2].
agents other than [18F]FDG in patients with cancers of the The general outcome of OC remains very poor (overall
gastrointestinal tract, pancreatic cancer, and hepatobiliary 5-year survival rates are ~10%, and 5-year post-
cancers. oesophagectomy survival rates are ~15–40%) [3], mainly
• Summarize the available clinical evidence cumulated so because most OCs are diagnosed in an advanced stage.
far with the use of [18F]FDG PET/MR in patients with Development of OC has been related to genetic factors
cancers of the gastrointestinal tract, pancreatic cancer, together with smoking and alcohol consumption (for squa-
and hepatobiliary cancers. mous cell carcinoma, ~90% of cases, typically found in the
upper two-thirds of the oesophagus) and obesity and reflux
(for adenocarcinoma, ~10% of cases but rapidly increasing
32.1 Introduction in incidence, typically in the distal oesophagus and at the
gastroesophageal junction).
Hybrid imaging with 2-deoxy-2-[18F]fluoro-d-glucose ([18F] OC most frequently presents with progressive dysphagia
FDG) positron emission tomography/computed tomography and weight loss. The main diagnostic procedure of OC is
(PET/CT) has become a powerful imaging tool in oncology. upper endoscopy with biopsy, guided with endoscopic ultra-
The combination of data on local glucose utilization and the sonography (US) when the lesion does not collapse the
relationship of metabolism to local lymph nodes and/or specific lumen. The best imaging tool for diagnosing the exact depth
sites in organs enhance the sensitivity and specificity of the of invasion into the oesophageal wall and local invasion to
hybrid imaging findings. PET/CT is currently important in the adjacent structures is contrast-enhanced CT (ceCT), which
initial workup of oesophageal and anal cancers, additional data frequently shows a loss of the fat planes between the tumour
supporting [18F]FDG PET/CT in other gastrointestinal (GI) and adjacent structures in the mediastinum and displacement
malignancies. In addition, [18F]FDG PET/CT has been advo- or indentation of adjacent structures [4].
32 Hybrid Imaging for Malignant Conditions of the Gastrointestinal Tract 843
Staging and prognostic groups of OC are based on the and neoadjuvant/palliative treatment when disease is more
TNM classification from the American Joint Committee on widespread. Despite having insufficient accuracy for detect-
Cancer (AJCC) (Table 32.1). Staging should include a ceCT ing tumour depth (T staging) as compared to endoscopic US
of the neck, chest, and abdomen. (which has a sensitivity of 85–90%), [18F]FDG PET/CT can
detect distant metastases (Fig. 32.1), especially extra-
mediastinal lymph nodes [5]. The accuracy of [18F]FDG
32.2.1 Indications for [18F]FDG PET/CT PET/CT for assessing the nodal status varies widely, but a
specificity of 90% has been reported. The false-positive
[18F]FDG PET/CT is useful for staging and guiding manage- results are often related to inflammatory processes
ment of OC: surgery when there is local/locoregional disease (Table 32.2) [6]. Since the oesophagus has a rich submucosal
lymphatic drainage and no serous wall, nodal spread in OC
Table 32.1 Oesophageal cancer TNM definitions (AJCC 2017) is frequent, not only for T3–T4 tumours but also for T1–T2
Primary tumour (T) tumours. Endoscopic US combined with [18F]FDG PET/CT
Tx Primary tumour cannot be assessed increases the sensitivity to detect locoregional nodal involve-
T0 No evidence of primary tumour ment improving the differentiation of nodal infiltration in the
Tis High-grade dysplasia vicinity of the OC lesion.
T1 Tumour invades the lamina propria, muscularis mucosae, [18F]FDG PET/CT can detect metastatic OC in unex-
or submucosa
T1a Tumour invades the lamina propria or muscularis mucosae
pected locations (Fig. 32.1) [7, 8]. In a meta-analysis, the
T1b Tumour invades the submucosa sensitivity and specificity of [18F]FDG PET/CT for the detec-
T2 Tumour invades the muscularis propria tion of distant metastases were 67% and 97%, respectively
T3 Tumour invades the adventitia [9], superior to CT. Early use of [18F]FDG PET/CT in the
T4 Tumour invades adjacent structures staging of OC identifies unsuspected metastases in up to
T4a Resectable tumour invading the pleura, pericardium, or 20% of patients.
diaphragm [18F]FDG PET/CT has been assessed for early prediction
T4b Unresectable tumour invading other adjacent structures,
and monitoring response to therapy (including neoadjuvant
such as the aorta, vertebral body, and trachea
Regional lymph nodes (N)
therapy); a reduction of [18F]FDG uptake (35% of maximal
Nx Regional lymph node(s) cannot be assessed standard uptake value—SUVmax) after two cycles of chemo-
N0 No regional lymph node metastasis therapy can differentiate responders from non-responders,
N1 Metastasis in 1–2 regional lymph nodes with a sensitivity and specificity of 93% and 95%, respec-
N2 Metastasis in 3–6 regional lymph nodes tively [10].
N3 Metastasis in 7 or more regional lymph nodes Local recurrence in patients treated with surgery is ~30%.
Distant metastasis (M) Endoscopic US or CT often cannot differentiate local recur-
M0 No distant metastasis rence from postsurgical fibrosis or oedema. In a recent meta-
M1 Distant metastasis
Fig. 32.1 Maximum intensity projection [18F]FDG PET image (on the The fused PET/CT axial slices (on the right) show muscular metastatic
left) and sagittal slices of CT, PET, and fused PET/CT (in the middle) disease
showing an oesophageal tumour in the middle thoracic oesophagus.
Table 32.2 [18F]FDG PET/CT pitfalls [18F]FDG PET/CT can detect nodal metastases out of the
False-negative results initial planned radiation field, improving radiotherapy plan-
Tumour size (small), type (mucinous, cystic, with low GLUT1 ning [12].
transporters) and location
Non-tumour increased [18F]FDG masking tumour uptake
• normal stomach (diffuse or regional, most typically involving
proximal stomach) may mask GC. Gastric distension by water 32.2.2 [18F]FDG PET/CT Interpretation
or water-based contrast agent reduces [18F]FDG uptake and Reporting
• normal large and small bowel (intense and frequent in diabetics
receiving metformin) may mask tumour in the bowel and
mesentery OC location is determined by the position of the upper edge
• normal liver (diffuse and somewhat heterogeneous) of the tumour in the oesophagus (Table 32.3). Tumours in the
• anastomosis (masking local tumour recurrence) gastroesophageal junction are considered OC if tumour epi-
Elevated blood glucose levels (frequent in patients with pancreatic centre is within the lower thoracic oesophagus, at the gastro-
pathology)
esophageal junction, or within the proximal 5 cm of the
False-positive results
Non-tumour increased [18F]FDG activity in
stomach (if the tumour extends into the oesophagus).
• normal oesophagus (focal or diffuse) Although T1 and T2 tumours are usually seen on [18F]
• normal gastroesophageal sphincter uptake (typically FDG PET/ceCT as an asymmetric thickening of the
circumferential, uniform, and without mass/soft tissue oesophageal wall, the lesions should be confirmed by biopsy
irregularity)
• hiatal hernia
under endoscopic US guidance. T3 tumours can be seen as
• normal stomach (diffuse or regional, most typically involving an eccentric or circumferential wall thickening >5 mm or as
proximal stomach) an oesophageal mass that causes luminal obstruction (and
• pancreas serous cystadenoma proximal dilation), with soft tissue and fat stranding that
• normal large and small bowel, benign polyps
• hepatic adenomas, hepatic haemangiomas, and nodular focal
generate a feathery appearance, showing a weak and diffuse
hyperplasia [18F]FDG uptake related to the perioesophageal fat
• normal lymphoid tissue of the terminal ileum and caecum invasion.
• anal sphincter uptake (circumferential, uniform, and without The presence of tracheobronchial, aortic, and bone
mass/soft tissue irregularity)
invasion is important because it determines the suitability
Inflammatory and infectious processes
• reactive lymph nodes of resection (T4a tumours are resectable, whereas T4b are
• oesophagitis not). Tracheobronchial invasion can be suspected when
• gastritis there is a discrete indentation on the posterior tracheal wall
• peptic ulcer disease or displacement of the trachea/bronchus by the tumour.
• pancreatitis (acute or chronic active)
• biliary stent (likely due to both procedural inflammation and Aortic invasion is suggested when the contact area between
attenuation artefact) the tumour and aorta is >90% or if there is obliteration of
• recent endoscopic retrograde cholangiopancreatography the triangular fat space between the oesophagus, aorta, and
• retroperitoneal fibrosis spine adjacent to the primary tumour.
• cholangitis
• hepatic abscess Detection of pathologic lymph nodes by CT depends pri-
• colitis (segmental or diffuse) marily on size criteria (>1 cm in short-axis diameter or 5 mm
• diverticulitis (associated diverticular disease with pericolonic in supraclavicular location), but normal-sized lymph nodes
fat stranding) can be invaded, which can be discovered by an increase of
• granulomatous disease
• biopsy sites [18F]FDG uptake.
Recent surgery and chemoradiation (due to the induced
inflammatory response)
There are no definitive recommendations on how long to wait after
Table 32.3 Oesophageal cancer location classification
such treatment to perform an [18F]FDG PET/CT, but it is considered
adequate for measurement of response an interval of: OC
• >6 weeks after surgery location Position of the upper edge of the tumour
• >10 days after chemotherapy Cervical Between the cricopharyngeus muscle and the level of
• >2 months after radiotherapy the sternal notch; 15–20 cm from the incisors at
oesophagoscopy
Upper Between the sternal notch and the azygos arch;
thorax 20–25 cm from the incisors
analysis, [ F]FDG PET/CT had a sensitivity of 96%, and a
18
Middle Between the azygos arch and the level of the inferior
specificity of 78% for recurrence. Histologic confirmation of thorax pulmonary vein; 25–30 cm from the incisors
suspected lesions is required due to the false-positive rate Lower At lower oesophageal sphincter; 30–40 cm from the
(mainly due to inflammation) [11]. thorax incisors
Distant metastases have been reported at initial presenta- associated with hereditary nonpolyposis colorectal cancer
tion in 20–30% of OC, especially in the liver (35%), lungs (CRC), familial adenomatous polyposis CRC, hereditary dif-
(20%), bones (9%), and adrenal glands (5%) and rarely in the fuse GC, gastric adenocarcinoma and proximal polyposis of
peritoneum and brain. On CT, liver metastases usually appear the stomach, and Peutz-Jeghers syndrome.
as hypoattenuating lesions that are best visualized during the Adenocarcinomas constitute 90% of GC. Less common
portal venous phase of liver enhancement. Pulmonary metas- GC such as GI stromal tumours (GISTs), lymphomas, and
tases are usually of round shape, smooth-bordered, and non- neuroendocrine tumours are covered elsewhere.
calcified on CT. [18F]FDG PET/CT has the advantage of GC most frequently presents with progressive upper abdom-
allowing total-body coverage, and its primary role is to depict inal discomfort, commonly with weight loss, anorexia, early
distant metastases, with superior capability as compared to satiety, nausea, vomiting (particularly when lesions are located
CT, especially in the liver and bone. Moreover, [18F]FDG in the pylorus (or with widespread disease)), dysphagia (when
PET/CT can depict metastases in unexpected locations, lesions are located in the cardia), and GI bleeding. If a diagnosis
including the brain, skeletal muscles (Fig. 32.1), subcutane- of GC is suspected, it should be confirmed by gastroscopy and
ous tissues, thyroid gland, and pancreas. biopsy, reporting histology according to the WHO criteria.
GC spreads primarily to the local lymph nodes, liver, and
peritoneum. Systemic spread is not common; metastases go
first to the liver and then to the lungs, adrenal glands, kid-
Key Learning Points
neys, bones, and brain.
• [18F]FDG PET/CT is not sufficiently accurate for T
Initial staging and risk assessment should include ceCT of
staging as compared to endoscopic US but has a
the thorax and abdomen ± pelvis to detect local/distant lymph-
better diagnostic performance to detect distant
adenopathy and metastatic disease or ascites. Laparoscopy is
metastases as compared to CT.
recommended for patients with resectable GC.
• When recurrence is suspected, [18F]FDG PET/CT
can help to differentiate it from fibrosis and oedema,
but due to high false-positive findings, histologic
32.3.1 Indications of [18F]FDG PET/CT
confirmation is required.
• [18F]FDG PET/CT can be useful for radiotherapy
[18F]FDG PET/CT has a detection rate of only ~ 55% for the
planning since it can identify nodal infiltration ini-
primary tumour and therefore has no role in the diagnosis
tially not considered to be included in the radiation
and T staging of GC. The lesions are often less metabolically
field, thus potentially improving treatment results.
active [14–16].
[18F]FDG PET/CT may improve GC staging by detecting
involved lymph nodes or metastatic disease as it has a signifi-
32.3 Gastric Cancer cantly higher accuracy in preoperative staging (68%) than
CT (53%) [17, 18].
GC is the sixth most common tumour worldwide, with an The National Comprehensive Cancer Network (NCCN)
incidence of 7% of all tumour types [13]. It is more common recommends PET/CT for initial staging when there is no evi-
in males (ratio 2:1), with the highest rates in Eastern Asia, dence of metastases (but it may not be appropriate for T1
Eastern Europe, and South America and lower rates in North tumours) and for post-treatment assessment in unresectable
America and Western Europe, where a gradual decline has disease or non-surgical candidates following primary treatment
been registered. More recent declines in high-prevalence if they have renal failure or allergy to intravenous (IV) iodin-
countries have also been reported and differ from the relative ated contrast medium [18]. [18F]FDG PET/CT is also useful for
increase in tumours of the gastroesophageal junction. GC is predicting response to preoperative chemotherapy [19], as well
the fourth most common cancer-related cause of death in the as for the evaluation of recurrence when ceCT is non-diagnos-
world (8.8% of all annual cancer deaths) [13]. tic [20, 21]. However, it is not sensitive for detecting peritoneal
Distal or antral GC is associated with risk factors such as involvement, which requires laparoscopy and biopsy [18].
H. pylori infection, alcohol consumption, and diets with high
salt, large quantities of processed meat, and/or low fruit and
vegetable intake. Proximal GC (cardia) is associated with 32.3.2 [18F]FDG PET/CT Interpretation
obesity. Tumours of the gastroesophageal junction are asso- and Reporting
ciated with reflux and Barrett’s oesophagus and predominate
in non-Asian countries. Some GC is familial (~10% of Focal or diffuse thickening of the gastric wall detected on CT
cases). There is an inherited genetic predisposition in 1–3% is an important but nonspecific sign of gastric disease. With
adequate distension of the stomach with air or contrast agent, PC may present with obstructive jaundice, weight loss,
wall thickening >5 mm can be considered abnormal. The pri- and abdominal or mid-back pain. Imaging has led to
mary tumour may appear as a focal, nodular, or irregular increased detection of pancreatic abnormalities, with ~33%
thickening of the gastric wall or as a polypoid intraluminal of pancreatic incidental findings finally resulting in PC [24].
mass of soft tissue attenuation. CeCT is the best-validated and most widely available modal-
Extension of the tumour into the perigastric fat (very com- ity for diagnosis and initial management [25].
mon when wall thickness >2 cm) blurs the serosal surface,
and strands and nodules of tumour are seen in the adjacent fat.
Malignant lymph nodes are frequently of round shape and 32.4.1 Indications for [18F]FDG PET/CT
have short-axis diameter >6 mm in perigastric region, with
central necrosis and heterogeneous or high enhancement. For staging, [18F]FDG/ceCT or PET/MR is helpful to
However, the sensitivity of CT for lymph node staging is detect both local extent of disease and identify metastasis,
variable (62.5–91.9%) [22], and global consensus is lacking particularly in patients with high risk of metastases (bor-
on specific diagnostic criteria. Nodes near the celiac axis and derline resectable disease, markedly elevated carbohydrate
in the gastrohepatic ligaments and those showing high FDG antigen 19-9 (CA 19-9), large primary tumours, large
uptake are most likely to be involved. regional lymph nodes, or very symptomatic patients). [18F]
Local recurrence appears as focal wall thickening at the FDG PET/CT can also be useful to distinguish recurrence
anastomosis or in the remaining stomach. Nodal recurrence from fibrosis after surgical or radiotherapy treatment [26,
is most common along the course of the hepatic artery or in 27]. The use of [18F]FDG PET/CT following ceCT showed
the para-aortic region. Peritoneal recurrence is seen in the increased sensitivity for the detection of metastatic disease
cul-de-sac, on parietal peritoneal surfaces, or on the surface when compared with ceCT and [18F]FDG PET/CT alone in
of the bowel. a retrospective study. The sensitivity of detecting metasta-
ses for PET/CT alone, ceCT, and the combination of both
were 61%, 57%, and 87%, respectively, and additionally,
Key Learning Points the clinical management of 11% of patients with invasive
• [18F]FDG PET/CT has no role in the diagnosis and pancreatic cancer was changed as a result of the [18F]FDG
T staging of GC since it is associated with a low PET/CT [28].
detection rate.
• [18F]FDG PET/CT is recommended for initial stag-
ing when there is no evidence of metastatic disease 32.4.2 [18F]FDG PET/CT Interpretation
and for post-treatment assessment in unresectable and Reporting
disease or non-surgical candidates following pri-
mary treatment if they have renal insufficiency or PC usually appears as a hypoattenuating mass within pancre-
allergy to IV contrast. atic parenchyma on CT. However, in ~10% of cases, the
• [18F]FDG PET/CT is not sensitive to detect perito- lesion is isoattenuating, especially small tumours (≤2 cm),
neal disease. thus making diagnosis difficult. Indirect signs of obstruction
of pancreatic and common bile duct such as upstream
pancreatic duct dilation or the double-duct sign (i.e. the pres-
ence of simultaneous dilatation of the common bile and pan-
32.4 Pancreatic Cancer creatic ducts) are helpful for diagnosis [29].
(Adenocarcinoma)
The incidence of PC is steadily increasing in most countries. Key Learning Points

PC is the seventh most common cause of cancer death, with • [18F]FDG PET CT should be considered when there
overall 5-year survival of 5.5%. PC presents as a locally are equivocal findings on ceCT or MR.
advanced disease in 30%, and <20% of patients are consid- • [18F]FDG PET/CT has higher sensitivity than CT
ered for curative surgery. Furthermore, PC presents with for the detection of metastases.
metastases in >50% and may only be treated with palliative • [18F]FDG PET/CT can be useful to distinguish
chemotherapy. Among the patients who survive surgical recurrence from fibrosis after surgical or radiother-
resection, the 5-year survival rate remains low (~15–40%) apy treatment.
[23]. Surgeons often underestimate the extent of disease and • [18F]FDG PET/CT improves the evaluation of
resectability; therefore, improvement of current imaging recurrent PC, especially in the setting of elevated
modalities to better diagnose and stage PC may help avoid CA 19-9 and negative or equivocal CT findings.
futile surgeries.
32.5 Colorectal Cancer Table 32.4 Colorectal cancer TNM definitions (AJCC 2017)
Primary tumour (T)
CRC is the third most common tumour in men and second in TX Primary tumour cannot be assessed
women, with an incidence of 10% of all tumour types world- T0 No evidence of primary tumour
wide [1, 30]. Incidence is higher in males (ratio 1.2:1). Tis Carcinoma in situ, intramucosal carcinoma (involvement
of the lamina propria with no extension through the
Risk factors for CRC include a family history of colon muscularis mucosae)
cancer in a first-degree relative, a history of colorectal ade- T1 Tumour invades the submucosa (through the muscularis
nomas or ovarian cancer, personal history of chronic ulcer- mucosae but not into the muscularis propria)
ative colitis or Crohn’s colitis, and diets high in red or T2 Tumour invades the muscularis propria
processed meat and low in fibre. These patients are often T3 Tumour invades through the muscularis propria into
pericolorectal tissues
obese and are not physically active. Several genetic factors
T4 Tumour invades the visceral peritoneum or invades or
are associated with CRC including Lynch syndrome and adheres to adjacent organ or structure
familial adenomatous polyposis [31]. Chronic nonsteroidal T4a Tumour invades through the visceral peritoneum
anti-inflammatory treatment, especially aspirin, is associated (including gross perforation of the bowel through tumour
with reduced incidence. and continuous invasion of tumour through serial of
CRC is the fifth most common cancer-related cause of inflammation to the surface of the visceral peritoneum)
T4b Tumour directly invades or adheres to adjacent organs or
death in the world (8.5% of all annual cancer deaths) [13]. The structures
CRC-related 5-year survival rate approaches 60% [32]. Regional lymph nodes (N)
Mortality has declined in many Western countries, mainly due NX Regional lymph nodes cannot be assessed
to cancer screening programmes, removal of colorectal adeno- N0 No regional lymph node metastasis
mas, early detection of cancerous lesions, and availability of N1 Infiltration of 1–3 regional lymph nodes (tumour in lymph
more effective therapies, mostly for early-stage disease. node measuring ≥0.2 mm) or any number of tumour
deposits are present, and all identifiable lymph nodes are
Left-sided CRC presents most commonly with rectal negative
bleeding, altered bowel habits, and abdominal or back pain. N1a Infiltration of 1 regional lymph node
Cecal and ascending CRC usually presents with symptoms N1b Infiltration of 2–3 regional lymph nodes
of anaemia, occult blood in stool, or weight loss. N1c No regional nodal infiltration, but there are tumour
Complications of CRC include bowel obstruction, volvulus, deposits in the subserosa, mesentery, or non-
perforation, and fistulae. peritonealized pericolic or perirectal/mesorectal tissue
metastasis
Spread of CRC includes direct extension with penetration
N2 Infiltration of ≥4 regional lymph nodes
of the colon wall, lymphatic drainage to regional nodes, hae- N2a Infiltration of 4–6 regional lymph nodes
matogenous drainage through portal veins to the liver or N2b Infiltration of ≥7 regional lymph nodes
through systemic paravertebral venous to the lungs (rectosig- Distant metastasis (M)
moid CRC), and intraperitoneal seeding. CRC metastasizes M0 No distant metastasis by imaging, etc.; no evidence of
in ~50% of patients (at initial diagnosis in ~25% of patients). tumour in distant sites or organs (this category is not
Metastatic CRC (mCRC) contributes to the high mortality assigned by pathologists)
M1 Metastasis to ≥1 distant site or organ or peritoneum
rate of this malignancy [32]. The clinical outcome for mCRC
M1a Metastasis to 1 site or organ without peritoneal metastasis
has improved mainly due to earlier detection of metastatic
M1b Metastasis to ≥2 sites or organs without peritoneal
disease and improvement in the efficacy of systemic thera- metastasis
pies. Staging of CRC is important in determining therapy M1c Metastasis to the peritoneal surface, alone or with other
and prognosis. Treatment of CRC is determined primarily by site or organ metastases
TNM staging (Table 32.4). The main procedure for diagnosis
of CRC is biopsy during endoscopy by either sigmoidoscopy
(as >35% of tumours are located in the rectosigmoid) or, when other imaging modalities are equivocal regarding the
preferably, total colonoscopy. Abdominopelvic ceCT is indi- presence of distant metastases.
cated for subsequent staging. MR is best for the detection In case of elevated tumour markers (carcinoembryonic anti-
and characterization of small lesions in liver and gen, CEA) without knowledge of the location of relapse or in
peritoneum. the presence of equivocal or suspicious lesions on ceCT, [18F]
FDG PET/CT is indicated for the detection of local recurrence
at the site of the initial colorectal surgery and especially for the
32.5.1 Indications of [18F]FDG PET/CT detection of metastatic disease (Figs. 32.2 and 32.3) [34–36]. A
stepwise imaging approach is recommended to evaluate
Routine use of [18F]FDG PET/CT is not recommended for patients with suspected CRC, starting with a contrast-enhanced
initial staging, since it does not modify the treatment CT of the chest, abdomen, and pelvis. If the findings on CT are
approach in most patients [33]. However, it can be performed not diagnostic, an additional evaluation with MR is helpful to
Fig. 32.2 Maximum intensity projection [18F]FDG PET image (on the left), with corresponding axial slices of CT, PET, and fused PET/CT local-
izing recurrence of colorectal cancer as small but hypermetabolic liver metastases
a b
Fig. 32.3 Maximum intensity projection [18F]FDG PET image (a), with corresponding axial slices of CT (b), PET (c), and fused PET/CT (d)
localizing recurrence of colorectal cancer as a solitary right pulmonary metastatic nodule with increased [18F]FDG uptake
Fig. 32.4 Maximum intensity

projection [18F]FDG PET image a b
(a), with corresponding axial
slices of CT (b), PET (c), and
fused PET/CT (d) showing an
incidental hypermetabolic right
colon polyp, which resulted in
malignancy after colonoscopy
and biopsy
evaluate the liver, peritoneum, and pelvis, while PET/CT is hypometabolism and low attenuation due to haemorrhage or
helpful to detect extrahepatic disease (Fig. 32.3). necrosis can be seen. Air within the tumour indicates ulcer-
The role of [18F]FDG PET/CT for the assessment of response ation. Loss of fat planes between the tumour and adjacent
after chemoradiation is under investigation, but not superior to structures suggests local invasion.
CT except for rectal cancer. The extent of subsequent rectal sur- CRC recurrences are uncommon, occurring in 6.4% of
gery should not be modified yet based on this [37]. patients within 5 years of surgery. The risk of developing
recurrence is highest within the first year after surgery
[38]. Recurrences occur at the site of the original tumour
32.5.2 [18F]FDG PET/CT Interpretation in ~50% of the cases (location of the original tumour may
and Reporting be identified by high-attenuation metallic bowel clips in
the anastomosis), whereas the remainder recur at distant
CRC may be a colon polyp seen incidentally as hypermeta- sites, especially in the liver (Fig. 32.2). Presacral soft tis-
bolic well-defined oval or round intraluminal projection. sue densities seen in patients with abdominoperineal
Therefore, incidental focal colonic [18F]FDG uptake may resection may represent a recurrence or fibrosis. While
require colonoscopy to rule out (pre)malignancy (Fig. 32.4). recurrent tumour shows increased [18F]FDG uptake and
CRC can also appear as a larger hypermetabolic intralu- tends to be nodular and convex anteriorly, fibrosis tends to
minal mass with nodular contours and irregular mucosal sur- be more uniform and concave anteriorly, without hyperme-
faces that may narrow the lumen of the colon. Central mass tabolism (unless inflamed).
Table 32.5 Anal cancer TNM definitions (AJCC 2017)

Key Learning Points Primary tumour (T)
• PET/CT is not routinely indicated for initial staging TX Primary tumour cannot be assessed
but can be performed in addition to MR in the case T0 No evidence of primary tumour
of resectable metastatic disease (initial workup or Tis Carcinoma in situ (Bowen disease, high-grade squamous
intraepithelial lesion [HSIL], AIN II–III
recurrence) to exclude the presence of other meta-
T1 Tumour ≤2 cm in greatest dimension
static sites. T2 Tumour >2 cm but ≤5 cm in greatest dimension
• Consider [18F]FDG PET/CT in front suspicion of T3 Tumour >5 cm in greatest dimension
recurrence in the case of serial CEA elevation, after T4 Tumour of any size invades adjacent organ(s) (e.g. vagina,
normal colonoscopy and CT for colon cancer, and urethra, bladder); direct invasion of the rectal wall,
as first-line imaging in the case of rectal cancer. perirectal skin, subcutaneous tissue, or the sphincter
muscle(s) is not classified as T4
• Colonoscopy is recommended in the case of inci-
Regional lymph nodes (N)
dental focal colonic [18F]FDG uptake. NX Regional lymph nodes cannot be assessed
N0 No regional lymph node metastasis
N1 Metastasis in perirectal lymph node(s)
32.6 Anal Cancer N2 Metastasis in unilateral internal iliac and/or inguinal lymph
node(s)
N3 Metastasis in perirectal and inguinal lymph nodes and/or
Epidermoid anal cancer or squamous cell carcinoma of the
bilateral internal iliac and/or inguinal lymph nodes
anus (SCCA) is an infrequent tumour (2–4% of CRC and Distant metastasis (M)
anal tumours). The incidence of SCCA is higher in females, M0 No distant metastasis
and it is increasing due to the strong association with human M1 Distant metastasis
papillomavirus (HPV) infection, which is the contributing AIN Anal intraepithelial neoplasia
agent in 80–85% of patients [39]. Factors increasing the risk
of HPV infection (anal intercourse and a high lifetime num-
ber of sexual partners) and/or modulating host response and After completion of chemoradiation therapy, SCCA tends
the persistence of HPV infection (human immunodeficiency to regress slowly. Clinical rectal and inguinal examinations
virus (HIV), immune suppression, a history of other HPV- are the mainstay of determining complete response. Pelvic
related cancers, autoimmune disorders, social deprivation, MR or CT is also necessary. To date, few [18F]FDG PET/CT
and smoking) have also influence on the epidemiology of studies have assessed treatment response, and the timing of
SCCA. assessment is controversial. The benefit of PET/CT is to
SCCA prognosis has remained stable in the last two detect residual subclinical pelvic or extrapelvic/para-aortic
decades, and 5-year survival rates are ~70%. node involvement.
SCCA commonly presents with bleeding, which may be
associated with a mass, non-healing ulcer, pain, itching, dis-
charge, faecal incontinence, and fistulae. The diagnosis is made 32.6.2 [18F]FDG PET/CT Interpretation
on biopsy-proven histology. At diagnosis, lymph node involve- and Reporting
ment is present in 30–40% of patients, whereas systemic spread
is uncommon (5–8%). Imaging should include pelvic MR or When searching for nodal involvement, one should bear in
endo-anal US. [18F]FDG PET/CT has a high sensitivity in iden- mind that drainage of anal canal lesions differs from that of
tifying involved lymph nodes, causing a change in staging in anal margin tumours. Proximal lesions drain to perirectal
~20% of patients, and changes patient management in ~3–5% nodes along the inferior mesenteric artery. Immediately
of cases. Treatment is mostly determined by TNM staging above the dentate line of the anal canal, drainage is to inter-
(Table 32.5), usually with locoregional scope owing to the low nal pudendal nodes and to the internal iliac system. Below
rate of distant metastases and aimed to achieve cure with the dentate line of the anal canal and perianal skin, drainage
locoregional control and preservation of anal function. is to the inguinal, femoral, and external iliac nodes.

Key Learning Point
[18F]FDG PET/CT has a high sensitivity for the detection of • [18F]FDG PET/CT has a high sensitivity in identify-
involved lymph nodes, thus influencing radiation therapy ing involved lymph nodes in SCCA and is recom-
planning, and is recommended in the diagnostic workup of mended for treatment planning.
SCCA.
32.7 Hepatobiliary Tract Cancer nation and 5-year overall survival prognosis. Abdomen
(Cholangiocarcinoma) ceMR cholangiopancreatography and ceCT are the primary
modalities for staging.
CC refers to both intrahepatic and extrahepatic tumours of
the bile tract, including gallbladder and ampulla of Vater.
Although it represents only 3% of all GI malignancies [40], 32.7.1 Indications of [18F]FDG PET/CT
prognosis is generally poor.
CC is associated with risk factors such as chronic inflam- The small tumour size and the growth pattern of CC (small
mation, especially primary sclerosing cholangitis, but also nests of cells embedded in a fibrous stroma) may compro-
with cirrhosis, hepatitis B and C, and fatty liver disease. The mise the accuracy of [18F]FDG PET/CT to detect low volume
incidence of CC increases with age except those cases related deposits in regional lymph node metastases, resulting in lim-
to primary sclerosing cholangitis. ited sensitivity but high specificity (Fig. 32.5) [41]. [18F]FDG
Symptoms are those derived from bile duct obstruction PET/CT should be considered in front of equivocal findings
including jaundice, abdominal pain, cola-coloured urine, and on CT or MR, especially if a positive finding would change
skin pruritus. Symptoms are common in extrahepatic CC, management [42, 43], but its routine use in the preoperative
but intrahepatic CC and, particularly, gallbladder cancer are setting has not been yet established in prospective trials.
often diagnosed incidentally during a laparoscopic cholecys- Several studies have shown [18F]FDG PET/CT to be supe-
tectomy performed for cholelithiasis. rior to CT for detecting distant metastases [44, 45].
The staging system of CC is separated depending on its The role of [18F]FDG PET/CT in monitoring response to
location on the bile duct. Each location has a T stage desig- therapy is controversial with no ideal imaging modality yet
a b c
d e
f g
Fig. 32.5 Cholangiocarcinoma with hepatic hilar lymph node infiltra- librium phase (d), with the corresponding axial slice of fused PET/CT
tion. Maximum intensity projection [18F]FDG PET image (a) and ceCT (e) demonstrate the primary tumour. Additionally, axial slices of CT (f)
axial slices of liver arterial phase (b), portal venous phase (c), and equi- and fused PET/CT (g) show locoregional lymph node infiltration
available. In addition, more trials are needed to establish the from palliative treatments, and those with a very poor life
utility of [18F]FDG PET/CT in identifying tumour recurrence expectancy (stage D).
[40] and for radiotherapy planning.

32.7.2 [ F]FDG PET/CT Interpretation
18
and Reporting [18F]FDG PET/CT is not recommended for the detection of

HCC because of its limited sensitivity due to variable [18F]
CC is typically homogeneously low in attenuation on unen- FDG tumour uptake relative to the normal liver uptake
hanced CT and demonstrates heterogeneous minor periph- (Table 32.2), and it has been reported to be only slightly bet-
eral enhancement with gradually central enhancement on ter than CT and US and inferior to MR [48].
ceCT (Fig. 32.5). Duct ectasia with or without a visible mass While there is insufficient data to support the routine
can be seen in intraductal tumours. detection of primary HCC by [18F]FDG PET/ CT, there is
It is important to keep in mind that lymph node involve- some evidence supporting its potential for detecting extrahe-
ment at the porta does not necessarily contraindicate sur- patic disease and tumour recurrences after treatment. HCC
gical resection but metastases to more distant nodes such usually spreads to lung and bone tissue, and [18F]FDG PET/
as in the celiac, retro-pancreatic, and para-aortic chains CT can help to diagnose metastatic patients in whom treat-
do [46]. ment should not have a curative intention. Further investiga-
tions are necessary to prove favourable cost to benefit ratio
for these potential indications.
Key Learning Points

• [18F]FDG PET/CT has limited sensitivity but high 32.8.2 [18F]FDG PET/CT Interpretation
specificity in the detection of regional lymph node and Reporting
metastases in CC.
• [18F]FDG PET/CT is an emerging indication in On CT, HCC enhances during the arterial phase because its
defining recurrent disease after treatment, espe- blood supply comes from abnormal hepatic arteries unlike
cially in gallbladder cancer. the surrounding liver parenchyma, which receives blood sup-
ply mostly from the portal veins, not yet opacified. In the
portal venous phase, the surrounding liver parenchyma
becomes relatively hyperattenuated, and the lesion is per-
32.8 Hepatocellular Cancer ceived to be hypoattenuated (washout effect) [49].
HCC is the third most common cause of cancer death

[47]. HCC is associated with viral hepatitis and cirrhosis, Key Learning Point
and strategies to reduce its high mortality rely on screen- • [18F]FDG PET/CT is not recommended for detec-
ing patients with these risk factors by US followed by CT tion of HCC because of limited sensitivity, due to
or MR and liver biopsy. Surgical resection and liver variable tumour [18F]FDG uptake relative to normal
transplant remain the best options for cure, but they are liver tissue.
frequently not possible due to tumour size, multifocal
disease, extrahepatic involvement, or insufficient hepatic
reserve.
Clinical presentation may include jaundice, anorexia, 32.9 C
ontraindications of [18F]FDG PET/CT
weight loss, upper abdominal pain, hepatomegaly, and ascites. (and/or Possible Side Effects)
HCC diagnosis involves multiphasic ceMR or multiphasic
ceCT liver protocol. Different staging systems have been pro- There are no specific contraindications for patients with GI
posed. The Barcelona Clinic Liver Cancer (BCLC) system is malignancies (Table 32.6) [50].
the reference staging system in which radiologic tumour extent
is only an element of classification. The system identifies those
patients with early HCC who may benefit from curative thera- 32.10 [18F]FDG PET/CT Acquisition
pies (≤3 lesions of <3 cm, stage 0 and A), those at intermediate
(multifocal disease, stage B) or advanced stage (vascular inva- The protocol of [18F]FDG PET/CT for malignancies of the
sion, nodal or metastatic disease, stage C) who may benefit gastrointestinal tract is summarized in Table 32.6 [50].
Table 32.6 FDG PET/CT protocol
Contraindications and/or possible side effects
There are no absolute contraindications
• In pregnant females a clinical decision is necessary in which the benefits are weighed against the possible harm to the foetus by the radiation exposure
• Breastfeeding should not be interrupted since little FDG is excreted in the milk. It is recommended to breastfeed the infant just before [18F]FDG injection and feed thereafter for 12 h
using a bottle filled with expressed breast milk to avoid prolonged contact infant/mother
Patient preparation
Avoid strenuous exercise for ≥6 h before [18F]FDG injection, and preferably for 24 h
NPO except for plain (unflavoured) water for ≥4 h before [18F]FDG injection (to ensure low blood glucose and low insulinaemia, as insulin is directly responsible for glucose uptake by
non-tumour cells)
Adequate prehydration during the 2 h prior to [18F]FDG injection to ensure a sufficiently low concentration of [18F]FDG in the urine (for a correct interpretation of abdominal/pelvic tumours
by reducing artefacts) and for radiation safety reasons
Stop parenteral nutrition and IV fluids containing glucose ≥4 h before [18F]FDG injection
In well-controlled diabetics stop insulin before [18F]FDG injection:
• 8–12 h intermediate-acting and long-acting insulin
• >6 h regular or short-acting insulin
• >4 h rapid-acting insulin
In diabetics on continuous insulin infusion, stop the insulin pump ≥4 h prior to [18F]FDG administration
Weigh the patient upon arrival in the PET unit on a calibrated weighing scale
Check blood glucose (for clinical studies <200 mg/dL (11 mmol/L))
• it is possible to reduce somewhat blood glucose levels by asking the patient to hydrate while ambulating
• hyperglycaemia of patients with unstable (“brittle”) or poorly controlled diabetes is not an absolute contraindication as mild to medium fasting hyperglycaemia does not hamper the
clinical value of the examination
Consider sedation for claustrophobia or anxiety. Patient kept warm for 30–60 min before [18F]FDG injection (to reduce [18F]FDG accumulation in brown fat)
32 Hybrid Imaging for Malignant Conditions of the Gastrointestinal Tract
If IV contrast agent is to be administered, check for contrast allergy and renal function (high risk of nephrotoxicity if creatinine >13 mmol/L (1.5 mg/dL) and/or glomerular filtration <60 mL/
min)
• if the patient is allergic, premedication to reduce the risk of recurrent anaphylaxis is required (but history of severe reaction recommends eluding contrast administration)
• if renal function is suboptimal, prior hydration is necessary, and if renal dysfunction persists, consider N-acetylcysteine, use of iso-osmolar contrast agents, and stop diuretics,
nonsteroidal anti-inflammatory agents, and aminoglycosides. In diabetics controlled with metformin, stop it at the time of the examination and withhold for 48 h after the examination (or
until renal function recovery if the risk of nephrotoxicity is high)
Administer 1 L of water or water-based contrast agent before [18F]FDG injection (consider this immediately before scanning for gastric and pancreatic cancers)
Radiopharmaceutical
Standard [18F]FDG dose: 0.14 mCi/kg (5.18 MBq/kg) IV (avoid dose infiltration)
Wait 60 ± 10 min with the patient kept quiet, silent, relaxed, and warm (to reduce [18F]FDG uptake in muscles and brown fat)
Void immediately prior to the PET/CT examination (to reduce bladder activity)
Image acquisition
Torso imaging (from the base of the skull to mid-thigh) during shallow respiration, lying still (usually in the supine position) with arms above the head (to avoid beam-hardening artefacts and
artefacts caused by truncation of the measured FOV; if the camera has extended CT FOV, arms may be positioned alongside the body to enhance patient comfort provided CT truncation is
avoided.)a
(continued)
853
Table 32.6 (continued)
854
Scout scan
Transmission scan:
(1) low-dose CT (70–80 mA, 140 kV) for CT AC and anatomical correlation of PET findings, or
(2) diagnostic CT (according to applicable local or national protocols and guidelines) for an accurate radiological interpretation (and also for CT ACb):
(a) administer an additional 50–100 mL of water or water-based oral contrast agent immediately before patient positioning (except for gastric and pancreatic cancers)
(b) if IV contrast is going to be administered: 100 mL of low-osmolar iodinated contrast, using a programmable fluid injector at a speed of 2.5 mL/s for a catheter of 20 G × 1.16″ in
the elbow
• Obtain CT scan in cranial-caudal direction following injection of IV contrast
– Typically acquired in portal-venous phase occurring 60–70 s after injection. Performing the CT AC scan in this phase of the IV contrast bolus injection (or thereafter) reduces
potential artefact from IV contrast materialb
– Preceding arterial phase obtained 30–40 s after injection may be helpful with certain GI tumours such as pancreatic and cholangiocarcinoma
Emission scan: 3D acquisition in caudal-cranial direction (scan direction limits time for filling of bladder with excreted FDG and potential for obscuring pelvic lesions and changing in
bladder volume between transmission and emission scans), 2–3 min per bed position for a total of 7–8 bed positions (depending on patient size)
Online random correction based on the “delayed coincidence time window” technique or random correction using a model based on (block) single count rates
Image processing
Iterative reconstruction with decay correction “on”. Truncation algorithms may be applied
• matrix sizes and zoom factors should be chosen such that reconstructed voxel sizes are within 3–4 mm in any direction
• Time-of-flight (TOF) information should be used (when available)
• Resolution modelling may be applied
• Spatial filters should not exceed a full width at half maximum of 7 mm
Image interpretation and reporting
Image display using an SUV scale: MIP rotating PET image, together with PET, CT, and fused PET/CT three axes transaxial slices
Review both AC images and non-AC images (to correctly identify potential artefacts)
Normal distribution of [18F]FDG must be known
Visual assessment is generally performed for diagnostic purposes. Quantification of [18F]FDG uptake complements the visual impression
Correlate findings with those of other diagnostic tests, interpret them in that context, and consider them in relation to the clinical data
For response assessment, use the same dynamic grey scale or colour scale range
Describe the location, extent, and intensity (compared to background or liver, e.g. mild, moderate, intense; ±SUV) of pathological [18F]FDG accumulation
Describe relevant CT findings and their relationship to FDG uptake
If present, describe confounding factors that might influence the accuracy of PET/CT (e.g. partial volume effect in small lesions, inflammatory changes, muscle activity, GI uptake, high blood
glucose levels, dose infiltration, and artefacts)
Clearly identify the study as normal or abnormal and address the question asked in the study requisition
If possible make a diagnosis and provide a staging assessment (TNM or other) stating whether there are categories of uncertainty and differential diagnosis
Recommend to repeat the examination or additional examinations to clarify or confirm the findings
AC Attenuation correction, FOV Field of view, GI Gastrointestinal, IV Intravenous, NPO Nil per os, SUV Standard uptake value
a
If [18F]FDG PET/CT data are used for radiation planning, the examination should be performed in the same position and using the same dedicated positioning devices used for the radiotherapy.
Respiratory gating may be also required
b
Contrast-enhanced blood absorbs 40% of the photons used in CT imaging (~80 keV), but only 2% of the higher-energy photons (511 keV) inherent to PET imaging; despite this difference, IV
contrast rarely induces a diagnostically significant artefacts
J. Duch and A. Flotats
32.11 [18F]FDG PET/CT Process respectively, but some 18F-FLT-avid primary tumours were
almost undetected because they were covered by a high
Emission data must be corrected for geometrical response and background hepatic uptake. The sensitivity of 18F-FLT
detector efficiency (normalization), system dead time, random PET/CT versus [18F]FDG PET/CT for detecting liver
coincidences, scatter, and attenuation (Table 32.6) [50]. metastases and bone metastases was 30% versus 100% and
20% versus 100%, respectively. Metabolically positive
findings of lymph node, peritoneal, and ovarian metastases
32.12 [18F]FDG PET/CT Interpretation were similar between the two tracers. Therefore 18F-FLT
and Reporting does not seem competent enough to evaluate GC liver and
bone metastases; moreover, the high 18F-FLT background
Normal biodistribution of [18F]FDG must be known in order hepatic uptake may cover the gastric primary tumours
to avoid false results (Table 32.2). located adjacent to the liver [52].
[11C]Acetate PET/CT has shown a sensitivity of 87–100%
in detecting moderate to well-differentiated HCC but with
32.13 Future Perspectives high rate of false-positive results. Combination of PET/CT
with [18F]FDG and [11C]acetate has shown a sensitivity and
Future perspectives of hybrid imaging include the use of har- specificity of 93.8% and 100% for HCC, respectively [53].
monized quantitative interpretation, radiopharmaceuticals PET/CT with 18F-fluorocholine has also been explored,
other than [18F]FDG, contrast media, state-of-the-art PET/ resulting in higher sensitivity and lower specificity than [18F]
CT scanners, and PET/magnetic resonance (MR) scanners. FDG PET/CT [54].
32.13.1 Harmonized Quantitative 32.13.3 State-of-the-Art PET/CT and Contrast

Interpretation Media in PET/CT
SUV represents the most commonly used semiquantitative The present generation of PET/CT scanners introduces new
parameter for analysis of tracer uptake. Presently, changes in hardware (continuous bed motion and solid-state digital
SUV during therapy are the most robust parameters if com- photodetectors), reconstruction software (time of flight,
parable protocols have been performed (25% threshold in point spread function, Bayesian penalized likelihood,
SUVmax variation and a 30% variation in SUVpeak). smaller voxel, and metal artefact reduction), and acquisition
Additionally, SUV-related quantitative parameters, such as methods (respiratory gating and administration of quadratic
metabolic tumour volume (MTV) and total lesion glycolysis weight-dependent [18F]FDG activity). These innovations
(TLG), have achieved relevance for therapy response moni- result in better image quality, improved small lesion detect-
toring and prognostic assessment [50]. ability, and more accurate quantification of radiopharma-
The use of these quantitative parameters as imaging bio- ceutical uptake [55].
markers in multicentre trials or in sites equipped with multi- On the other hand, most of the evidence of PET/CT in
ple scanners requires that these parameters be comparable GI malignancies is based on older technology using low-
among patients, regardless of the PET/CT system used. dose unenhanced CT. The use of high-dose contrast
Accreditation programmes to harmonize PET/CT procedure enhancement PET/CT has shown to improve the assess-
have been created [51]. ment of resectability of PC [56], preoperative nodal staging
in rectal cancer [57], and recurrence of CRC [58], and
although it increases the acquisition time and radiation bur-
32.13.2 Radiopharmaceuticals Other den, it may decrease the need for additional imaging
than [18F]FDG examinations.
The real impact of these developments on diagnosis and
Due to the physiologic uptake of [18F]FDG in the GI tract, staging of GI malignancies, as well as therapy response moni-
other PET radiopharmaceuticals have been used to evaluate toring and radiotherapy planning, needs further assessment.
tumours with aspects of cellular activity other than glucose
metabolism, but the experience with these is limited, and
results are not convincing. 32.13.4 PET/MR Imaging
Evaluation of 18F-fluorothymidine (18F-FLT) in the pre-
treatment assessment of metastatic GC in comparison with Depending on the CT protocol used for a PET/CT examina-
[18F]FDG showed sensitivities of 92.3% and 94.9%, tion, patient radiation exposure can be 2–3 times the associ-
ated [18F]FDG administered dose. MR does not expose the nocarcinomas of esophagogastric junction by metabolic imaging. J
patient to ionizing radiation. In addition, MR has an improved Clin Oncol. 2001;19:3058–65.
11. Goense L, van Rossum PS, Reitsma JB, Lam MG, Meijer GJ, van
sensitivity for soft tissue, liver, and bone. It is particularly Vulpen M, et al. Diagnostic performance of 18F-FDG PET and PET/
useful in abnormalities in pelvic and abdominal organs such CT for the detection of recurrent esophageal cancer after treatment
as the liver, biliary tract, and pancreas. with curative intent: a systematic review and meta-analysis. J Nucl
PET/MR may provide an important diagnostic imaging Med. 2015;56:995–1002.
12. Lu J, Sun XD, Yang X, Tang XY, Qin Q, Zhu HC, et al. Impact
modality combining the metabolic parameters of PET with of PET/CT on radiation treatment in patients with esopha-
the superior soft tissue imaging quality of MR [59], all along geal cancer: a systematic review. Crit Rev Oncol Hematol.
with a significant reduction (40–60%) of ionizing radiation 2016;107:128–37.
(significant for patients who require multiple scans and chil- 13. GLOBOCAN 2012 website. globocan.iarc.fr. Accessed 25 July
2017.
dren). However, specific MR sequences have to be chosen in 14. Alakus H, Batur M, Schmidt M, Drebber U, Baldus SE, Vallböhmer
advance so as not to lengthen the study disproportionately. D, et al. Variable 18F-fluorodeoxyglucose uptake in gastric cancer is
Standardization of PET/MR protocols is not yet accom- associated with different levels of GLUT-1 expression. Nucl Med
plished, and the choice of MR sequences also depends on the Commun. 2010;31:532–8.
15. Yamada A, Oguchi K, Fukushima M, Imai Y, Kadoya M. Evaluation
equipment architecture (simultaneous versus sequential). of 2-deoxy-2-[18F]fluoro-d-glucose positron emission tomography
Additionally, PET/MR technology is not yet widely avail- in gastric carcinoma: relation to histological subtypes, depth of
able and is more expensive than PET/CT in terms of machin- tumour invasion, and glucose transporter-1 expression. Ann Nucl
ery, maintenance, and operation. Med. 2006;20:597–604.
16. Mukai K, Ishida Y, Okajima K, Isozaki H, Morimoto T, Nishiyama
Since PET/MR is an emerging technology, no large pro- S. Usefulness of preoperative FDG-PET for detection of gastric
spective studies validating [18F]FDG PET/MR in oncologic cancer. Gastric Cancer. 2006;9:192–6.
patients are currently available. However, there is emerging 17. Smyth EC, Verheij M, Allum W, Cunningham D, Cervantes A,
data in its potential role in diagnosis and treatment of GI Arnold D, ESMO Guidelines Committee. Gastric cancer: ESMO
Clinical Practice Guidelines for diagnosis, treatment and follow-up.
malignancies [60]. Ann Oncol. 2016;27(Suppl 5):v38–49.
18. NCCN Clinical Practice Guidelines in Oncology. Gastric Cancer
Version 1.2017 – March 21, 2017. https://www.nccn.org/profes-
sionals/physician_gls/pdf/gastric.pdf.
References 19. Vallböhmer D, Hölscher AH, Schneider PM, Schmidt M, Dietlein
M, Bollschweiler E, et al. [18F]-fluorodeoxyglucose-positron emis-
1. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, sion tomography for the assessment of histopathologic response
Coebergh JW, Comber H, et al. Cancer incidence and mortality pat- and prognosis after completion of neoadjuvant chemotherapy in
terns in Europe: estimates for 40 countries in 2012. Eur J Cancer. gastric cancer. J Surg Oncol. 2010;102:135–40.
2013;49:1374–403. 20. Jadvar H, Tatlidil R, Garcia AA, Conti PS. Evaluation of recurrent
2. Key statistics for esophageal cancer website. cancer.org/cancer/ gastric malignancy with [F-18]-FDG positron emission tomog-
esophagus-cancer. raphy. Clin Radiol. 2003;58:215–21. Erratum in: Clin Radiol.
3. Huang FL, Yu SJ. Esophageal cancer: risk factors, genetic associa- 2003;58:570.
tion, and treatment. Asian J Surg 2016. pii: S1015-9584(16)30201-9. 21. Wu LM, Hu JN, Hua J, Gu HY, Zhu J, Xu JR. 18F-fluorodeoxyglucose
https://doi.org/10.1016/j.asjsur.2016.10.005. [Epub ahead of print]. positron emission tomography to evaluate recurrent gastric cancer:
4. Hong SJ, Kim TJ, Nam KB, Lee IS, Yang HC, Cho S, et al. New a systematic review and meta-analysis. J Gastroenterol Hepatol.
TNM staging system for esophageal cancer: what chest radiologists 2012;27:472–80.
need to know. Radiographics. 2014;34:1722–40. 22. Kwee RM, Kwee TC. Imaging in assessing lymph node status in
5. Gauthé M, Richard-Molard M, Cacheux W, Michel P, Jouve JL, gastric cancer. Gastric Cancer. 2009;12:6–22.
Mitry E, et al. Role of fluorine 18 fluorodeoxyglucose positron 23. Okano K, Suzuki Y. Strategies for early detection of resectable pan-
emission tomography/computed tomography in gastrointestinal creatic cancer. World J Gastroenterol. 2014;20:11230–40.
cancers. Dig Liver Dis. 2015;47:443–54. 24. Lahat G, Ben Haim M, Nachmany I, Sever R, Blachar A, Nakache
6. Meltzer CC, Luketich JD, Friedman D, Charron M, Strollo D, R, et al. Pancreatic incidentalomas: high rate of potentially malig-
Meehan M, et al. Whole-body FDG positron emission tomographic nant tumors. J Am Coll Surg. 2009;209:313–9.
imaging for staging esophageal cancer comparison with computed 25. Dibble E, Karantanis D, Mercier G, Peller PJ, Kachnic LA,

tomography. Clin Nucl Med. 2000;25:882–7. Subramaniam RM. PET/CT of cancer patients: Part 1, Pancreatic
7. Erasmus JJ, Munden RF. The role of integrated computed tomography neoplasms. AJR Am J Roentgenol. 2012;199:952–67.
positron-emission tomography in esophageal cancer: staging and assess- 26. Cameron K, Golan S, Simpson W, Peti S, Roayaie S, Labow D,
ment of therapeutic response. Semin Radiat Oncol. 2007;17:29–37. et al. Recurrent pancreatic carcinoma and cholangiocarcinoma:
8. Luo LN, He LJ, Gao XY, Huang XX, Shan HB, Luo GY, et al. 18
F-fluorodeoxyglucose positron emission tomography/computed
Evaluation of preoperative staging for esophageal squamous cell tomography (PET/CT). Abdom Imaging. 2011;36:463–71.
carcinoma. World J Gastroenterol. 2016;22:6683–9. 27. NCCN Clinical Practice Guidelines in Oncology. Pancreatic

9. Van Westreenen HL, Westerterp M, Bossuyt PM, Pruim J, Sloof Adenocarcinoma. Version 2.2017 – April 27, 2017. https://www.
GW, van Lanschot JJ, et al. Systematic review of the staging per- nccn.org/professionals/physician_gls/pdf/pancreatic.pdf.
formance of 18F-fluorodeoxyglucose positron emission tomography 28. Farma JM, Santillan AA, Melis M, Walters J, Belinc D, Chen DT,
in esophageal cancer. J Clin Oncol. 2004;22:3805–12. et al. PET/CT fusion scan enhances CT staging in patients with
10. Weber WA, Ott K, Becker K, Dittler HJ, Helmberger H, Avril NE, pancreatic neoplasms. Ann Surg Oncol. 2008;15:2465–71.
et al. Prediction of response to preoperative chemotherapy in ade-
29. Lee ES, Lee JM. Imaging diagnosis of pancreatic cancer: a state-of- operable cholangiocarcinoma: a prospective study compared with
the-art review. World J Gastroenterol. 2014;20:7864–77. conventional imaging. Am J Gastroenterol. 2008;103:1145–51.
30. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo 46. Gourgiotis S, Kocher HM, Solaini L, Yarollahi A, Tsiambas E,
M, et al. Cancer incidence and mortality worldwide: sources, Salemis NS. Gallbladder cancer. Am J Surg. 2008;196:252–64.
methods and major patterns in GLOBOCAN 2012. Int J Cancer. 47. World Health Organization. www.who.int.
2015;136:E359–86. 48. Böhm B, Voth M, Geoghegan J, Hellfritzsch H, Petrovich A,

31. Kirkegaard H, Johnsen NF, Christensen J, Frederiksen K, Overvad Scheele J, et al. Impact of positron emission tomography on strat-
K, Tjønneland A. Association of adherence to lifestyle recommen- egy in liver resection for primary and secondary liver tumors. J
dations and risk of colorectal cancer: a prospective Danish cohort Cancer Res Clin Oncol. 2004;130:266–72.
study. BMJ. 2010;341:c5504. 49. McEvoy SH, McCarthy CJ, Lavelle LP, Moran DE, Cantwell CP,
32. Van Cutsem E, Cervantes A, Nordlinger B, Arnold D, ESMO
Skehan SJ, et al. Hepatocellular carcinoma: illustrated guide to
Guidelines Working Group. Metastatic colorectal cancer: ESMO systematic radiologic diagnosis and staging according to guide-
Clinical Practice Guidelines for diagnosis, treatment and follow- lines of the American Association for the Study of Liver Diseases.
up. Ann Oncol. 2014;25(Suppl 3):iii1–9. https://doi.org/10.1093/ Radiographics. 2013;33:1653–68. Erratum in: Radiographics
annonc/mdu260. Erratum in: Ann Oncol 2015;26 Suppl 5:v174–7. 2014;34:15A. Dosage error in article text.
33. Niekel MC, Bipat S, Stoker J. Diagnostic imaging of colorectal liver 50. Boellaard R, Delgado-Bolton R, Oyen WJG, Giammarile F, Tatsch
metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: K, Eschner W, et al. FDG PET/CT: EANM procedure guidelines
a meta-analysis of prospective studies including patients who have for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging.
not previously undergone treatment. Radiology. 2010;257:674–84. 2015;42:328–54.
34. Hendlisz A, Golfinopoulos V, Garcia C, Covas A, Emonts P, Ameye 51. Aide N, Lasnon C, Veit-Haibach P, Sera T, Sattler B, Boellaard
L, et al. Serial FDG-PET/CT for early outcome prediction in R. EANM/EARL harmonization strategies in PET quantification:
patients with metastatic colorectal cancer undergoing chemother- from daily practice to multicentre oncological studies. Eur J Nucl
apy. Ann Oncol. 2012;23:1687–93. Med Mol Imaging. 2017;44(Suppl 1):17–31.
35. Petersen RK, Hess S, Alavi A, Hoilund-Carlsen PF. Clinical impact 52. Zhou M, Wang C, Hu S, Yao Z, Li J, Guo W, Zhang Y. 18F-FLT
of FDG-PET/CT on colorectal cancer staging and treatment strat- PET/CT imaging is not competent for the pretreatment evaluation
egy. Am J Nucl Med Mol Imaging. 2014;4:471–82. of metastatic gastric cancer: a comparison with 18F-FDG PET/CT
36. Maffione AM, Lopci E, Bluemel C, Giammarile F, Herrmann K, imaging. Nucl Med Commun. 2013;34:694–700.
Rubello D. Diagnostic accuracy and impact on management of 53. Cheung TT, Ho CL, Lo CM, Chen S, Chan SC, Chok KS, et al.
18
F-FDG PET and PET/CT in colorectal liver metastasis: a meta- 11
C-acetate and 18FDG PET/CT for clinical staging and selection
analysis and systematic review. Eur J Nucl Med Mol Imaging. of patients with hepatocellular carcinoma for liver transplantation
2015;42:152–63. on the basis of Milan criteria: surgeons perspective. J Nucl Med.
37. van Stiphout RG, Lammering G, Buijsen J, Janssen MH,
2013;54:192–200.
Gambacorta MA, Slagmolen P, et al. Development and external 54. Talbot JN, Fartoux L, Balogova S, Nataf V, Kerrou K, Gutman
validation of a predictive model for pathological complete response F, et al. Detection of hepatocellular carcinoma with PET/CT:
of rectal cancer patients including sequential PET-CT imaging. a prospective comparison of 18F-fluorocholine and 18F-FDG
Radiother Oncol. 2011;98:126–33. in patients with cirrhosis or chronic liver disease. J Nucl Med.
38. Elferink MA, Visser O, Wiggers T, Otter R, Tollenaar RA,
2010;51:1699–706.
Langendijk JA, Siesling S. Prognostic factors for locoregional 55. van der Vos CS, Koopman D, Rijnsdorp S, Arends AJ, Boellaard R,
recurrences in colon cancer. Ann Surg Oncol. 2012;19:2203–11. van Dalen JA, et al. Quantification, improvement, and harmoniza-
39. Jemal A, Simard EP, Dorell C, Noone AM, Markowitz LE, Kohler tion of small lesion detection with state-of-the-art PET. Eur J Nucl
B, et al. Annual Report to the Nation on the Status of Cancer, 1975– Med Mol Imaging. 2017;44(Suppl 1):4–16.
2009, featuring the burden and trends in human papillomavirus 56. Strobel K, Heinrich S, Bhure U, Soyka J, Veit-Haibach P, Pestalozzi
(HPV)-associated cancers and HPV vaccination coverage levels. J BC, et al. Contrast-enhanced 18F-FDG PET/CT: 1-stop-shop imag-
Natl Cancer Inst. 2013;105:175–201. ing for assessing the resectability of pancreatic cancer. J Nucl Med.
40. de Groen PC, Gores GJ, LaRusso NF, Gunderson LL, Nagorney 2008;49:1408–13.
DM. Biliary tract cancers. N Engl J Med. 1999;341:1368–78. 57. Tateishi U, Maeda T, Morimoto T, Miyake M, Arai Y, Kim

41. Butte JM, Redondo F, Waugh E, Meneses M, Pruzzo R, Parada H, EE. Non-enhanced CT versus contrast-enhanced CT in integrated
et al. The role of PET-CT in patients with incidental gallbladder PET/CT studies for nodal staging of rectal cancer. Eur J Nucl Med
cancer. HBP (Oxford). 2009;11:585–91. Mol Imaging. 2007;34:1627–34.
42. Srinivasa S, McEntee B, Koea JB. The role of PET scans in the 58. Kitajima K, Murakami K, Yamasaki E, Domeki Y, Tsubaki M,
management of cholangiocarcinoma and gallbladder cancer: a sys- Sunagawa M, et al. Performance of integrated FDG PET/contrast-
tematic review for surgeons. Int J Diagnos Imaging. 2015;2:1–9. enhanced CT in the diagnosis of recurrent colorectal cancer: com-
43. Corvera CU, Blumgart LH, Akhurst T, DeMatteo RP, D’Angelica parison with integrated FDG PET/non-contrast-enhanced CT and
M, Fong Y, et al. 18F-fluorodeoxyglucose positron emission tomog- enhanced CT. Eur J Nucl Med Mol Imaging. 2009;36:1388–96.
raphy influences management decisions in patients with biliary 59. Torigian DA, Zaidi H, Kwee TC, Saboury B, Udupa JK, Cho ZH,
cancer. J Am Coll Surg. 2008;206:57–65. et al. PET/MR imaging: technical aspects and potential clinical
44. Lan BY, Kwee SA, Wong LL. Positron emission tomography in applications. Radiology. 2013;267:26–44.
hepatobiliary and pancreatic malignancies: a review. Am J Surg. 60. Matthews R, Choi M. Clinical utility of positron emission tomog-
2012;204:232–41. raphy magnetic resonance imaging (PET-MRI) in gastrointestinal
45. Kim JY, Kim MH, Lee TY, Hwang CY, Kim JS, Yun SC, et al. cancers. Diagnostics (Basel). 2016;6(3):E35.
Clinical role of 18F-FDG PET-CT in suspected and potentially
Radionuclide Therapy for Tumors
of the Liver and Biliary Tract 33
Federica Guidoccio, Giuseppe Boni, Duccio Volterrani,
Contents
33.2 Rationale of Radioembolization Therapy for Liver Tumors 860
33.3 Radioactive Lipiodol 860
33.4 Commercially Available Radiolabeled Microspheres for TARE/SIRT 861
33.4.1 90
Y-Microspheres 861
33.4.2 Ho-Microspheres
166
861
33.5 Patient Selection 862
33.5.1 Assessment of Arterial Anatomy 862
33.5.2 Pretreatment Imaging with 99mTc-MAA 862
33.6 Radiodosimetric Aspects 865
33.7 Y-Microsphere Administration
90
867
33.8 Posttreatment Scan 868
33.9 Patient Follow-Up and Assessment of Response to Therapy 869
33.10 Y-Microsphere Radioembolization Combined with Other Therapies
90
871
33.11 linical Indications in Primary Liver Tumors
C 873
33.11.1 Hepatocellular Carcinoma 873
33.11.2 Intrahepatic Cholangiocarcinoma 873
33.12 linical Indications in Metastatic Liver Tumors
C 874
33.12.1 Metastatic Colorectal Carcinoma 874
33.12.2 Metastatic Neuroendocrine Tumors 874
33.13 Absolute and Relative Contraindications 874
33.14 Early and Late Toxicities 874
33.15 Perspectives on Radioembolization for Primary and Metastatic Liver Tumors 875
33.16 Ho-PLLA-Microspheres
166
875
References 875
Learning Objectives
F. Guidoccio (*) · D. Volterrani · G. Mariani • Acquire the basic notions on various types of radionu-
clide therapy for tumors of the liver and biliary tract.
University of Pisa, Pisa, Italy • Learn advantages and disadvantages of radioemboliza-
tion therapy for tumors of the liver and biliary tract.
G. Boni
Regional Center of Nuclear Medicine, • Learn the properties of radioactive preparations used
University Hospital of Pisa, Pisa, Italy for radioembolization therapy and the rationale to their

use in different clinical settings (hepatocellular carci-

noma, intrahepatic cholangiocarcinoma, metastatic Key Learning Points
cholangiocarcinoma, metastatic neuroendocrine • The liver can be the site of both primary and meta-
tumors). static malignancies.
• Learn the most common protocols for radioemboliza- • Curative resection or liver transplantation results in
tion therapy (pretreatment imaging with 99mTc-MAA, significant survival benefit in selected patients with
90
Y/166Ho-microsphere administration, posttreatment hepatocellular carcinoma.
PET/TC scan). • Radiofrequency ablation, microwave ablation,
• Understand the principles of interpretation for radio- cryoablation, and irreversible electroporation are
nuclide imaging associated with radioembolization potentially curative therapies.
therapy of liver tumors. • The efficacy of trans-arterial chemoembolization is
proved only for palliative treatments of primary and
metastatic liver cancers.
33.1 Introduction • Trans-arterial radioembolization (TARE) with 90Y-
or 166Ho-labeled particles constitutes a novel
The liver can be the site of both primary and metastatic approach for locoregional therapy of liver tumors.
malignancies. Besides hepatocellular carcinoma (HCC,
ranking fifth among the most common malignancies world-
wide) [1], different solid tumors primarily arising in other
tissues/organs, most importantly colorectal cancer (CRC), 33.2 Rationale of Radioembolization
frequently give rise to metastases in the liver [2]. Therapy for Liver Tumors
Curative resection or liver transplantation results in sig-
nificant survival benefit in selected patients with HCC [3], While blood to normal hepatic tissue is supplied for >70%
although <15% of the newly diagnosed HCC cases can be by the portal system, blood to malignant tissue is supplied
treated with curative intents. mostly by the arterial system. Radiation doses to the tumor
Radiation therapy has a limited role in the treatment of lesions >70 Gy (up to 200–300 Gy) can thus be selectively
HCC, mostly because of the risk of inducing “radiation- delivered through locoregional administration of agents
induced liver disease” (RILD) or radiation hepatitis for doses emitting β− particles (labeled with either 131I, 90Y, 188Re, or
greater than 40 Gy to normal hepatic tissue [4, 5]. 166
Ho), while the non-affected liver (where blood supply
Radiofrequency ablation (RFA), microwave ablation, derives from the portal vein) receives low radiation levels,
cryoablation, and irreversible electroporation (IRE) are with minimal risk of inducing RILD [13–17].
potentially curative therapies for either HCC or metastatic
liver disease [6], in particular for patients who do not meet
the criteria for surgery. 33.3 Radioactive Lipiodol
The use of transcatheter arterial chemoembolization
(TACE), a catheter-based locoregional therapy for unresect- I-Lipiodol (a suspension of lipidic particles) injected into
131
able primary liver cancer, is expanding to include also liver the hepatic artery of patients with HCC follows the prefer-
metastases from other cancers [7]. TACE is based on the ential blood flow toward the tumor, where the radiolabeled
injection of chemotherapeutic agents mixed with lipiodol micellae are retained by pinocytosis both in the tumor cells
and embolic particles selectively into the branch of the and in the endothelium arterial vessels feeding the tumor [18].
hepatic artery that feeds the tumor [8]. Nevertheless, the effi- This route of administration is performed under angiographic
cacy of TACE (including the use of drug-eluting beads that monitoring and results in >75% of the injected 131I-lipiodol
provide sustained and controlled drug release over time) is being retained in the liver, the remainder distributing mainly
proved only for palliative treatments of primary and meta- to the lungs. Thyroidal uptake of free radioiodine released
static liver cancers [9–12]. after treatment must be prevented with adequate medication.
Trans-arterial radioembolization (TARE) with Either a fixed activity of 131I-lipiodol (2.4 GBq, or 65 mCi)
90
Y-labeled particles has more recently been introduced for can be administered or a patient-specific dosimetry estimate
locoregional treatment of liver tumors, thus adding a new can guide definition of the activity to be injected. Tumor/
tool for locoregional therapy with radionuclides in addition non-tumor ratios >5 are generally achieved, with >10% of
to the earlier form of treatment based on the use of lipiodol the injected radioactivity remaining in the tumor with an
containing 131I. The use of non-reabsorbable 90Y-labeled effective half-life >6 days [14, 15, 19]. Because of such long
particles has also been defined as “selective internal radia- half-life of 131I-lipiodol in the tumor, legislation regarding
tion therapy” (SIRT); thus, the terms TARE and SIRT are radioprotection of the general population requires in some
interchangeable. countries hospitalization for about 1 week.
33 Radionuclide Therapy for Tumors of the Liver and Biliary Tract 861
There are usually no serious adverse side effects caused 33.4 C

ommercially Available Radiolabeled
by treatment with 131I-lipiodol, interstitial pneumopathy due Microspheres for TARE/SIRT
to trapping and retention of the radiolabeled particle suspen-
sion being reported as the main risk of such therapy [20]. 33.4.1 90
Y-Microspheres
The objective response rates of trans-arterial therapy with
131
I-lipiodol are reported in 40–50% range [21–25]. In HCC TARE/SIRT with 90Y-microspheres has shown definite clini-
patients with portal thrombosis, 71% survival was reported cal benefits in patients with inoperable primary or metastatic
at 3 months versus 10% with the best supportive care; the liver tumors [31]. The pure β− emitter 90Y decays to stable
median overall survivals were 26 weeks and 10 weeks, 90
Zr with a physical half-life of 64.2 h. The average energy
respectively [24]. The benefits of treatment with 131I-lipiodol of β− emission is 0.936 MeV, with a mean tissue penetration
have been proved in several respects, including median time of 2.5 mm and a maximum tissue range of 10 mm.
to recurrence; overall survival at 3 and 5 years; recurrence The 90Y-microspheres commercially available are made
rate; 5-, 7-, and 10-year disease-free survival; and overall of glass (TheraSphere®, MDS Nordion, Ottawa, Ontario,
survival [26, 27]. Compared with TACE, embolization Canada) or of resin (SIR-Spheres®, Sirtex Medical, Sydney,
with131I-lipiodol ensued similar therapeutic efficacy associ- Australia), respectively [32]. The TheraSphere® particles are
ated however with better tolerance [23]. 20–30 μm in diameter and carry 2500 Bq of 90Y per particle
An alternative approach to therapy with 131I-lipiodol is con- (high specific activity); about 1.2 million microspheres are
stituted by lipiodol labeled with 188Re using 4-hexadecyl injected intra-arterially for a single treatment, corresponding
2,2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol (HDD) as the to a total administered activity of about 3 GBq (81 mCi). The
chelating agent [28] in patients with inoperable large and/or SIR-Spheres® particles are 20–60 μm in diameter and carry-
multifocal HCCs. 188Re has a shorter half-life than 131I (16.9 h ing 50 Bq of 90Y per particle (low specific activity); 40–80
versus 8 days), emission of β− particles with high energy million microspheres are injected for a single treatment to
(Emax = 2.1 MeV), and emission of 155-keV γ-rays suitable for achieve a similar total administered activity of 3 GBq [32].
gamma camera imaging (for the purpose of dosimetry esti-
mates). 188Re can be obtained through the 188W/188Re generator,
the parent radionuclide having a physical half-life of 69 days, 33.4.2 166
Ho-Microspheres
which is particularly suitable for distribution logistics.
188
Re-HDD-lipiodol represents an excellent alternative to This recently approved formulation known as
131
I-lipiodol, because of higher tumor-killing efficacy QuiremSpheres® (distributed in Europe by Quirem Medical
combined with lower toxicity [29]. In patients with inopera- BV, Deventer, The Netherlands) consists of a poly(l-
ble HCC, therapy with 188Re-HDD-lipiodol resulted in favor- lactic acid) acid matrix containing 166Ho (166Ho-PLLA-
able safety profile and good clinical response, as shown also microspheres); this radionuclide emits β− particles with two
by objective responses [30]. main energy peaks (1.74 MeV and 1.85 MeV, with intensity
48.7% and 50.0%, respectively) and decays to stable 166Er
with a physical half-life of 26.8 h. The maximum range in
Key Learning Points tissues of the β− particles emitted by 166Ho is 8.7 mm, with
• While blood to normal hepatic parenchyma is sup- an average range of 2.5 mm.
plied for >70% by the portal system, blood to A standard administered therapeutic activity of
malignant tissue is supplied mostly by the arterial 166
Ho-PLLA-microspheres contains about 30 million micro-
system. spheres having a mean diameter of 30 μm (with 97% of the
• High radiation doses to the tumor lesions can there- particles being included in the 15–60 μm size range).
fore be selectively delivered through intra-arterial In addition to its predominant decay with emission of β−
administration of radioactive agents emitting particles, 166Ho also emits low-energy γ-rays (81 keV), which
β− particles. allow for scintigraphic visualization posttreatment. This fea-
• At the same time, following this administration ture is very useful for the following three main reasons:
route, the non-affected liver (where blood supply
derives from the portal vein) receives low radiation • Prior to actual TARE/SIRT therapy, a small scout activity of
levels—therefore with minimal risk of causing 166
Ho-PLLA-MS can be administered for predicting distri-
radiation-induced liver disease. bution of the treatment dose, although the manufacturer rec-
• 131I-Lipiodol and 188Re-HDD-lipiodol (suspensions ommends use of a pretreatment scout scintigraphy with
of lipidic particles carrying 131I or 188Re, respec- 99m
Tc-MAA—similarly as recommended for the
tively) have been used in patients with inoperable 90
Y-microspheres; thus, a potential advantage over
hepatocellular carcinoma. 90
Y-microspheres is not yet fully exploited in the clinical
routine.
• Quantitative analysis of the data based on the pre-therapy or bilobar. Ascites indicates poor hepatic reserve or perito-
scout scan with 166Ho-microspheres enable to assess/pre- neal metastasis, therefore poor prognosis.
dict the radiation dose delivered both to the tumor For final decision on treatment with 90Y-microsphere
lesion(s) and to normal liver parenchyma. TARE/SIRT, all functional and anatomic parameters should
• Trivalent holmium is the element with the highest para- be carefully considered within a multidisciplinary team
magnetic properties among all known elements (106 μB); involving interventional radiologists, surgical oncologists,
therefore, it can in principle be visualized also using MR medical oncologists, nuclear medicine physicians, radiation
imaging, with potential clinical applications that are cur- oncologists, medical physicists, and radiation safety experts.
rently being explored.
33.5.1 Assessment of Arterial Anatomy

Key Learning Points
• Three types of radiolabeled microspheres for radio- As an essential requisite for the therapeutic procedure, pre-
embolization therapy are commercially available: treatment angiography serves to assess the vascular anatomy
90
Y-microspheres made of glass, 90Y-microspheres of the liver, patency of the portal vein, and presence of artero-
made of resin, and 166Ho-microspheres made of a portal shunting and/or shunting to extrahepatic territories
poly(l-lactic acid) matrix, respectively. (the most important being the liver-to-lung shunt). Blood
• There are small variations in size of the microspheres flow carrying the radiolabeled microspheres outside the liver
among the three preparations, all of them allowing vasculature is prevented by prophylactic embolization of the
efficient embolization of the precapillary vessels. responsible vessels identified during angiography, such as the
• Each type of preparation has specific features con- gastroduodenal artery and the right gastric artery [5]. This safe
cerning the amount of radioactivity carried by each and effective procedure minimizes the risks of hepato-enteric
microsphere and the number of microspheres flow. Mesenteric angiography ensures that blood supply to
administered per treatment. the lesions has been adequately identified, since incomplete/
• Whereas 90Y is an almost pure β− emitter (there is in inaccurate definition of the pattern of blood supply may lead
fact an extremely small proportion of decay through to incomplete/ineffective targeting of the tumor lesion.
internal pair production, with 32 events per million
decays), 166Ho also emits low-energy γ-rays that
allow for scintigraphic visualization with a standard 33.5.2 Pretreatment Imaging with 99mTc-MAA
gamma camera.
• Holmium has very high paramagnetic properties; During pretreatment angiography, 99mTc-macroaggregated
the potential clinical applications linked to this fea- serum albumin (99mTc-MAA), or alternatively 99mTc-HSA-
ture are being explored. microspheres, is injected into the hepatic artery to confirm
homing of the radiolabeled particles in the tumor lesion(s)
and to assess for the presence of shunting to extrahepatic
circulation, either in the splanchnic bed or in pulmonary
33.5 Patient Selection bed. Scintigraphy of the chest and upper abdomen by either
planar or SPECT/CT imaging (the optimal imaging tech-
Before treatment, functional status of the liver is evaluated nique) is performed for this purpose (Figs. 33.1, 33.2, and
by routine blood chemistry workup. Patients with an ECOG 33.3) [33]. The optimal time window for imaging after
performance status >2 are not considered ideal candidates 99mTc- MAA injection is within 1 h post-administration,
for TARE/SIRT. Treatment with 90Y-microspheres is contra- since intrahepatic degradation of MAA is relatively fast
indicated by inadequate liver function, with total bilirubin and radioactivity (smaller fragments and/or free 99mTcO4−)
>2.0 mg/dL and serum albumin <3 g/dL. In patients with can redistribute from the capillary bed of the liver to the
concomitant renal failure, the use of iodinated contrast capillary bed of the lung. Therefore, the liver-to-lung
medium during angiography must be minimized. shunt fraction can be overestimated at later time points
Before TARE/SIRT with 90Y-microspheres, cross- after administration [34], or free 99mTcO4− can accumu-
sectional imaging and arteriograms are performed in each late in the stomach and can be misinterpreted as shunting
patient, after ascertaining that the patient has liver-dominant to extrahepatic sites. Sodium perchlorate is administered
unresectable disease. Pretreatment workup includes CT or orally about 30 min before 99mTc-MAA injection in order
MR imaging of the liver for assessing tumoral and non- to avoid such occurrence [35]. Instead, images can be
tumoral volume, portal vein patency, and extent of extrahe- acquired within 4 h after the administration of the more
patic disease. Tumor disease is typically defined as unilobar stable 99mTc-HSA-microspheres.
Planer Anterior Axial SPECT/CT
Fig. 33.1 Pretreatment angiographic and scintigraphic evaluation of with minimal perfusion of the main site of the tumor. Based on the
patient with a large infiltrating HCC. (a) Left panel: contrast-enhanced pretreatment angiographic and scintigraphic evaluation, this patient
CT shows massive infiltration of segment I by a large HCC, extending was excluded from trans-arterial radioembolization with
to the adjacent liver segments. Right panel shows early-phase digital 90
Y-microspheres, because treatment would have exposed the non-
subtraction angiography through the hepatic artery (catheter indicated tumor liver parenchyma to excessive, unjustified radiation doses with-
by black arrow): intense diffuse enhancement of the left branches of the out actually targeting the tumor lesion (reproduced with permission
portal vein, with an artero-portal fistula (white arrows). (b) Left panel: from: Boni G, Guidoccio F, Volterrani D, Mariani G. Radionuclide
planar scintigraphy (anterior view) demonstrates widespread diffusion therapy of tumor of the liver and biliary tract. In: Strauss HW, Mariani
of the injected 99mTc-MAA to virtually the entire liver. Right panel: the G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
fused axial SPECT/CT image better defines the intrahepatic distribu- Pathophysiology to Clinical Applications. New York: Springer;
tion of 99mTc-MAAs, mostly to both the right and the left liver lobes, 2017:1337–60)
The liver-to-lung shunt fraction (LSF) is calculated based adjusting the administered activity in order to minimize the
on ROI analysis to estimate the radiation dose delivered to risk of radiation pneumonitis (see further below).
the lungs for any given amount of radioactivity, thus properly
Key Learning Points • Prior to actual radioembolization therapy, 99mTc-MAA

• Patients are selected as potential candidates to radio- is injected into the hepatic artery to confirm homing of
embolization therapy on the basis of clinical parame- the radiolabeled particles in the tumor lesion(s) and to
ters and cross-sectional imaging including CT or MR assess for the presence of shunting to extrahepatic cir-
imaging. culation, either in the splanchnic bed or in pulmonary
• Final selection of patients is based on consensus reached bed.
within a multidisciplinary team involving interventional • Planar and/or SPECT/CT imaging serves to estimate
radiologists, surgical oncologists, medical oncologists, the liver-to-lung shunt fraction, which is generally
nuclear medicine physicians, radiation oncologists, calculated based on ROI analysis based on planar
medical physicists, and radiation safety experts. imaging.
Fig. 33.2 Patient with metastasis from an urothelial carcinoma in both the falciform artery induced vasoconstriction, thus greatly reducing
lobes of the liver; selective angiography performed separately into the inadvertent 90Y-microsphere deposition in the periumbilical region and
right, the median, and the left branches of the hepatic artery. (a) Left preventing delivery of a high radiation dose to this region. The proce-
panel: digital subtraction angiography of the median hepatic artery dure was uneventful, and the posttreatment PET/CT scan acquired on
reveals a thin branch (white arrow) identified as a patent falciform the basis on internal pair production of 90Y (c) showed excellent tumor
artery. Right panel: the corresponding contrast-enhanced CT phase targeting without visualization of the area fed by the patent falciform
shows the falciform artery (white arrow). (b) Pretreatment 99mTc-MAA artery (fused coronal image in left panel, fused sagittal image in right
scintigraphy clearly shows the abnormal arterial branch (arrows) in the panel) (reproduced with permission from: Boni G, Guidoccio F,
planar anterior view (upper left panel) as well as in the fused SPECT/ Volterrani D, Mariani G. Radionuclide therapy of tumor of the liver and
CT images (coronal in upper right, sagittal in lower left, axial in lower biliary tract. In: Strauss HW, Mariani G, Volterrani D, Larson SM, Eds.
right panels, respectively). During administration of therapeutic Nuclear Oncology – From Pathophysiology to Clinical Applications.
90
Y-microspheres, an ice pack positioned on the cutaneous projection of New York: Springer; 2017:1337–60)
90Y-PET/CT
Fig. 33.3 Pretreatment 99mTc-MAA scintigraphy (fused axial SPECT/ terns, further attempts to 90Y-microsphere TARE/SIRT can only be
CT images at various levels) of patient with tumor between segments V undertaken after performing coil embolization of the arteries feeding
and VI of the liver demonstrates satisfactory distribution to tumor extrahepatic territories (reproduced with permission from: Boni G,
lesion, but also visualization of the gallbladder wall, due to blood flow Guidoccio F, Volterrani D, Mariani G. Radionuclide therapy of tumor
through the cholecystic artery. The patient was excluded from therapy of the liver and biliary tract. In: Strauss HW, Mariani G, Volterrani D,
with 90Y-microsperes, because of the high risk of causing necrotizing Larson SM, Eds. Nuclear Oncology – From Pathophysiology to Clinical
radiation-induced cholecystitis. In patients with abnormal vascular pat- Applications. New York: Springer; 2017:1337–60)
33.6 Radiodosimetric Aspects Flow/reflux of part of the 90Y-microspheres to arteries feed-

ing the gastrointestinal tract and excessive lung irradiation due
The main adverse side effects of TARE/SIRT with to a high hepato-pulmonary shunt (as frequently observed in
90
Y-microspheres are due to excessive irradiation of non- patients with large tumor burden) constitute additional absolute
target tissue. The maximum acceptable cumulative radia- and relative contraindications for the procedure.
tion dose to the whole liver is 35 Gy, as estimated by prior Quantification of the LSF, as determined during the pre-
experience with external beam radiation therapy [36], while treatment 99mTc-MAA scan, is crucial to calculate the
the radiation dose to kill a solid tumor is >70 Gy. The risk expected radiation dose to the lung, therefore to warrant
of liver failure rises sharply for radiation doses to the liver safety of the procedure. Based on data extrapolated from
parenchyma >35 Gy. prior data obtained with external beam radiation therapy, the
highest tolerable dose to the lungs is 30 Gy for a single while sparing normal liver tissue. The MIRD-based partition
administration and less than 50 Gy as the cumulative dose model considers three different compartments for radiation
for multiple treatments [36]. dose estimates: the tumor, the non-tumoral liver, and the
The geometric mean of the lung and liver counts in the lungs [37]. The partition model assumes that the LSF and
anterior and posterior views, respectively, is used to calculate distribution of 99mTc-MAAs in the tumor and non-tumor liver
the LSF by ROI analysis performed on the planar scans compartments (expressed as T/N ratio) reliably predict dis-
acquired after 99mTc-MAA administration, according to the tribution of the 90Y-microspheres that are then administered
following equation: during another interventional radiology procedure performed
a few days later by the same radiologist, who tries to repli-
Lung counts
LSF ( % ) = ×100 cate exactly the same position of the intra-arterial catheter as
Lung counts + Liver counts during 99mTc-MAA administration. This approach is adopted

for TARE/SIRT with the glass 90Y-microspheres
Any LSF value <10% constitutes no restriction for TARE/ (TheraSphere®) and yields safe and reproducible estimates
SIRT with 90Y-microspheres, which is instead contraindi- of the expected toxicity and clinical outcomes.
cated for LSF values >20%. For LSF values between 10% For treatment with the resin 90Y-microspheres (SIR-
and 20%, administered activity is adjusted as follows: for Spheres®), the activity to be administered can be estimated
LSF values between 10% and 15%, activity is reduced by according to the empiric method or to the body surface area
20%, whereas for LSF values included between 15% and (BSA) method. For the empiric method, the fraction of liver
20%, activity is reduced by 40% (see Fig. 33.4). occupied by the tumor relative to the overall liver volume is
Different models have been developed to calculate the 90Y estimated by either CT or MRI. For a tumor/liver fraction
activity allowing delivery of the highest dose to the tumor <0.25, an activity of 2 GBq is recommended, increasing to
LSF = 6.4% LSF = 12% LSF = 24%
Fig. 33.4 Estimation of the lung shunt fraction (LSF) estimated in was therefore the full amount planned on the basis of the dosimetry
three different patients as derived from planar gamma camera imaging estimate. A 20% reduction for administered 90Y-microsphere activity
after trans-arterial injection of 99mTc-MAA. The anterior and posterior was instead applied to the patient shown in the center panel (with 12%
views are displayed for each patient, with drawing of ROIs over the LSF), while the patient shown in right panel was excluded from treat-
liver and the lung fields, respectively; each ROI is first manually drawn ment because of a LSF exceeding 20% (reproduced with permission
on the anterior view, and then it is flipped on the horizontal axis to from: Boni G, Guidoccio F, Volterrani D, Mariani G. Radionuclide
match the posterior view. The LSF value is calculated using geometric therapy of tumor of the liver and biliary tract. In: Strauss HW, Mariani
means of the ROI counts from the two orthogonal views according to G, Volterrani D, Larson SM, Eds. Nuclear Oncology – From
the equation described in the text. LSF for the patient shown in left Pathophysiology to Clinical Applications. New York: Springer;
panel was 6.4%; the 90Y-microsphere activity injected in this patient 2017:1337–60)
2.5 GBq for tumor fractions between 0.25 and 0.5 and fur-
ther on to 3 GBq for tumor fractions >0.5. On the other hand, Key Learning Points
a tumor/liver fraction >0.7 constitutes an absolute contrain- • Pretreatment radiation dosimetry estimates are cru-
dication to treatment. cial in order to minimize possible adverse side
The following equation is used to calculate the effects of radioembolization due to excessive irra-
90
Y-microsphere activity (A) to be administered (BSA being diation of nontarget tissues (either normal liver
expressed in sq m): parenchyma, the pulmonary parenchyma, or other
sites in the splanchnic arterial bed).
Tumor volume
A ( GBq ) = ( BSA − 0.2 ) + • Geometric mean values derived from planar imag-
Total liver volume ing are normally utilized to calculate the liver-to-
Although these two methods are routinely used, both of lung shunt fraction (LSF).
them frequently overestimate the activity to be administered • No restrictions to the planned administered activity
[38, 39], and the MIRD-based partition model is recom- are applied for LSF values <10%.
mended also when treating patients with the resin • LSF >20% constitute an absolute contraindication
90
Y-microspheres. to radioembolization therapy.
Concerning instead pretreatment dosimetry estimates when • Administered activity is reduced by 20% for LSF
using 166Ho-PLLA-microspheres, the following equation is values between 10% and 15%.
recommended by the manufacturer to calculate the activity to • Administered activity is reduced by 40% for LSF
be administered (AHo166) in order not to exceed a certain radia- values between 15% and 20%.
tion burden to the whole liver (LD, expressed in Gy): • Different models have been developed to calculate
the 90Y activity allowing delivery of the highest
A
Ho166
( MBq ) = LD ( Gy ) × LW ( kg ) × 63 ( MBq / J ) dose to the tumor while sparing normal liver tissue,
where LW is weight of the liver as assessed by either CT or the MIRD-based partition model and the model
MRI. By optimizing the formula to a maximum liver dose of based on body surface area for glass microspheres
60 Gy, the following equation is obtained: and for resin microspheres, respectively.
• A specific equation based on threshold radiation
A
Ho166
( MBq ) = LD ( Gy ) × 3781( MBq / kg ) × LW ( kg ) dose to the whole liver and on liver mass is utilized
when planning radioembolization therapy with
From all the above equations, it follows that accurate evalu- 166
Ho-microspheres.
ation of the liver and tumor mass (generally based on anatomic • The predictive value of 99mTc-MAA scintigraphy as
imaging) and of 99mTc-MAA distribution (based on scinti- to the actual distribution of therapeutic radiolabeled
graphic imaging) is required for patient-specific dosimetry, microspheres in the liver is still debated.
although the predictive value of 99mTc-MAA scintigraphy as to • A definite dose-response correlation has been shown
the distribution of 90Y-microspheres in the liver is still debated by most of the retrospective studies so far reported.
[40–42]. There are in fact some parameters that potentially
induce discrepancies between intrahepatic 99mTc-MAA distri-
bution and 90Y-microsphere distribution, such as interval differ-
ences in catheter position during injection in the two separate 33.7 90Y-Microsphere Administration
interventional radiology sessions, physiologic changes in
hepatic blood flow, and differences in size and shape of the Under fluoroscopic guidance during transcutaneous arterial
99m
Tc-MAA particles and the 90Y-microspheres [41, 42]. catheterization (and trying to replicate as closely as possible
Thus, some investigators consider as suboptimal the abil- the same catheter positioning as during pretreatment admin-
ity of 99mTc-MAA to predict radiation dosimetry expected istration of 99mTc-MAA), the suspension of 90Y-microspheres
from 90Y-microsphere administration; however, a definite is injected into the artery feeding the target lesion(s). When
dose-response correlation has been shown by most of the ret- treatment is directed to one liver lobe, the procedure is
rospective studies so far reported using this technique, with defined as “selective,” whereas the term “super-selective” is
threshold values for objective tumor response varying from used for treatment directed to specific liver segment(s).
120 to 205 Gy (and even up to 500 Gy) [43–45]. Nevertheless, There are different procedures for delivering the therapeu-
no univocal threshold value has been identified as yet regard- tic amount of 90Y-microspheres, either manually using a dedi-
ing objective tumor response. cated delivery system operating in a step-by-step manner,
Dosimetry-based methods utilized to calculate the activ- which is useful to avoid early full embolization of the arterial
ity to be administered for TARE/SIRT are described in detail bed that might prevent successful infusion of the total intended
in Chap. 11 of this book. amount of radioactivity. The procedure is usually performed
alternating injections of iodinated contrast medium and sterile Instead, PET imaging based on the emission of β+ parti-
water/glucose solution (when administering the resin cles by 90Y provides a better way of assessing distribution of
90
Y-microspheres) or saline solution (when administering the the radiolabeled microspheres [47]. In this regard, in addi-
glass 90Y-microspheres). Continuous fluoroscopy monitoring tion to the largely predominant emission of β− particles, 90Y
is employed to ensure the occurrence of no stasis during infu- decays also through internal pair production that generates
sion and to confirm that flow of the infused material is similar 511 keV annihilation photons, even if in the extremely small
to that observed during the prior angiographic workup. fraction of 32 events per million. When performing radio-
embolization of liver tumors with 90Y-microspheres, a very
high concentration of the radionuclide is reached in a rela-
33.8 Posttreatment Scan tively small volume at the administration site; thus, even the
extremely small fraction of 90Y decays that occurs through
The pattern of 90Y-microsphere deposition is assessed by internal pair production is sufficient to acquire clinically
high-quality imaging early after treatment, to confirm that useful PET images for validation and dosimetric purposes
no significant radioactivity accumulation in gastrointestinal [48–51] (Fig. 33.5). In this manner, PET imaging allows
tract has occurred and to evaluate the radiation-absorbed dose detection of possible extrahepatic distribution of the
delivered to the tumor. 90
Y-microspheres, and the absorbed dose delivered during
Although a post-therapy planar and SPECT/CT scan based the radioembolization procedure can reliably be estimated
on the bremsstrahlung emission generated by the high-energy [47, 48].
β− particles of 90Y has been described to confirm correct depo- Depending on local regulation of radioprotection, treat-
sition of the radiolabeled microspheres in the tumor lesions ment with 90Y-microspheres can be performed as an outpa-
[46], images obtained in this manner are usually of very poor tient procedure, thus making it possible to discharge the
quality and do not allow an accurate quantification of micro- patient from the hospital on the same day of treatment or on
sphere distribution, especially when treating small lesions. the next day.
99mTc-MAA SPECT/CT 90Y-PET/CT
Fig. 33.5 Correspondence between intrahepatic distribution of pre- image in upper left panel, fused PET/CT image in lower left panel, and
treatment 99mTc-MAA scintigraphy (SPECT/CT) after injection into the 3D surface volumetric rendering in lower right panel (reproduced with
right hepatic artery (left) and posttreatment 90Y-PET/CT (right). For permission from: Boni G, Guidoccio F, Volterrani D, Mariani
SPECT/CT, the MIP image is displayed in upper right panel, the axial G. Radionuclide therapy of tumor of the liver and biliary tract. In:
CT image in upper left panel, the fused axial SPECT/CT image in lower Strauss HW, Mariani G, Volterrani D, Larson SM, Eds. Nuclear
left panel, and the 3D surface volumetric rendering in lower right panel. Oncology – From Pathophysiology to Clinical Applications. New York:
Similar displays for PET/CT: MIP image in upper right panel, axial CT Springer; 2017:1337–60)
els of the tumor-associated marker alpha-fetoprotein (AFP)

Key Learning Points constitute a basis to assess evolution of the disease. In gen-
• The suspension of therapeutic radiolabeled micro- eral, contrast-enhanced CT is performed at 1 month post-
spheres is injected through selective (directed to treatment; subsequently non-contrast-enhanced CT scans
one liver lobe) or super-selective (directed to one or are repeated every 3 months to assess response to treatment.
more liver segments) arterial catheterization. In addition to the general RECIST criteria, tumor response
• Posttreatment images to verify distribution of the to therapy can be assessed by specific criteria for liver
therapeutic 90Y-microspheres can be obtained by malignancies defined by the World Health Organization
acquiring the bremsstrahlung emission generated (WHO) and by the European Association for the Study of
by the high-energy β− particles or preferably by the Liver (EASL); such criteria are based on size parameters
PET acquisition based on internal pair and on necrosis parameters in the target lesions, respectively
production. [13, 52].
• When using 166Ho-microspheres for radioemboliza- Metabolic assessment with [18F]FDG PET/CT provides
tion, posttreatment imaging can be acquired using a highly useful prognostic information to assess response to
standard gamma camera based on the 81 keV γ-rays TARE/SIRT with 90Y-microspheres [53–56]. Similarly as
emitted during decay. observed for other malignancies in different clinical settings,
functional metabolic imaging with PET performs better than
conventional criteria based on morphology, such as RECIST,
33.9 P
atient Follow-Up and Assessment for early assessment of tumor response to treatment, as con-
of Response to Therapy vincingly demonstrated in patients with either HCC [57, 58],
intrahepatic cholangiocarcinoma [56] (Fig. 33.6), metastatic
Clinical and radiological monitoring is useful to assess tumor CRC [59, 60] (Figs. 33.7 and 33.8), liver metastases from
response to radioembolization, while routine blood chemis- breast cancer [61], or metastatic neuroendocrine malignan-
try is useful to monitor possible toxicity due to treatment; cies (the latter using 68Ga-DOTANOC as the PET tracer) [62]
when treating patients with HCC, changes in the serum lev- (Fig. 33.9).
a b 1.0 Responder
Non-responder
censored
censored
0.5
Cumulative survival
0.6
0.4
c d
0.2
0.0
0 50 100 150 200

Weeks
Fig. 33.6 Left panel: response to trans-arterial therapy with survival curves as a function of ΔSUV2SD. Patients responding to TARE
90
Y-microspheres in a patient with intrahepatic cholangiocarcinoma. (a) (blue line) had significantly longer survival (P < 0.05) than patients
Axial fused pre-TARE [18F]FDG-PET/CT section; (b) corresponding who did not respond (green line) (modified from: Haug AR, Heinemann
section of diagnostic CT; (c) axial fused post-TARE [18F]FDG-PET/CT V, Bruns CJ, Hoffmann R, Jakobs T, Bartenstein P, Hacker M. 18F-FDG
section; (d) corresponding section of diagnostic CT. SUVmax declined PET independently predicts survival in patients with cholangiocellular
by 70% 3 months after TARE, and serum CA 19-9 declined from 85.2 carcinoma treated with 90Y microspheres. Eur J Nucl Med Mol Imaging.
to 49.2 U/mL; the patient was still alive 12 months after TARE without 2011;38:1037–45)
evidence of progression within the liver. Right panel: Kaplan-Meier
a b
c d
Liver PFS
1.0
PR
SD
0.8
0.6
0.4
0.2
0.0
0.00 5.00 10.00 15.00 20.00

Months
Fig. 33.7 Left panel: coronal PET (left), axial fused [18F]FDG-PET/ having taken place in some of them (red arrow indicating the same
CT (a, c), and axial contrast-enhanced CT (ce-CT) sections (b, d) in a lesion in segment VIII). This ce-CT pattern was incorrectly reported as
patient with metastatic colorectal cancer before and 6 weeks after proof of disease progression. Right panel: Kaplan-Meier plots of
TARE with 90Y-microspheres. (a) Pre-TARE PET/CT shows increased progression-free survival (PFS) in relation to responses seen on [18F]
[18F]FDG uptake in metastatic lesions in segments I, II, IVa, VII, and FDG-PET/CT. Patients who had PR or stable disease on [18F]FDG-
VIII. (b) Pre-TARE ce-CT shows some of the metabolically active PET/CT had median PFS of 12 months and 5 months, respectively
lesions as low attenuation lesions, while several lesions are isointense (modified and reproduced with permission from: Zerizer I, Al-Nahhas
and difficult to distinguish (red arrow in segment VIII). (c) Post-TARE A, Towey D, Tait P, Ariff B, Wasan H, et al. The role of early 18F-FDG
PET/CT shows excellent partial response (PR) with marked reduction PET/CT in prediction of progression-free survival after 90Y radioembo-
in intensity and extent of uptake in metastatic lesions. (d) Post-TARE lization: comparison with RECIST and tumour density criteria. Eur J
ce-CT shows multiple new low attenuation lesions, necrotic changes Nucl Med Mol Imaging. 2012;39:1391–9)
a c
b d
1.0 metabolic response

non-responders
responders
censored
censored
0.8
Cumulative survival
0.6
0.4
0.2
0.0
0 10 20 30 40 50
Overall survival (mo)
Fig. 33.8 Left panel: axial fused [18F]FDG PET/CT and CT images Meier plots of overall survival in relation to responses seen on [18F]
pre-TARE (a) and 4 weeks post-TARE (b) in a patient with metastatic FDG-PET/CT. Patients with good metabolic response survived signifi-
colorectal cancer; remarkable metabolic response was associated with a cantly longer than patients who did not (modified and reproduced with
12-month survival after treatment. Center panel: axial fused [18F]FDG permission from: Sabet A, Meyer C, Aouf A, Sabet A, Ghamari S, Pieper
PET/CT and CT images pre-Tare (a) and 4 weeks post-TARE (b) in a CC, et al. Early post-treatment FDG PET predicts survival after 90Y
patient with metastatic colorectal cancer; poor metabolic response was microsphere radioembolization in liver-dominant metastatic colorectal
associated with 5-month survival after treatment. Right panel: Kaplan- cancer. Eur J Nucl Med Mol Imaging. 2015;42:370–6)
33.10 90Y-Microsphere Radioembolization specific clinical conditions—even in patients with advanced-

Combined with Other Therapies stage HCC [63] or with unresectable liver-predominant
metastases from colorectal cancer.
Although TARE/SIRT with 90Y-microspheres is a valid ther- In the following clinical conditions, patients with nonre-
apeutic option per se (especially in patients with early-stage sectable/ablatable HCC can benefit from SARE/SIRT com-
nonresectable liver-predominant malignancies), it is also bined with other therapies:
useful in combination with systemic therapies, depending on
a b c
100
80
Survival probability (%)
60
40
20
0
10 15 20 25 30 35 40
Time
Fig. 33.9 Left upper panel: (a) axial 68Ga-DOTANOC PET section mulation in the tumor mass. (c) Axial 68Ga-DOTANOC PET slice
pre-TARE shows intense tracer uptake in large metastasis from a neuro- acquired 6 weeks post-TARE shows substantially unchanged tumor
endocrine neoplasm in the right hepatic lobe (arrow). (b) Fused axial uptake (arrow), consistent with no response (ΔT/S was −24.6%).
90
Y-PET/CT image post-TARE demonstrates radioactivity accumula- Overall survival was 23 months. Right panel: Kaplan-Meier survival
tion in the tumor mass (with doughnut-like pattern due to necrotic analysis in relation to ΔT/S measured 6 weeks post-TARE. Patients
core). (c) Axial 68Ga-DOTANOC PET slice acquired 6 weeks post- with ΔT/S less than −50% (dashed line) had significantly lower
TARE shows significant reduction of tracer uptake in the tumor (arrow), (P < 0.001) overall survival than patients with ΔT/S more than −50%
consistent with molecular response (ΔT/S was −73.4%); overall sur- (solid line) (modified and reproduced with permission from: Filippi L,
vival was 34 months. Left lower panel: (a) axial 68Ga-DOTANOC PET Scopinaro F, Pelle G, Cianni R, Salvatori R, Schillaci O, et al. Molecular
section pre-TARE shows intense tracer accumulation in metastasis response assessed by 68Ga-DOTANOC and survival after 90Y micro-
from neuroendocrine neoplasm in segment IV (arrow). (b) Fused axial sphere therapy in patients with liver metastases from neuroendocrine
90
Y-PET PET/CT slice acquired post-TARE shows radioactivity accu- tumours. Eur J Nucl Med Mol Imaging. 2016;43:432–40)
• As a neoadjuvant treatment before planned resection or tumor size does not completely define all types of response
transplantation [64–66]. to antitumor therapy. When tumor size changes and arterial
• As an alternative to other ablative treatments in patients contrast enhancement are combined with the response crite-
with portal vein occlusion—who are generally excluded ria of the European Society for the Study of the Liver
from a targeted therapy [67]. (EASL), overall tumor response rates between 40% and 90%
• In combination with systemic therapies based on biologi- can be observed, with an 80–100% disease control rate in
cal targeting agents, e.g., sorafenib [68–71]. targeted lesion. Changes in vascular enhancement can be
• As first-line treatment in combination with chemotherapy, detected about 2 months after TARE/SIRT, earlier than
to downstage the disease [72]. changes in tumor size [75].
• As a second-line treatment and as salvage treatment in When liver transplantation is considered as curative treat-
patients with liver-predominant metastatic colorectal can- ment for patients with HCC (i.e., in patients with a single
cer, in combination with systemic chemotherapy based on lesion <5 cm in diameter, or with ≤3 lesions, all <3 cm
radiosensitizing drugs (oxaliplatin, 5-FU, irinotecan) [73]. without extrahepatic metastases or portal vein thrombosis),
90
Y-microsphere radioembolization slows down the progres-
sion of HCC, thus allowing the patients more time to wait for
Key Learning Points donor organs [76–78].
• Tumor response to radioembolization with Good candidates to TARE/SIRT are also those patients
90
Y-microspheres is assessed by clinical and imag- whose disease is too advanced to meet transplant criteria, but
ing criteria. do not have malignant portal vein thrombosis or metastatic
• Assessment of tumor-associated markers can be HCC. In these conditions, radioembolization downstages the
useful to monitor the efficacy of radioembolization disease to the point that >50% of the patients who were ini-
with 90Y-microspheres in case of measurable, tially non-eligible for transplant according to Milan criteria
abnormally increased serum levels pretreatment. then become eligible. Furthermore, TARE/SIRT prolongs
• In addition to morphological RECIST criteria, overall survival of these patients [76–78].
tumor response to therapy can be assessed by spe- Improved survival has been observed also in patients with
cific criteria for liver malignancies, based on size malignant portal vein thrombosis involvement [79]. Instead,
parameters or on necrosis parameters in the target the presence of distant metastases contraindicates radioem-
lesions. bolization with 90Y-microspheres.
• Functional metabolic imaging with [18F]FDG PET
(or with 68Ga-DOTANOC in case of neuroendocrine
malignancies) performs better than RECIST for 33.11.2 Intrahepatic Cholangiocarcinoma
early assessment of tumor response to treatment.
• Radioembolization with 90Y-microspheres can be Even in patients who have resectable intrahepatic cholan-
combined with systemic treatments, either conven- giocarcinoma (ICC), surgery is associated with only mod-
tional chemotherapy or therapy with novel biologi- est survival benefit. On the other hand, TACE ensues some
cal targeting drugs. improvement in survival but is associated with high tox-
icity, whereas, although based on limited clinical experi-
ence, TARE/SIRT has shown favorable tumor response and
improved survival, especially in patients with better ECOG
33.11 C
linical Indications in Primary Liver performance status [80]. In particular, partial response and
Tumors stable disease rates between 28% and 54% at 3 months post-
treatment have been reported, with overall median survival
33.11.1 Hepatocellular Carcinoma of 15.5 months and downstaging to surgery in about 10% of
patients [81]. Encouraging results have been obtained also
Treatment of HCC patients with TARE/SIRT is character- with the combination of TARE/SIRT with chemotherapy as
ized by overall clinical safety and improved survival, espe- a strategy for downstaging ICC to achieve surgical resect-
cially if employed earlier in the course of disease [74]. ability [82]. The conclusion of investigations on different
Variable tumor biology is associated with variable locoregional treatments for patients with unresectable ICC
response rates to radioembolization with 90Y-microspheres, is that TARE/SIRT with 90Y-microspheres is best suited for
also depending on variations in timing of assessment of patients who are not eligible for intra-arterial chemotherapy
tumor response and/or in treatment intensity. When evalu- [83, 84].
ated according to the WHO criteria, the median time to As to assessment of tumor response to TARE/SIRT, [18F]
response after TARE/SIRT is approximately 6 months for FDG-PET is the best independent predictor for patient out-
reduction in tumor size to occur [75]. However, evaluation of come after radioembolization treatment [56].
33.12 C
linical Indications in Metastatic Liver
Tumors • Radioembolization with 90Y-microspheres results in
clinical benefit also for patients with inoperable
The use of TARE/SIRT in patients with hepatic metastases intrahepatic cholangiocarcinoma, especially if com-
from other malignancies [85] has been extensively inves- bined with chemotherapy as a strategy for down-
tigated, either as a single treatment or associated with sys- staging the tumor to achieve surgical resectability.
temic chemotherapy. • In combination with systemic chemotherapy, radio-
embolization with 90Y-microspheres results in gen-
eral in enhanced tumor response in terms of
33.12.1 Metastatic Colorectal Carcinoma objective response, longer time to progression and
longer survival than chemotherapy alone.
TARE/SIRT is an effective treatment for metastatic CRC • Although based on more limited clinical experience,
to the liver, particularly for patients with unresectable liver radioembolization with 90Y-microspheres results in
metastases on systemic chemotherapy or who have not very high response rates in >95% of patients with
responded to first- or second-line chemotherapy. In these liver metastatic disease from neuroendocrine tumors.
patients, [18F]FDG-PET is especially suited to assess tumor
response to radioembolization [53, 54, 86]. The overall safety
and efficacy of TARE/SIRT for unresectable, chemorefrac-
tory metastatic CRC have been largely validated, resulting 33.13 Absolute and Relative
in a median time to intrahepatic progression of 9 months and Contraindications
overall survival of 12 months [87].
Several studies have investigated the efficacy of The main absolute contraindication to TARE/SIRT is the
90
Y-microsphere TARE/SIRT in combination with systemic presence of important hepato-pulmonary shunting demon-
chemotherapy (exploring different antitumor agents), result- strated by the pretreatment 99mTc-MAA scan, because of the
ing in general in enhanced tumor response in terms of objec- risk to deliver a radiation dose to the lungs >30 Gy with a
tive response, longer time to progression, and longer survival single infusion (or even 50 Gy for repeat treatment). The
than with chemotherapy alone [88–96]. inability to prevent deposition of the radiolabeled micro-
spheres in the gastrointestinal tract is an additional absolute
contraindication to TARE/SIRT.
33.12.2 Metastatic Neuroendocrine Tumors Concerning instead relative contraindications, the choice
of performing TARE/SIRT should be carefully evaluated in
Liver metastases from neuroendocrine tumors such as carci- patients with reduced pulmonary function, poor functional
noids, VIPomas, gastrinomas, and somatostatinomas are gen- liver reserve, serum creatinine >2.0 mg/dL, and platelet
erally well arterialized and represent therefore an ideal target count <75 × 109/L. In these conditions, the possibility of
for locoregional trans-arterial therapies. TARE/SIRT with using radiolabeled lipiodol rather than the radiolabeled
90
Y-microspheres in metastatic neuroendocrine tumors to the microspheres should also be considered [99].
liver is safe and effective, with very high response rates: any
response in >95% of patients (lasting for over 2 years) and
progressive disease in only 4.9% of the patients [97, 98]. 33.14 Early and Late Toxicities
A mild post-embolic syndrome (including fatigue, vague

Key Learning Points abdominal discomfort, pain, and fever) is the most com-
• Based on clinical experience acquired in different mon clinical toxicity, similarly as experienced with TACE
clinical settings, radioembolization with [100, 101]. In patients in whom deposition of the radiola-
90
Y-
microspheres is indicated in patients with beled microspheres in nontarget sites cannot be adequately
inoperable hepatocellular carcinoma, also a sal- avoided by adequate pre-therapy procedures and treatment
vage therapy in patients who are candidates to planning, additional possible toxicities include radiation
liver transplantation. cholecystitis, gastric ulceration, gastroduodenitis, pancreati-
• The only contraindication to radioembolization tis, radiation pneumonitis, and RILD [76, 79, 80, 102]. In the
with 90Y-microspheres in patients with hepatocellu- immediate post-TARE/SIRT period, lymphopenia is com-
lar carcinoma is the occurrence of predominant dis- monly observed, due to the well-known radiosensitivity of
tant metastatic disease. lymphocytes; nevertheless, no serious infectious complica-
tions have been reported [13, 33, 52–54, 100–103].
33.15 Perspectives on Radioembolization

for Primary and Metastatic Liver liver disease—when deposition of the radiolabeled
Tumors microspheres in nontarget sites cannot be prevented
by adequate pre-therapy procedures and treatment
TARE/SIRT with 90Y-microspheres is increasingly being dem- planning.
onstrated to constitute an effective monotherapy in patients • Although based on more limited clinical experience
with unresectable HCC. Nevertheless, newer approaches than with 90Y-microspheres, there is evidence that
being currently explored include combination therapies with radioembolization with 166Ho-microspheres results
systemic and locoregional agents, specifically sorafenib and in satisfactory overall tumor response rates.
TARE/SIRT, in the adjuvant or neoadjuvant setting—although
evidence confirming the actual clinical benefit of these combi-
nation therapies in patients with HCC is scarce as yet [68, 70].
Similarly, no sufficient data are available concerning References
treatment of metastatic liver disease (from, e.g., CRC or neu-
roendocrine tumors) with TARE/SIRT combined with other 1. El-Serag HB. Epidemiology of viral hepatitis and hepatocellular
carcinoma. Gastroenterology. 2012;142:1264–73.
systemic therapies [104]. 2. van der Pool AE, Damhuis RA, Ijzermans JN, de Wilt JH,
Eggermont AM, Kranse R, et al. Trends in incidence, treat-
ment and survival of patients with stage IV colorectal cancer: a
33.16 166Ho-PLLA-Microspheres population-based series. Color Dis. 2012;14:56–61.
3. Mazzaferro V, Regalia E, Doci R, Andreola S, Pulvirenti A,
Bozzetti F, et al. Liver transplantation for the treatment of small
As the latest addition to the armamentarium of formulations hepatocellular carcinomas in patients with cirrhosis. N Engl J
commercially available for TARE/SIRT [105–111], published Med. 1996;334:693–9.
evidence is limited regarding the clinical benefit of therapy 4. Ingold JA, Reed GB, Kaplan HS, Bagshaw MA. Radiation hepati-
tis. Am J Roentgenol Radium Ther Nucl Med. 1965;93:200–8.
with 166Ho-PLLA-microspheres in large groups of patients, 5. Lawrence TS, Robertson JM, Anscher MS, Jirtle RL, Ensminger
although data are very promising regarding intrahepatic distri- WD, Fajardo LF. Hepatic toxicity resulting from cancer treatment.
bution (tumor-to-normal liver ratios), safety of toxicity profile Int J Radiat Oncol Biol Phys. 1995;31:1237–48.
of the administration procedure, and efficacy of these particles 6. Yu H, Burke CT. Comparison of percutaneous ablation technolo-
gies in the treatment of malignant liver tumors. Semin Intervent
in both preclinical and pilot clinical studies [112–116]. Radiol. 2014;31:129–37.
In the latest report on these topics, Prince et al. described 7. Stuart K. Chemoembolization in the management of liver tumors.
a 73% overall response rate in patients with unresectable Oncologist. 2003;8:425–37.
liver metastases of any origin (either complete response, par- 8. Geschwind JFH. Chemoembolization for hepatocellular car-
cinoma: where does the truth lie? J Vasc Interv Radiol.
tial response, or stable disease as assessed by RECIST crite- 2002;13:991–4.
ria); in this study, disease control in the liver was achieved in 9. Varela M, Real MI, Burrel M, Forner A, Sala M, Brunet M, et al.
49% of patients [117]. Chemoembolization of hepatocellular carcinoma with drug elut-
ing beads: efficacy and doxorubicin pharmacokinetics. J Hepatol.
2007;46:474–81.
10. Llovet JM, Real MI, Montaña X, Planas R, Coll S, Aponte J, et al.
Key Learning Points Arterial embolisation or chemoembolisation versus symptomatic
• Radioembolization therapy is contraindicated by treatment in patients with unresectable hepatocellular carcinoma:
important hepato-pulmonary shunting demon- a randomised controlled trial. Lancet. 2002;359(9319):1734–9.
11. Kettenbach J, Stadler A, Katzler IV, Schernthaner R, Blum M,
strated by the pretreatment 99mTc-MAA scan. Lammer J, et al. Drug-loaded microspheres for the treatment
• The inability to prevent deposition of the radiola- of liver cancer: review of current results. Cardiovasc Intervent
beled microspheres in the gastrointestinal tract is an Radiol. 2008;31:468–76.
additional absolute contraindication to radioembo- 12. European Association For The Study Of The Liver; European
Organisation For Research And Treatment Of Cancer. EASLEORT
lization therapy. Clinical practice guidelines: management of hepatocellular carci-
• Reduced pulmonary function, poor functional liver noma. J Hepatol. 2012;56:908–43.
reserve, serum creatinine >2.0 mg/dL, and platelet 13. Salem R, Lewandowski RJ, Atassi B, Gordon SC, Gates VL,
count <75 × 109/L constitute relative contraindica- Barakat O, et al. Treatment of unresectable hepatocellu-
lar carcinoma with use of 90Y microspheres (TheraSphere):
tions to radioembolization therapy. safety, tumor response, and survival. J Vasc Interv Radiol.
• A mild post-embolic syndrome is the most common 2005;16:1627–39.
clinical toxicity of radioembolization therapy. 14. Raoul JL, Bourguet P, Bretagne JF, Duvauferrier R, Coornaert S,
• Possible additional toxicities include radiation cho- Darnault P, et al. Hepatic artery injection of I-131-labelled lipi-
odol. Part I. Biodistribution study results in patients with hepato-
lecystitis, gastric ulceration, gastroduodenitis, pan- cellular carcinoma. Radiology. 1988;168:541–5.
creatitis, radiation pneumonitis, or radiation-induced
15. Nakajo M, Kobayashi H, Shimabukuro K, Shirono K, Sakata H, cancer with yttrium-90 microspheres (TheraSphere): assessment
Taguchi M, et al. Biodistribution and in vivo kinetics of iodine-131 of hepatic arterial embolization. Cardiovasc Intervent Radiol.
lipiodol infused via the hepatic artery of patients with hepatic can- 2006;29:522–9.
cers. J Nucl Med. 1988;29:1066–77. 32. Salem R, Thurston KG. Radioembolization with 90Yttrium micro-
16. Smits ML, Nijsen JF, van den Bosch MA, Lam MG, Vente MA, spheres: a state-of-the-art brachytherapy treatment for primary
Huijbregts JE, et al. Holmium-166 radioembolization for the and secondary liver malignancies. Part 1: Technical and methodo-
treatment of patients with liver metastases: design of the phase logic considerations. J Vasc Interv Radiol. 2006;17:1251–78.
I HEPAR trial. J Exp Clin Cancer Res. 2010;29:70. https://doi. 33. Hamami ME, Poeppel TD, Müller S, Heusner T, Bockisch A,
org/10.1186/1756-9966-29-70. Hilgard P, Antoch G. SPECT/CT with 99mTc-MAA in radioem-
17. Nowicki ML, Cwikla JB, Sankowski AJ, Shcherbinin S, Grimmes bolization with 90Y microspheres in patients with hepatocellular
J, Celler A, et al. Initial study of radiological and clinical effi- cancer. J Nucl Med. 2009;50:688–92.
cacy radioembolization using 188Re-human serum albumin (HSA) 34. Grosser OS, Rufi J, Kupitz D, Pethe A, Ulrich G, Genseke P, et al.
microspheres in patients with progressive, unresectable primary or Pharmacokinetics of 99mTc-MAA- and 99mTc-HSA microspheres
secondary liver cancers. Med Sci Monit. 2014;20:1353–62. used in preradioembolization dosimetry: influence on the liver–
18. Bhattacharya S, Dhillon AP, Winslet MC, Davidson BR, Shukla lung shunt. J Nucl Med. 2016;57:925–7.
N, Gupta SD, et al. Human liver cancer cells and endothelial cells 35. Sabet A, Ahmadzadehfar H, Muckle M, Haslerud T, Wilhelm K,
incorporate iodised oil. Br J Cancer. 1996;73:877–81. Biersack HJ, et al. Significance of oral administration of sodium
19. Madsen MT, Park CH, Thakur ML. Dosimetry of iodine-131 perchlorate in planning liver-directed radioembolisation. J Nucl
ethiodol in the treatment of hepatoma. J Nucl Med. Med. 2011;52:1063–7.
1988;29:1038–44. 36. Dale RG. Dose-rate effects in targeted radiotherapy. Phys Med
20. Monsieurs MA, Bacher K, Brans B, Vral A, De Ridder L, Dierckx Biol. 1996;41:1871–84.
RA, et al. Patient dosimetry for 131I-lipiodol therapy. Eur J Nucl 37. Ho S, Lau WY, Leung TW, Chan M, Chan KW, Lee WY, et al.
Med Mol Imaging. 2003;30:554–61. Partition model for estimating radioation doses from yttrium-90
21. Bhattacharya S, Novell JR, Dusheiko GM, Hilson AJ, Dick R, microspheres in treating hepatic tumors. Eur J Nucl Med.
Hobbs KE. Epirubicin-Lipiodol chemotherapy versus 131iodine- 1996;23:947–52.
lipiodol radiotherapy in the treatment of unresectable hepatocel- 38. Kennedy AS, Dezarn WA, McNeillie P, Overton C, England M,
lular carcinoma. Cancer. 1995;76:2202–10. Sailer SL. Fractionation, dose selection, and response of hepatic
22. Yoo HS, Park CH, Lee JT, Kim KW, Yoon CS, Suh JH, et al. metastases of neuroendocrine tumors after 90Y-microsphere
Small hepatocellular carcinoma: high dose internal radia- brachytherapy. Brachytheraphy. 2006;5:103–4.
tion therapy with superselective intra-arterial injection of 39. Kennedy AS, Dezarn WA, McNeillie P, Overton C, England M,
I-131-labeled Lipiodol. Cancer Chemother Pharmacol. Sailer SL. Dose selection of resin 90Y-microspheres for liver brachy-
1994;33:S128–S33. therapy: a single center review. Brachytheraphy. 2006;5:104.
23. Leung WT, Lau WY, Ho S, Chan M, Leung N, Lin J, et al. 40. Flamen P, Vanderlinden B, Delatte P, Ghanem G, Ameye L, Van
Selective internal radiation therapy with intra-arterial iodine-131- Den Eyden M, et al. Multimodality imaging can predict the meta-
Lipiodol in inoperable hepatocellular carcinoma. J Nucl Med. bolic response of unresectable colorectal liver metastases to radio-
1994;35:1313–8. embolisation therapy with yttrium-90 labeled resin microspheres.
24. Raoul JL, Guyader D, Bretagne JF, Duvauferrier R, Bourguet Phys Med Biol. 2008;53:6591–3.
P, Bekhechi D, et al. Randomized controlled trial for hepato- 41. Jiang M, Fischman A, Nowakowski FS, Sherif Heiba S, Zhuangyu
cellular carcinoma with portal vein thrombosis: intra-arterial Zhang Z, Kneusaurek K, et al. Segmental perfusion differences
iodine-131-iodized oil versus medical support. J Nucl Med. on paired Tc-99m macroaggregated albumin (MAA) hepatic per-
1994;35:1782–7. fusion imaging and yttrium-90 (Y-90) bremsstrahlung imaging
25. Boucher E, Garin E, Guillygomac’h A, Olivie D, Boudjema K, studies in SIR-sphere radioembolization: associations with angi-
Raoul JL. Intra-arterial injection of iodine-131-labeled lipiodol ography. J Nucl Med Radiat Ther. 2012;3:122.
for treatment of hepatocellular carcinoma. Radiother Oncol. 42. Wondergem M, Smits MLJ, Elschot M, de Jong HWAM,
2007;82:76–82. Verkooijen HM, van den Bosch MAAJ, et al. 99mTc-
26. Partensky C, Sassolas G, Henry L, Paliard P, Maddern GJ. Intra- Macroaggregated albumin poorly predicts the intrahepatic distri-
arterial iodine 131-labeled lipiodol as adjuvant therapy after cura- bution of 90Y resin microspheres in hepatic radioembolization. J
tive liver resection for hepatocellular carcinoma: a phase 2 clinical Nucl Med. 2013;54:1294–301.
study. Arch Surg. 2000;135:1298–300. 43. Lau WY, Sangro B, Chen PJ, Cheng SQ, Chow P, Lee RC, et al.
27. Lau WY, Lai EC, Leung TW, Yu SC. Adjuvant intra-arterial Treatment for hepatocellular carcinoma with portal vein tumor
iodine-131-labeled lipiodol for resectable hepatocellular car- thrombosis: the emerging role for radioembolization using
cinoma: a prospective randomized trial-update on 5-year and yttrium-90. Oncology. 2013;84:311–8.
10-year survival. Ann Surg. 2008;247:43–8. 44. Garin E, Lenoir L, Rolland Y, Edeline J, Mesbah H, Laffont
28. Lee YS, Jeong JM, Kim YJ, Chung JW, Park JH, Suh YG, et al. S, et al. Dosimetry based on 99mTc-macroaggregated albumin
Synthesis of 188Re-labelled long chain alkyl diaminedithiol for SPECT/CT accurately predicts tumor response and survival in
therapy of liver cancer. Nucl Med Commun. 2002;23:237–42. hepatocellular carcinoma patients treated with 90Y-loaded glass
29. De Ruyck K, Lambert B, Bacher K, Gemmel F, De Vos F, Vral microspheres: preliminary results. J Nucl Med. 2012;53:255–63.
A, et al. Biologic dosimetry of 188Re-HDD/lipiodol versus 45. Mazzaferro V, Sposito C, Bhoori S, Romito R, Chiesa C,
131
I-lipiodol therapy in patients with hepatocellular carcinoma. J Morosi C, et al. Yttrium-90 radioembolization for intermediate-
Nucl Med. 2004;45:612–8. advanced hepatocellular carcinoma: a phase 2 study. Hepatology.
30. Kumar A, Srivastava DN, Chau TT, Long HD, Bal C, Chandra 2013;57:1826–37.
P, et al. Inoperable hepatocellular carcinoma: transarterial 188Re 46. Ahmadzadehfar H, Muckle M, Sabet A, Wilhelm K, Kuhl C,
HDD-labeled iodized oil for treatment. Prospective multicenter Biermann K, et al. The significance of bremsstrahlung SPECT/
clinical trial. Radiology. 2007;243:509–19. CT after yttrium-90 radioembolisation treatment in the predic-
31. Sato K, Lewandowski RJ, Bui JT, Omary R, Hunter RD, Kulik tion of extrahepatic side effects. Eur J Nucl Med Mol Imaging.
L, et al. Treatment of unresectable primary and metastatic liver 2011;39:309–15.
47. Lhommel R, Goffette P, Van del Eynde M, Jamar F, Pauwels S, ses from neuroendocrine tumours. Eur J Nucl Med Mol Imaging.
Bilbao JI, et al. Yttrium-90 TOF PET scan demonstrates high- 2016;43:432–40.
resolution biodistribution after liver SIRT. Eur J Nucl Med Mol 63. Fidelman N, Kerlan RK Jr. Transarterial chemoembolization and
Imaging. 2009;36:1696. 90
Y radioembolization for hepatocellular carcinoma: review of
48. Lhommel R, van Elmbt L, Goffette P, Van den Eynde M, Jamar F, current applications beyond intermediate-stage disease. AJR Am J
Pauwels S, et al. Feasibility of 90Y TOF PET-based dosimetry in Roentgenol. 2015;205:742–52.
liver metastasis therapy using SIR-Spheres. Eur J Nucl Med Mol 64. Braat AJ, Huijbregts JE, Molenaar IQ, Borel Rinkes IH, van den
Imaging. 2010;37:1654–62. Bosch MA, Lam MG. Hepatic radioembolization as a bridge to
49. Kao YH, Tan EH, Lim KY, Eng CE, Goh SW. Yttrium-90 internal liver surgery. Front Oncol. 2014;4:199. https://doi.org/10.3389/
pair production imaging using first generation PET/CT provides fonc.2014.00199.. eCollection 2014
high resolution images for qualitative diagnostic purposes. Br J 65. Lewandowski RJ, Donahue L, Chokechanachaisakul A, Kulik L,
Radiol. 2012;85:1018–9. Mouli S, Caicedo J, et al. 90Y radiation lobectomy: outcomes fol-
50. Wright CL, Zhang J, Tweedle MF, Knopp MV, Hall NC. Theranostic lowing surgical resection in patients with hepatic tumors and small
imaging of yttrium-90. Biomed Res Int. 2015;2015:481279. future liver remnant volumes. J Surg Oncol. 2016;114:99–105.
51. Gnesin S, Canetti L, Adib S, Cherbuin N, Silva-Monteiro M, Bize 66. Teo JY, Allen JC Jr, Ng DC, Choo SP, Tai DW, Chang JP, et al. A
P, et al. Partition model based 99mTc-MAA SPECT/CT predictive systematic review of contralateral liver lobe hypertrophy after uni-
dosimetry compared to 90Y TOF PET/CT post treatment dosim- lobar selective internal radiation therapy with Y90. HPB (Oxford).
etry in radioembolisation of hepatocellular carcinoma: a quantita- 2016;18:7–12.
tive agreement comparison. J Nucl Med. 2016;57:1672–8. 67. Johnson GE, Monsky WL, Valji K, Hippe DS, Padia
52. Riaz A, Memon K, Miller FH, Nikolaidis P, Kulik LM, SA. Yttrium-90 radioembolization as a salvage treatment fol-
Lewandowski RJ, et al. Role of the EASL, RECIST, and WHO lowing chemoembolization for hepatocellular carcinoma. J Vasc
response guidelines alone or in combination for hepatocel- Interv Radiol. 2016;27:1123–9.
lular carcinoma: radiologic-pathologic correlation. J Hepatol. 68. Chow PK, Poon DY, Khin MW, Singh H, Han HS, Goh AS, et al,
2011;54:695–704. Asia-Pacific Hepatocellular Carcinoma Trials Group. Multicenter
53. Wong CY, Qing F, Savin M, Campbell J, Gates VL, Sherpa KM, phase II study of sequential radioembolization-sorafenib
et al. Reduction of metastatic load to liver after intraarterial therapy for inoperable hepatocellular carcinoma. PLoS One.
hepatic yttrium-90 radioembolization as evaluated by [18F]fluo- 2014;9(3):e90909. https://doi.org/10.1371/journal.pone.0090909.
rodeoxyglucose positron emission tomographic imaging. J Vasc eCollection 2014.
Interv Radiol. 2005;16:1101–6. 69. Kulik L, Vouche M, Koppe S, Lewandowski RJ, Mulcahy MF,
54. Miller FH, Keppke AL, Reddy D, Huang J, Jin J, Mulcahy MF, Ganger D, et al. Prospective randomized pilot study of Y90 +/−
Salem R. Response of liver metastases after treatment with sorafenib as bridge to transplantation in hepatocellular carcinoma.
yttrium-90 microspheres: role of size, necrosis, and PET. AJR Am J Hepatol. 2014;61:309–17.
J Roentgenol. 2007;188:776–83. 70. Ricke J, Bulla K, Kolligs F, Peck-Radosavljevic M, Reimer
55. Flamen P, Vanderlinden B, Delatte P, Ghanem G, Ameye L, Van P, Sangro B, et al. Safety and toxicity of radioembolization
Den Eynde M, et al. Multimodality imaging can predict the meta- plus Sorafenib in advanced hepatocellular carcinoma: analy-
bolic response of unresectable colorectal liver metastases to radio- sis of the European multicentre trial SORAMIC. Liver Int.
embolization therapy with Yttrium-90 labeled resin microspheres. 2015;35:620–6.
Phys Med Biol. 2008;53:6591–603. 71. Lorenzin D, Pravisani R, Leo CA, Bugiantella W, Soardo G,
56. Haug AR, Heinemann V, Bruns CJ, Hoffmann R, Jakobs T, Carnelutti A, et al. Complete remission of unresectable hepatocel-
Bartenstein P, Hacker M. 18F-FDG PET independently predicts sur- lular carcinoma after combined sorafenib and adjuvant yttrium-90
vival in patients with cholangiocellular carcinoma treated with 90Y radioembolization. Cancer Biother Radiopharm. 2016;31:65–9.
microspheres. Eur J Nucl Med Mol Imaging. 2011;38:1037–45. 72. Sangha BS, Nimeiri H, Hickey R, Salem R, Lewandowski
57. Sabet A, Ahmadzadehfar H, Bruhman J, Sabet A, Meyer C, RJ. Radioembolization as a treatment strategy for metastatic
Wasmuth JC, et al. Survival in patients with hepatocellular carci- colorectal cancer to the liver: what can we learn from the
noma treated with 90Y-microsphere radioembolization. Prediction SIRFLOX trial? Curr Treat Options Oncol. 2016;17:26.
by 18F-FDG PET. Nuklearmedizin. 2014;53:39–45. 73. Dutton SJ, Kenealy N, Love SB, Wasan HS, Sharma RA,
58. Hartenbach M, Weber S, Albert NL, Hartenbach S, Hirtl A, FOXFIRE Protocol Development Group and the NCRI Colorectal
Zacherl MJ, et al. Evaluating treatment response of radioemboliza- Clinical Study Group. FOXFIRE protocol: an open-label,
tion in intermediate-stage hepatocellular carcinoma patients using randomised, phase III trial of 5-fluorouracil, oxaliplatin and
18
F-fluoroethylcholine PET/CT. J Nucl Med. 2015;56:1661–6. folinic acid (OxMdG) with or without interventional Selective
59. Zerizer I, Al-Nahhas A, Towey D, Tait P, Ariff B, Wasan H, et al. Internal Radiation Therapy (SIRT) as first-line treatment for
The role of early 18F-FDG PET/CT in prediction of progression- patients with unresectable liver-only or liver-dominant meta-
free survival after 90Y radioembolization: comparison with static colorectal cancer. BMC Cancer. 2014;14:497. https://doi.
RECIST and tumour density criteria. Eur J Nucl Med Mol org/10.1186/1471-2407-14-497.
Imaging. 2012;39:1391–9. 74. Geschwind JF, Salem R, Carr BI, Soulen MC, Thurston KG, Goin
60. Sabet A, Meyer C, Aouf A, Sabet A, Ghamari S, Pieper CC, et al. KA, et al. Yttrium-90 microspheres for the treatment of hepatocel-
Early post-treatment FDG PET predicts survival after 90Y micro- lular carcinoma. Gastroenterology. 2004;127:S194–205.
sphere radioembolization in liver-dominant metastatic colorectal 75. Salem R, Lewandowsky RJ, Mulcahy MF, Riaz A, Ryu RK,
cancer. Eur J Nucl Med Mol Imaging. 2015;42:370–6. Ibrahim S, et al. Radioembolisation for hepatocellular carcinoma
61. Haug AR, Tiega Donfack BP, Trumm C, Zech CJ, Michl M, using Yttrium-90 microspheres. a comprehensive report of long
Laubender RP, et al. 18F-FDG PET/CT predicts survival after term outcomes. Gastroenterology. 2010;138:52–64.
radioembolization of hepatic metastases from breast cancer. J 76. Kulik LM, Atassi B, van Holsbeeck L, Souman T, Lewandowski
Nucl Med. 2012;53:371–7. RJ, Mulcahy MF, et al. Yttrium-90 microspheres (TheraSphere®)
62. Filippi L, Scopinaro F, Pelle G, Cianni R, Salvatori R, Schillaci O, treatment of unresectable hepatocellular carcinoma: downstag-
et al. Molecular response assessed by 68Ga-DOTANOC and suring to resection, RFA and bridge to transplantation. J Surg Oncol.
vival after 90Y microsphere therapy in patients with liver metasta- 2006;94:572–86.
77. Tohme S, Sukato D, Chen HW, Amesur N, Zajko AB, Humar liver metastases from colorectal cancer by using yttrium-90 resin
A, et al. Yttrium-90 radioembolization as a bridge to liver trans- microspheres plus concomitant systemic irinotecan chemother-
plantation: a single-institution experience. J Vasc Interv Radiol. apy. J Clin Oncol. 2009;27:4089–95.
2013;24:1632–8. 93. Gibbs P, Heinemann V, Sharma NK, Findlay MPN, Ricke J,
78. Abdelfattah MR, Al-Sebayel M, Broering D, Alsuhaibani Gebski V, et al. SIRFLOX Study Group. SIRFLOX: randomized
H. Radioembolization using yttrium-90 microspheres as bridging phase III trial comparing first-line mFOLFOX6 ± bevacizumab
and downstaging treatment for unresectable hepatocellular carci- (bev) versus mFOLFOX6 + selective internal radiation therapy
noma before liver transplantation: initial single-center experience. (SIRT) ± bev in patients (pts) with metastatic colorectal cancer
Transplant Proc. 2015;47:408–11. (mCRC). J Clin Oncol. 2015;33(Suppl):3502.
79. Kulik LM, Carr BI, Mulcahy MF, Lewandowski RJ, Atassi B, 94. Hong K, McBride JD, Georgiades CS, Reyes DK, Herman JM,
Ryu RK, et al. Safety and efficacy of 90Y radiotherapy for hepa- Kamel IR, et al. Salvage therapy for liver-dominant colorectal
tocellular carcinoma with and without portal vein thrombosis. metastatic adenocarcinoma: comparison between transcatheter
Hepatology. 2008;47:71–81. arterial chemoembolization versus yttrium-90 radioembolization.
80. Ibrahim SM, Mulcahy MF, Lewandowski RJ, Sato KT, Ryu RK, J Vasc Interv Radiol. 2009;20:360–7.
Masterson EJ, et al. Treatment of unresectable cholangiocarci- 95. Saxena A, Bester L, Shan L, Perera M, Gibbs P, Meteling B, et al.
noma using yttrium-90 microspheres: results from a pilot study. A systematic review on the safety and efficacy of yttrium-90 radio-
Cancer. 2008;113:2119–28. embolization for unresectable, chemorefractory colorectal cancer
81. Al-Adra DP, Gill RS, Axford SJ, Shi X, Kneteman N, Liau liver metastases. J Cancer Res Clin Oncol. 2014;140:537–47.
SS. Treatment of unresectable intrahepatic cholangiocarcinoma 96. Kennedy AS, Ball D, Cohen SJ, Cohn M, Coldwell DM, Drooz A,
with yttrium-90 radioembolization: a systematic review and et al. Multicenter evaluation of the safety and efficacy of radioem-
pooled analysis. Eur J Surg Oncol. 2015;41:120–7. bolization in patients with unresectable colorectal liver metastases
82. Rayar M, Sulpice L, Edeline J, Garin E, Levi Sandri GB, Meunier selected as candidates for 90Y resin microspheres. J Gastrointest
B, et al. Intra-arterial yttrium-90 radioembolization combined Oncol. 2015;6:134–42.
with systemic chemotherapy is a promising method for downstag- 97. Rhee TK, Lewandowski RJ, Liu DM, Mulcahy MF, Takahashi G,
ing unresectable huge intrahepatic cholangiocarcinoma to surgical Hansen PD, et al. 90Y Radioembolization for metastatic neuroen-
treatment. Ann Surg Oncol. 2015;22:3102–8. docrine liver tumors: preliminary results from a multi-institutional
83. Boehm LM, Jayakrishnan TT, Miura JT, Zacharias AJ, Johnston experience. Ann Surg. 2008;247:1029–35.
FM, Turaga KK, et al. Comparative effectiveness of hepatic artery 98. Kennedy AS, Dezarn WA, McNeillie P, Coldwell D, Nutting C,
based therapies for unresectable intrahepatic cholangiocarcinoma. Carter D, et al. Radioembolization for unresectable neuroen-
J Surg Oncol. 2015;111:213–20. docrine hepatic metastases using resin 90Y-microspheres: early
84. Valle J, Wasan H, Palmer DH, Cunningham D, Anthoney A, results in patients. Am J Clin Oncol. 2008;31:271–9.
Maraveyas A, et al. Cisplatin plus gemcitabine versus gemcitabine 99. Giammarile F, Bodei L, Chiesa C, Flux G, Forrer F, Kraeber-
for biliary tract cancer. N Engl J Med. 2010;362:1273–81. Bodere F, et al. EANM procedure guideline for the treatment of
85. Welsh JS, Kennedy AS, Thomadsen B. Selective Internal liver cancer and liver metastases with intra-arterial radioactive
Radiation Therapy (SIRT) for liver metastases secondary to compounds. Eur J Nucl Med Mol Imaging. 2011;38:1393–406.
colorectal adenocarcinoma. Int J Radiat Oncol Biol Phys. 100. Shepherd FA, Rotstein LE, Houle S, Yip TC, Paul K, Sniderman
2006;66:S62–73. KW. A phase I dose escalation trial of yttrium-90 microspheres
86. Wong CY, Salem R, Raman S, Gates VL, Dworkin HJ. Evaluating in the treatment of primary hepatocellular carcinoma. Cancer.
90
Y-glass microsphere treatment response of unresectable colorec- 1992;70:2250–4.
tal liver metastases by [18F]FDG PET: a comparison with CT or 101. Yan ZP, Lin G, Zhao HY, Dong YH. An experimental study and
MRI. Eur J Nucl Med Mol Imaging. 2002;29:815–20. clinical pilot trials on yttrium-90 glass microspheres through
87. Van den Eynde M, Flamen P, El Nakadi I, Liberale G, Delatte the hepatic artery for treatment of primary liver cancer. Cancer.
P, Larsimont D, et al. Inducing resectability of chemotherapy- 1993;72:3210–5.
refractory colorectal liver metastasis by radioembolization with 102. Lau WY, Leung WT, Ho S, Cotton LA, Ensminger WD, Shapiro
yttrium-90 microspheres. Clin Nucl Med. 2008;33:697–9. B. Treatment of inoperable hepatocellular carcinoma with intrahe-
88. Gray B, Van Hazel G, Hope M, Burton M, Moroz P, Anderson J, patic arterial yttrium-90 microspheres: a phase I and II study. Br J
et al. Randomised trial of SIR-Spheres plus chemotherapy vs. che- Cancer. 1994;70:994–9.
motherapy alone for treating patients with liver metastases from 103. Andrews JC, Walker SC, Ackermann RJ, Cotton LA, Ensminger
primary large bowel cancer. Ann Oncol. 2001;12:1711–20. WD, Shapiro B. Hepatic radioembolization with yttrium-90 con-
89. Goin JE, Dancey JE, Hermann GA, Sickles CJ, Roberts CA, taining glass microspheres: preliminary results and clinical fol-
MacDonald JS. Treatment of unresectable metastatic colorectal low-up. J Nucl Med. 1994;35:1637–44.
carcinoma to the liver with intrahepatic Y-90 microspheres: a 104. Mahnken AH. Current status of transarterial radioembolization.
dose-ranging study. World J Nucl Med. 2003;2:216–25. World J Radiol. 2016;8:449–59.
90. Van Hazel G, Blackwell A, Anderson J, Price D, Moroz P, Bower 105. Nijsen JFW, Seppenwoolde JH, Havenith T, Bos C, Bakker CJG,
G, et al. Randomised phase 2 trial of SIR-Spheres plus fluoro- Van het Schip AD. Liver tumors: MR imaging of radioactive
uracil/leucovorin chemotherapy versus fluorouracil/leucovorin- holmium microspheres – Phantom and rabbit study. Radiology.
chemotherapy alone in advanced colorectal cancer. J Surg Oncol. 2004;231:491–9.
2004;88:78–85. 106. Seppenwoolde JH, Nijsen JFW, Bartels LW, Zielhuis SW, Van het
91. Sharma RA, Van Hazel GA, Morgan B, Berry DP, Blanshard K, Schip AD, Bakker CJ. Internal radiation therapy of liver tumors:
Price D, Bower G, et al. Radioembolization of liver metastases qualitative and quantitative magnetic resonance imaging of the
from colorectal cancer using yttrium-90 microspheres with con- biodistribution of holmium-loaded microspheres in animal mod-
comitant systemic oxaliplatin, fluorouracil, and leucovorin che- els. Magn Reson Med. 2004;53:76–84.
motherapy. J Clin Oncol. 2007;25:1099–106. 107. De Wit TC, Xiao J, Nijsen JF, Van het Schip FD, Staelens SG, Van
92. van Hazel GA, Pavlakis N, Goldstein D, Olver IN, Tapner MJ, Rijk PP, Beekman FJ. Hybrid scatter correction applied to quanti-
Price D, et al. Treatment of fluorouracil-refractory patients with tative holmium-166 SPECT. Phys Med Biol. 2006;51:4773–87.
108. Zielhuis SW, Nijsen JFW, De Roos R, Krijger GC, Van Rijk PP, loaded poly(L-lactic acid) microspheres in rats. Biomaterials.
Hennink WE, et al. Production of GMP-grade radioactive hol- 2007;28:4591–9.
mium loaded poly(l-lactic acid) microspheres for clinical applica- 114. Vente MAD, Nijsen JFW, De Wit TC, Seppenwoolde JH, Krijger
tion. Int J Pharm. 2006;311:69–74. GC, Seevinck PR, et al. Clinical effects of transcatheter hepatic
109. Zielhuis SW, Nijsen JFW, Krijger GC, Van het Schip AD, Hennink arterial embolization with holmium-166 poly(L-lactic acid)
WE. Holmium-loaded poly(L-lactic acid) microspheres: in vitro microspheres in healthy pigs. Eur J Nucl Med Mol Imaging.
degradation study. Biomacromolecules. 2006;7:2217–23. 2008;35:1259–71.
110. Bult W, Vente MA, Zonnenberg BA, Van het Schip AD, Nijsen 115. Smits ML, Nijsen JF, van den Bosch MA, Lam MG, Vente MA,
JF. Microsphere radioembolization of liver malignancies: current Mali WP, et al. Holmium-166 radioembolisation in patients with
developments. Q J Nucl Med Mol Imaging. 2009;53:325–35. unresectable, chemorefractory liver metastases (HEPAR trial): a
111. Elschot M, Nijsen JF, Dam AJ, de Jong HW. Quantitative evalu- phase 1, dose-escalation study. Lancet Oncol. 2012;13:1025–34.
ation of scintillation camera imaging characteristics of isotopes 116. Braat AJAT, Prince JF, van Rooij R, Bruijnen RCG, van den Bosch
used in liver radioembolization. PLoS One. 2011;6:e26174. MAAJ, Lam MGEH. Safety analysis of holmium-166 micro-
https://doi.org/10.1371/journal.pone.0026174.. Epub 2011 Nov 3 sphere scout dose imaging during radioembolisation work-up: a
112. Nijsen JFW, Rook D, Brandt CJWM, Meijer R, Dullens H, cohort study. Eur Radiol. 2018;28:920–8.
Zonnenberg BA, et al. Targeting of liver tumour in rats by selec- 117. Prince JF, van den Bosch MAAJ, Nijsen JFW, Smits MLJ, van den
tive delivery of holmium-166 loaded microspheres: a biodistribu- Hoven AF et al. Efficacy of radioembolization with holmium-166
tion study. Eur J Nucl Med. 2001;28:743–9. microspheres in salvage patients with liver metastases: a phase
113. Zielhuis SW, Nijsen JFW, Seppenwoolde JH, Bakker CJG, 2 study. J Nucl Med. 2017. pii: jnumed.117.197194. https://doi.
Krijger GC, Dullens HF, et al. Long-term toxicity of holmium- org/10.2967/jnumed.117.197194. Epub ahead of print
Hybrid Imaging for Gynecologic
Malignancies 34
Elisa Lodi Rizzini, Elena Tabacchi, and Cristina Nanni
Contents
34.1.1 Cervical Cancer 882
34.1.2 Endometrial Cancer 882
34.1.3 Ovarian Cancer 882
34.2 [18F]FDG PET/CT in Gynecological Imaging 883
34.2.1 Procedure 883
34.2.2 [18F]FDG PET/CT in Cervical Cancer........ 884
34.2.3 [18F]FDG PET/CT in Endometrial Cancer.. 887
34.2.4 [18F]FDG PET/CT in Epithelial Ovarian Cancer 888
34.3 PET/MR Imaging in Gynecologic Malignancies 890
34.3.1 Procedure 891
34.3.2 [18F]FDG PET/MR in Cervical Cancer....... 891
34.3.3 [18F]FDG PET/MR in Endometrial Cancer 893
34.3.4 [18F]FDG PET/MR in Epithelial Ovarian Cancer 893
34.4 Uncommon Gynecologic Tumors 894
34.4.1 Vulvar Cancer 894
34.4.2 Vaginal Cancer 894
References 895
Learning Objectives malignancies (initial staging, assessment of response to

• Summarize the main features regarding epidemiology, therapy, suspected tumor recurrence posttreatment).
etiology, histopathology, and the most common clinical • Understand the pathophysiologic basis for the use of
presentation of gynecologic malignancies, including PET/MR imaging with [18F]FDG in patients with gyneco-
uncommon gynecologic tumors. logic malignancies.
• Understand the pathophysiologic basis for the use of • Summarize the basic modalities of performing a PET/
PET/CT imaging with [18F]FDG in patients with gyneco- MR scan with [18F]FDG PET/CT in patients with gyne-
logic malignancies. cologic malignancies.
• Summarize the basic modalities of performing a PET/CT • List the advantages and indications for performing a PET/
scan with [18F]FDG PET/CT in patients with gynecologic MR scan with [18F]FDG in patients with gynecologic
malignancies. malignancies.
• List the established indications for performing a PET/CT
scan with [18F]FDG in patients with gynecologic
E. Lodi Rizzini · E. Tabacchi · C. Nanni (*)

Metropolitan Nuclear Medicine,
AOU S. Orsola-Malpighi, Bologna, Italy
e-mail: cristina.nanni@aosp.bo.it

882 E. Lodi Rizzini et al.
34.1 Introductory Background Although all types of high-grade endometrial carcino-

mas (including grade 3 endometrioid, serous, and clear cell
34.1.1 Cervical Cancer adenocarcinomas) have poor prognosis, they respond to
adjuvant therapy differently [3]. Since standard treatment is
Cervical cancer is the third most common neoplasm in hysterectomy with bilateral salpingo-oophorectomy, these
women [1]. Squamous cell carcinoma is found in about tumors have always been stratified according to the FIGO
three-quarters of the cases; the remaining cases are adeno- classification which is based on intraoperative findings;
carcinoma subtypes. The most common clinical symptom although revised in 2010, this classification does not include
is postcoital bleeding, followed by metrorrhagia or menor- imaging parameters for staging [6]. On the contrary, the
rhagia with subsequent chronic anemia; a combination of NCCN guidelines recommend [18F]FDG PET/CT imaging
rectal or bladder symptoms can also be present [2]. Cervical for evaluation of extrauterine disease in high-risk patients
cancer generally spreads progressively from local adjacent with endometrial cancer, with the purpose of better patient
structure and regional lymph nodes to the pelvic, para-aortic, selection and surgical planning (www.nccn.org/profession-
and supraclavicular lymph nodes. Dissemination to distant als/physician_gls/pdf/uterine.pdf).
organs occurs late in the course of disease [2]. At the time of In advanced disease, where the treatment of choice is
initial diagnosis, the most important prognostic parameters chemotherapy (alone or combined with radiation therapy),
include tumor size, parametrial invasion, and lymph node surgery has only a palliative role.
involvement.
34.1.3 Ovarian Cancer

34.1.2 Endometrial Cancer
In developed countries, ovarian cancer is one of the three most
Endometrial cancer is the sixth most common malignancy common gynecologic malignancies and the most lethal. The
among women worldwide and is the most commonly incidence of ovarian cancer increases with age. Overall, 90%
gynecologic malignancy in industrialized countries [3, 4], a of ovarian cancers are of epithelial origin and include serous,
factor that is usually correlated to obesity and high fat con- mucinous, and endometrioid histological subtypes. Less com-
sumption. A major risk factor is long-lasting endogenous mon histologic subtypes include clear cell, transitional cell,
or exogenous hyperestrogenism (nulliparity, anovulation, mixed epithelial tumors, and the undifferentiated subtype [7].
premature menarche and late menopause, polycystic ovary Ovarian cancer spreads to the inguinal, pelvic, and para-aortic
syndrome, tamoxifen therapy). Other comorbidities include lymph nodes, followed by involvement of mediastinal lymph
arterial hypertension, diabetes mellitus, and genetic predis- nodes. Distant metastasis occurs mainly to the pleura, liver,
position (e.g., the Lynch syndrome, associated with colorec- lungs, spleen, and bone marrow [8]. Since early ovarian can-
tal cancer). Over 90% of the cases occur in women older than cer is clinically silent, it often presents in an advanced stage.
50 years of age, with a median age of 63 years. There is no Cytoreductive surgery and postoperative adjuvant che-
routine screening test for endometrial cancer, but most cases motherapy are standard therapy. When surgery is not pos-
are diagnosed early because the earliest symptom is vaginal sible because of advanced disease at diagnosis, patients often
bleeding. This results in high survival rates, 82% at 5 years benefit of neoadjuvant chemotherapy, which can result in
and 79% at 10 years, respectively [5]. complete cytoreduction of all macroscopic visible disease
The two main histologic types of endometrial carcinoma [9]. The recurrence rate of ovarian cancer at 2 years after
are estrogen-dependent endometrioid carcinomas (Type 1) primary treatment varies according to the stage of disease at
and estrogen-independent non-endometrioid carcinomas diagnosis. The tumor recurs in 75–80% of patients with stage
(Type 2). Type 1 tumors constitute 80% of endometrial car- III and in 90–95% of patients with stage IV disease [10].
cinomas. Type 1 tumors are detected earlier and have bet- Patients with ovarian cancer are followed with the tumor-
ter prognosis than Type 2 tumors. In the well-differentiated associated marker carbohydrate antigen CA-125, which
form, they are composed of glands that resemble those of has an overall 94% sensitivity and 82% specificity [11].
the normal endometrium and can be associated with, or pre- However, in premenopausal women specificity decreases to
ceded by, endometrial hyperplasia. Genetic abnormalities about 60% because other conditions (such as endometriosis,
include microsatellite instability and mutations of the PTEN, cystadenoma, pelvic inflammatory disease, and peritoneal
PIK3CA, K-Ras, and β-catenin genes. Serous carcinomas dissemination of non-ovarian cancers) can also induce high
are considered the prototype of Type 2 endometrial cancer, serum levels of CA-125 [12]. In combination with imaging,
while clear cell carcinomas constitute a rare and heteroge- evaluation of the serum levels of CA-125 actually plays an
neous group of tumors with intermediate features between important role in assessing disease progression and response
Type 1 and Type 2. to therapy.
34 Hybrid Imaging for Gynecologic Malignancies 883
Similarly as for endometrial cancer, the FIGO staging to a cold environment (or simply to stress); this occurrence
system for ovarian cancer is based on surgical findings only. can be ascertained by integrating the metabolic informa-
However, intraoperative staging can underestimate early tion provided by PET with the anatomic information (low-
peritoneal dissemination or metastasis to extra-abdominal attenuation focus) provided by CT. Diffusely increased [18F]
sites, like the pleural space. A major favorable prognostic FDG uptake in skeletal muscles can suggest a postprandial
factor is resection of all visible tumor at the time of initial state in nondiabetic patients in response to insulin; if tracer
surgery. Therefore, presurgical diagnostic imaging plays a uptake is increased in distinct muscular groups, there could
central role in the management of ovarian cancer, as recom- be a recent history of exertion of those muscles [15]. Focally
mended by the NCCN guidelines for management of ovar- increased [18F]FDG uptake can be seen in nonmalignant
ian carcinoma [13]. Ultrasonography (US), CT, and MRI are conditions, such as infections, inflammation, benign tumor,
widely employed especially in initial assessment and charac- imaging artifacts, and misregistration of the PET and CT
terization of the ovarian mass. Hybrid [18F]FDG PET/CT or components. Since inflammation can be induced by surgery,
PET/MR improve tumor characterization and help identify radiotherapy, or chemotherapy, posttreatment PET scans
local extent of disease, nodal involvement, and metastases. should be performed after an interval of at least 2 weeks after
surgery or chemotherapy and 6–12 weeks after completion
of radiotherapy [16].
Besides benign tumors and infection/inflammation,
Key Learning Points benign conditions that can cause increased [18F]FDG uptake
• Gynecologic malignancies are relatively common include uterine fibroids, adenomyosis, endometriosis, and
diseases affecting the uterus (cervical epithelium or physiological cyclic activity of the uterus or ovary in pre-
endometrium) and the ovary. menopausal women [17]. Focal tracer accumulation in a
• These malignancies can be highly aggressive and ureter may mimic a pathological lymph node, or diaphragm
require treatment with chemotherapy, surgery, and motion and bowel peristalsis could result in ambiguous foci
radiation therapy—combined in different ways with increased [18F]FDG accumulation; when evaluating
depending on staging at diagnosis. patients with known or suspected gynecologic malignan-
cies, PET scanning in a caudal-cranial direction reduces
these artifacts. Furthermore, attenuation correction with CT
can cause artifacts at sites where CT attenuation differs sub-
34.2 [18F]FDG PET/CT in Gynecological stantially from bone or soft tissue (e.g., in presence of high
Imaging concentration of intravenous contrast or in case of metallic
implants) [18]. Diffuse [18F]FDG uptake can be seen in the
The CT component of integrated PET/CT imaging allows large and small bowel of diabetic patients on oral metformin
both accurate attenuation correction and anatomic correla- therapy [19].
tion of the metabolic data. The standardized uptake value
(SUV) of lesions is extremely useful in the characterization
of lesions. Performing the CT with X-ray contrast agent pro- Key Learning Point
vides a better assessment of pelvic anatomy and should be • [18F]FDG PET/CT is a functional imaging proce-
performed if possible. [18F]FDG PET/CT does not have an dure that can be clinically useful in gynecologic
important role in the evaluation of the primary tumor, mostly malignancies.
due to limited soft tissue contrast and spatial resolution
(≥4 mm), which often results in false-negative findings for
lesions <1 cm in size [14]. Other possible causes of false-
negative results of PET imaging include low-glycometabolic 34.2.1 Procedure
cancer lesions (such as grade I endometrial cancer or well-
differentiated mucinous ovarian cystadenocarcinoma) and Imaging with the hybrid technique PET/CT often combines
small pelvic tumors adjacent to sites of physiologic [18F] the anatomic information of a low-dose CT scan with the
FDG excretion (bladder or ureters). Peritoneal carcinomato- metabolic information derived from the tracer employed for
sis (a form of local metastatic dissemination often occurring the PET component of the scan, most frequently [18F]FDG
in patients with some gynecologic cancers) can be missed by for gynecologic malignancies—which provides a map of the
PET imaging in the early phases of tumor growth. tissue consumption of glucose. Assuming that the glucose
On the other hand, [18F]FDG PET/CT imaging can requirements of cancer cells are greater than those of their
result in false-positive findings, for example, in the well- normal counterparts, PET will produce a metabolic imaging
known case glycometabolic activation of the brown fat due map of the cancer cells present in tissues.
The [18F]FDG PET/CT is typically acquired 60–90 min choice of treatment [24]. Therefore, surgical staging is rec-
after an i.v. bolus injection of the tracer. The field of view ommended in patients with advanced disease for identifying
commonly includes an area of the body between the skull pelvic and para-aortic lymph node metastasis. Instead, this
base and the proximal thighs. The patient’s arms are usually procedure is not recommended in patients with early-stage
kept raised in order to minimize scatter when imaging the disease, since it is frequently associated with serious morbid-
torso. Depending on the PET scanner employed and on the ity and complications.
amount of [18F]FDG injected, the typical acquisition time for The latest NCCN guidelines recommend imaging, such as
whole body PET varies between 15–45 min and 2–5 min per [18F]FDG PET/CT (Fig. 34.1), for pretreatment assessment in
bed position. patients with cervical cancer and clinical stage IB2 or higher,
Though protocols might differ from institution to institu- as also when para-aortic lymph node disease is found at sur-
tion, fasting (except for water) is generally required for at gical staging or when invasive cervical cancer is found inci-
least 6 h before [18F]FDG administration, and patients should dentally at hysterectomy performed for other reasons [25].
avoid vigorous exercise for 24 h before the scan. The abso- Similarly, the American College of Radiology (ACR) recom-
lute cutoff of glucose blood level at the time of [18F]FDG mends pretreatment imaging with either MR or [18F]FDG
injection is 250 mg/dL, although lower values are routinely PET/CT for staging patients with stage IB1 or higher [26].
applied for some protocols. Since therapy with metformin [18F]FDG PET/CT performs better than conventional imag-
can significantly increase [18F]FDG uptake in the bowel, the ing for detecting local or extra-pelvic disease in patients in
drug should be withheld for at least 48 h before the proce- whom conventional imaging has been equivocal or negative.
dure, if clinically appropriate. In a meta-analysis of 41 studies, either stand-alone PET or
After [18F]FDG administration, the patient should be in PET/CT showed better diagnostic accuracy (with 82% sen-
a quiet, dimly lit setting, not speaking (in order to minimize sitivity and 95% specificity) for the detection of metastatic
uptake in laryngeal muscles), and consuming oral contrast lymph nodes than either MR or CT alone [27]. Another meta-
material. The patient should also be kept warm during the analysis of 72 studies found a higher positive probability ratio
uptake period, in order to minimize uptake in the brown fat. for PET or [18F]FDG PET/CT for the detection of metastatic
Occasionally, the patient requires sedation if suffering from lymph nodes (15.3) than either MR (6.4) or CT (4.3) [28].
anxiety attacks or claustrophobia [20]. Whereas, for local tumor staging, [18F]FDG PET/CT per-
forms worse than MR, which has superior soft tissue contrast
and spatial resolution [29, 30]. Furthermore, according to a
meta-analysis of ten studies, the overall diagnostic perfor-
Key Learning Points
mance of [18F]FDG PET/CT in patients with early-stage cer-
• Fasting is required for performing an [18F]FDG
vical cancer is suboptimal, and its efficacy is related to the
PET/CT.
initial probability of lymph node metastasis [31].
• Diabetic patients are required to discontinue met-
Detection of extra-pelvic metastasis in patients with strong
formin therapy for at least 48 h before the scan.
clinical suspicion or tissue diagnosis of pelvic recurrence is
• Pregnancy is an absolute contraindication.
very important, because this occurrence would preclude pel-
vic exenteration surgery—which is associated with a chance
of about one-third of long-term disease-free survival [32].
34.2.2 [18F]FDG PET/CT in Cervical Cancer The latest NCCN guidelines suggest performing an [18F]
FDG PET/CT 3–6 months after chemoradiation for locally
Cervical cancer is staged clinically according to the FIGO advanced cervical cancer, to detect early or asymptomatic
criteria, although many studies have demonstrated a dis- persistent/recurrent disease (Fig. 34.2) [25]. The Society of
crepancy between clinical and surgical staging, which varies Gynecologic Oncology (SGO) recommends [18F]FDG PET/
from 25% of the patients for early-stage disease to 65–90% CT imaging only if recurrence is suspected, stating that the
in more advanced tumors [21]. This is due to the fact that scan can be useful for detecting extra-pelvic disease before
FIGO staging in based on physical examination, cystoscopy, radiation or exenteration [33]. Instead, [18F]FDG PET/CT is
proctoscopy, colposcopy, and biopsy. This approach does not not recommended for routine surveillance without clinical
include many important prognostic factors (such as tumor suspicion or histologically proven local recurrence [34].
size, grade, local or retroperitoneal lymph node involvement) [18F]FDG PET/CT is sensitive for detection of recurrent
that can result in a different prognosis and require different disease: a meta-analysis of 20 studies found 87% sensitivity
clinical management [22, 23]. About one-third of patients and 97% specificity for distant metastasis and 82% sensi-
with locally advanced cervical cancer have metastatic dis- tivity with 98% specificity for local and regional recurrence
ease to the para-aortic lymph nodes, a factor that affects the [35]. Another meta-analysis of 23 studies demonstrated that
Fig. 34.1 Cervical cancer. Staging [18F]FDG PET/CT performed for staging in a 55-year-old woman with spinocellular carcinoma. [18F]FDG
PET/CT shows increased tracer uptake (SUVmax 12.9) in the cervical lesion, adjacent to the bladder
[18F]FDG PET/CT has better sensitivity for recurrence than are not surgically staged, considering that [18F]FDG PET/
PET and CT performed separately [36]. Patients with new, CT may detect metastatic lymph nodes not detected by CT
residual, or no disease on posttreatment imaging have 5-year alone (Figs. 34.3 and 34.4) [25]. [18F]FDG PET/CT is used
survival rates of 0%, 46%, and 92%, respectively [37]. to accurately define the treatment area [42], to define targets
The usefulness of [18F]FDG PET/CT as a prognostic with intensity-modulated radiotherapy planning (thus mini-
indicator for cervical cancer and for the need of additional mizing the radiation dose to the bone marrow/other organs),
treatment has also been explored [38]. A meta-analysis of 14 and to perform adaptative brachytherapy, by adjusting the
studies showed that a higher SUVmax in the primary lesion radiation field according to changes in the findings of serial
or in metastatic pelvic/para-aortic nodes (mostly para-aortic) [18F]FDG PET/CT scans [28].
increases the risk of adverse events or death [39].
Regarding the use of [18F]FDG PET/CT to assess
response to treatment, a meta-analysis of 16 studies found
that a positive posttreatment [18F]FDG study (either PET/ Key Learning Points
CT or stand-alone PET) is associated with poor survival and • [18F]FDG PET/CT is recommended to stage the dis-
metabolic response to therapy correlated with survival [40, ease for its high sensitivity in detecting lymph node
41]. Patients with no tumor uptake after therapy have better metastasis, especially in extra-pelvic sites.
survival (86–99%) than patients with increased or residual • [18F]FDG PET/CT is also indicated for restaging
uptake after therapy (0–18%) [22]. the disease and for assessing response to therapy.
According to current NCCN guidelines, [18F]FDG PET/ • Incomplete metabolic response at [18F]FDG PET is
CT imaging is useful for defining nodal volume to be associated with poor prognosis.
included in the radiation treatment plan in patients who
Fig. 34.2 Cervical cancer. Restaging [18F]FDG PET/CT performed for radiotherapy, no tracer uptake is detectable at the uterine cervix, indi-
assessing tumor response to therapy in the same patient as in Fig. 34.1 cating complete response
therapy assessment. Three months after completion of chemotherapy/
Fig. 34.3 Cervical cancer. [18F]FDG PET/CT performed for radiotherapy planning in an 85-year-old woman with cervical cancer. [18F]FDG PET/
CT shows markedly increased tracer uptake (SUVmax 20.4) at the uterine cervix, ideal to guide the radiotherapy planning
Fig. 34.4 Cervical cancer. [18F]FDG PET/CT performed in a 40-year- increased tracer uptake within the cervical lesion (SUVmax 35.6), in an
old woman for defining the radiotherapy target volume: primary lesion, external right iliac lymph node (SUVmax 10.4) and in at least two perito-
lymph nodes, and peritoneal implants. [18F]FDG PET/CT shows neal nodules (SUVmax 5.5)
34.2.3 [18F]FDG PET/CT in Endometrial Cancer CT. Nevertheless, the consistently high specificity justifies
the use of [18F]FDG PET/CT for initial staging of patients
In patients with endometrial cancer, [18F]FDG PET/CT has with endometrial cancer, with the main purpose of sparing
82% sensitivity and 90% specificity for detecting the primary unnecessary lymphadenectomy (minimizing morbidity and
lesion (see Fig. 34.5); the corresponding values for detecting complications of this surgical procedure) to patients who
lymph node metastasis are 72% and 93%, rising to 96% and will not benefit from lymphadenectomy because they already
95%, respectively, for distant metastasis [43]. Suboptimal have advanced disease [44].
sensitivity in detecting the primary lesion can be related to In patients with early-stage endometrial cancer, [18F]FDG
interference in the interpretation of images induced by meta- PET/CT is indicated when histology of the primary tumor
bolic changes associated with the ovulatory and menstrual exhibits high-risk features. The high diagnostic performance
phase in premenopausal women. Whereas, the relatively low of [18F]FDG PET/CT in the detection of locoregional and
sensitivity for the detection of metastatic lymph nodes can distant metastasis appears to be influenced by the site of
be due to the limited spatial resolution of [18F]FDG PET/ recurrence; in fact, the diagnostic performance of [18F]FDG
Fig. 34.5 Endometrial cancer. [18F]FDG PET/CT performed for staging in a 55-year-old woman with G2 endometrial carcinoma. [18F]FDG PET/
CT shows markedly increased tracer uptake in the uterus only (SUVmax 22.8)
PET/CT is higher for pelvic lymph node metastasis than for the scan is scheduled in days at some distance from ovulation
para-aortic lymph node metastasis [45]. or menstrual cycle (Fig. 34.7).
Results of pilot studies are promising regarding the use- Sensitivity, specificity, and accuracy of [18F]FDG PET/
fulness of [18F]FDG PET/CT for assessing response to ther- CT in the characterization of a pelvic mass were found to be
apy in patients with endometrial cancer [46, 47] (Fig. 34.6). 87%, 100%, and 92%, respectively, versus 90%, 61%, and
80% for transvaginal US [49]. Furthermore, [18F]FDG PET/
CT is more accurate than the other imaging modalities for
identifying metastasis in regional lymph nodes (including
Key Learning Points
normal-sized lymph nodes), in retroperitoneal lymph nodes,
• Endometrial cancer can be staged with [18F]FDG
as well as in distant metastases [50]. Detection of peritoneal
PET/CT, which has sufficient diagnostic accuracy
metastatic disease with [18F]FDG PET/CT is more sensitive
for detecting the primary lesion, and especially dis-
than CT, especially for lesions located in the right upper
tant metastasis.
abdomen and small bowel mesentery (Fig. 34.8).
• Patients with a negative PET scan for lymph node
At initial staging, the SUVmax of the primary cancer
metastasis can be considered for a less invasive
derived from [18F]FDG PET/CT is associated with FIGO
surgery.
stage and histology and is a predictor of overall survival
(OS) and disease-free survival (DFS). Patients with a low
primary tumor SUVmax had longer OS rate and DSF rate than
34.2.4 [18F]FDG PET/CT in Epithelial Ovarian patients with higher tumor SUVmax [51]. Furthermore, more
Cancer complex metabolic parameters such as pretreatment meta-
bolic volume (MTV) and total lesion glycolysis (TLG), mea-
[18F]FDG PET/CT has been shown to be superior to transvag- sured with [18F]FDG PET/CT, are inversely correlated with
inal US, CT, and MRI for staging and presurgical e valuation the progression-free interval [52].
of disease in patients with epithelial ovarian cancer [48]. In patients with locally advanced epithelial cancer treated
False-positive findings of [18F]FDG PET/CT are reduced if with neoadjuvant therapy, [18F]FDG PET/CT accurately
Fig. 34.6 Endometrial cancer. Radiotherapy assessment. An 83-year- had been performed 2 years earlier, and the findings at MR imaging
old woman, treated 2 years before with brachytherapy for uterine can- were equivocal. [18F]FDG PET/CT shows increased tracer uptake
cer, underwent [18F]FDG PET/CT performed in an 83-year-old woman (SUVmax 17) in the uterine body, consistent with persistent active
for assessing the long-term outcome after radiotherapy; brachytherapy disease
Fig. 34.7 Characterization of an ovarian mass. [18F]FDG PET/CT performed in a 55-year-old woman with equivocal findings at conventional
imaging. Increased tracer uptake (SUVmax 8.8) at the right ovary is consistent with malignancy
Fig. 34.8 Ovarian cancer. [18F]FDG PET/CT performed for staging in sistent with malignancy (SUVmax 5.8), associated with less intense
a 67-year-old woman with recent diagnosis of cancer of the left ovary. uptake in the right ovary (SUVmax 3.4), doubtful for malignancy
[18F]FDG PET/CT shows increased tracer uptake in the left ovary con-
discriminates responders from nonresponders [53]. The

superiority of [18F]FDG PET/CT over other imaging modal- Key Learning Points
ities is especially obvious in the identification of residual • [18F]FDG PET/CT is a diagnostic modality for ini-
viable tumor after therapy [53]. In association with serum tial staging of patients with epithelial ovarian
CA-125, [18F]FDG PET/CT allows early detection of recur- cancer.
rent disease. The sensitivity for detecting recurrent disease • SUVmax of the primary tumor at diagnosis is a strong
is particularly useful in asymptomatic patients who could prognostic index.
benefit from second-line surgery and chemo- or radiation • [18F]FDG PET/CT is an optimal diagnostic to iden-
therapy. The sensitivity, specificity, and accuracy of [18F] tify nonresponder patients and the site of relapse in
FDG PET/CT for detecting recurrent ovarian cancer are case of rising serum marker levels and to define
88.2–98.3%, 71.4–100%, and 71.4–97%, respectively [54, optimal radiotherapy plans.
55]. Based on the findings of [18F]FDG PET/CT, the thera-
peutic strategy can change in more than 60% of the patients
with ovarian cancer [56], particularly to select patients with
gross lesions as candidates for cytoreductive surgery [57]. 34.3 P
ET/MR Imaging in Gynecologic
Intensity-modulated radiotherapy (IMRT) can be guided Malignancies
by the [18F]FDG PET/CT findings, since the scan can rede-
fine the extent of metastasis and detect additional lymph Thanks to the higher soft tissue contrast, higher resolution,
node metastasis, thus leading to modifications in gross target multiplanar capabilities, and multimodality features, MR
volume (GTV) delineation in 35.7% of patients. [18F]FDG imaging is useful in the management of patients with cervi-
PET/CT-guided IMRT improves tumor response and 3-year cal and ovarian cancer, while its role in patients with endo-
overall survival compared with CT-guided IMRT [58]. metrial carcinoma is growing. Newer dynamic functional
imaging techniques such as DCE MR and DWI MR provide 34.3.1 Procedure

biological information enabling to better characterize tumors
at initial staging, to assess response to treatment, to detect PET/MR scans are performed on a sequential or simultane-
residual or recurrent disease, and to provide important bio- ous (3T) PET/MR imaging system. In sequential PET/MR
logical information about tumor oxygenation, perfusion, and systems, spatially separate individual PET and MR scanners
microstructure. In addition, quantitative parameters derived are connected by a common moving table to reduce changes
from functional MR imaging can potentially constitute bio- in patient’s positioning between the two scans. Instead, in
markers for assessment of tumor response to therapy, espe- simultaneous PET/MR systems, there is a single gantry
cially if combined with PET-derived parameters. for MR and PET scanners. The sequential scanner is more
An inverse correlation between the main metabolic [18F] economical, but it involves artifacts due to misregistration.
FDG-PET-derived parameter (SUVmax) and the MR-derived With both systems, the most critical limitation is the accu-
minimum apparent diffusion coefficient (ADCmin) has been racy of the MR-derived attenuation correction algorithms for
found in cervical and endometrial cancers [59–61]; ADCmin is PET. In fact, unlike PET/CT, the MR imaging signal acquired
a quantitative parameter derived from the exponential attenu- with a PET/MR scanner correlates with proton density and
ation of signal between at least two acquisitions with differ- tissue relaxation properties, and not with electron density.
ent amounts of diffusion weighting in DWI RM sequences. Possible solutions for this problem include the segmentation
Malignant tissues have low ADC values, which normalize in approach to assign MR data into tissue classes or the use of
response to effective therapy. This pattern of change becomes PET emission data and MR anatomic information to create
therefore a predictive biomarker of tumor response, not attenuation maps [5]. Despite the technological advances, in
unlike the SUV variations in case of favorable tumor response the simultaneous PET/MR scanners, there is still a lack of
to treatment. Since both biomarkers (ADCmin and SUV) are MR signal in cortical bone that can result in a lower SUV (by
functional in nature, changes in their values can precede 10–20%) [62]; furthermore, the low MR signal in the lung
reduction in tumor size, therefore allowing better assessment can lead to underestimate the pulmonary parenchymal atten-
of individual response to treatment or earlier identification of uation. However, the better soft tissue contrast generated by
residual/recurrent disease than morphological imaging pro- MR-specific sequences may be of much greater value than
vided by, e.g., CT. the possible problems originated by motion, especially when
PET/MR and [18F]FDG PET/CT provide complementary evaluating pelvic structures.
information in the evaluation of patients with gynecologic Even if protocols can change from center to center, an
cancers; nevertheless, further evidence is needed to better integrated PET/MR scan is generally based on acquisition
define and integrate the three fundamental sets of imaging of 3–5 bed positions, with 2–3 min acquisition time per
data, functional (MR), morphologic (CT), and molecular bed position. Simultaneously with the PET acquisition,
(PET), respectively. Since MR is superior to CT for tissue two MR sequences are acquired for each bed position, and
characterization, it can be speculated that [18F]FDG PET/ the total duration of a PET/MR scan is about 20 min, in
MR is superior to [18F]FDG PET/CT for detecting small 4–5 steps from skull vertex to thighs. Depending to the
metastatic lymph nodes and for distinguishing such lymph clinical indication, a dedicated study with high-resolution
nodes from anatomic structures with high physiologic tracer and contrast-enhanced (CE) MR is usually obtained. The
uptake/accumulation, such as the bowel loops, ovaries, and patient’s preparation needed is the same as for the [18F]
ureters. Choi et al. demonstrated a higher diagnostic perfor- FDG PET/CT procedure.
mance for fused [18F]FDG PET/MR than for [18F]FDG PET/
CT in patients with cervical cancer, but it is still unclear
whether the slight improvement in diagnostic performance
Key Learning Points
achieved by the combined MR imaging data significantly
• PET and MR can be simultaneously acquired.
improves patient’s management [27].
• Problems in image processing can derive from
attenuation correction based on MR data and from
the long time needed for image acquisition.
Key Learning Points
• The role of MR imaging is crucial in gynecological
malignancies, especially for the evaluation of 34.3.2 [18F]FDG PET/MR in Cervical Cancer
parameter T.
• The combination of PET and MR data could be the Since treatment of cervical cancer differs for different
future of imaging in patients with gynecologic stages of the disease, imaging plays an important role in
malignancies. defining the most appropriate management. According to
NCCN guidelines, imaging with CT, PET/CT, or MR is
recommended for initial evaluation of stage IB2 cervical and metastatic lymph nodes in premenopausal women.
cancer or higher, while it is optional for stage IB1 or lower. Furthermore, the improved spatial coregistration can mini-
For detecting distant metastasis and/or local recurrence, the mize the uncertainties about [18F]FDG uptake in the bowel
NCCN guidelines recommend imaging with specific imag- that could hide a nodal uptake in the pelvis of patients with
ing modalities as follows: CT-based treatment planning for minimal intrinsic body fat [68].
external beam radiotherapy, MR imaging for identifying Depending on histology and degree of tumor differen-
soft tissue and parametrial involvement, and PET for lymph tiation, there are different primary tumors that can exhibit
node involvement. different levels of SUVmax; the highest SUVmax is generally
As mentioned above, the use of [18F]FDG PET/CT observed in squamous cell carcinoma and poorly differen-
is growing in the clinical practice as an alternative to tiated tumors [69]. With integrated simultaneous [18F]FDG
lymph node dissection, as well as for determining tumor PET/MR imaging, an inverse correlation has been found
volume and for prognostic purposes [63]. Based on the between SUVmax and the minimum apparent diffusion coef-
experience so far obtained in different types of cancer, ficient (ADCmin) in cervical and endometrial cancers [59,
PET/MR imaging has high diagnostic image quality for 60, 70]. The combination of metabolic information from
assessing the primary tumor, including especially tumor [18F]FDG PET and functional information from MR imag-
infiltration into adjacent soft tissues. Similar results have ing may therefore provide complementary prognostic data
been obtained on the high diagnostic value of PET/MR in patients with gynecologic cancers, concerning especially
imaging for staging cervical cancer in pretreated patients. lymph node involvement. Patients with PET-positive lymph
According to Grueneisen et al. [64], [18F]FDG PET/ nodes have significantly worse survival rates than those with
MR imaging detected all cervical lesions in 27 patients PET-negative lymph nodes [69, 71]. In advanced cervical
and correctly defined T stage in 85% of the cases, while cancer (FIGO stages III and IVA), distant metastases in the
lymph node metastasis was identified with 91% sensi- para-aortic and supraclavicular lymph nodes are common
tivity, 94% specificity, and 93% diagnostic accuracy. and increase the probability of tumor recurrence and lower
High-resolution T2-weighted imaging is best for depict- survival rates.
ing the cervical stromal and parametrial invasion, with In most patients with advanced stage cervical cancer, the
multiplanar T2-weighted imaging serving as the standard tumor recurs within 2 years from initial treatment, although
approach. Newer-generation sequences, such as isotropic approximately 70% of the patients survive the recurrence.
3D T2-weighted (SPACE) imaging, enable to obtain high- Early detection of disease recurrence is crucial to improve
quality multiplanar images with a single acquisition, in the survival rates—and [18F]FDG PET/CT has a pivotal role
which isotropic voxels can be reconstructed in multiple in this scenario [72]. Published data indicate that the diag-
planes without significant loss of resolution [65]. While nostic accuracy of [18F]FDG PET/MR imaging is similar to
it is well established that [18F]FDG PET/CT detects that of [18F]FDG PET/CT. In one study of 19 patients with
regional/distant metastasis in patients with advanced dis- recurrent cervical cancer, [18F]FDG PET/MR imaging had
ease significantly better than CT or MR imaging alone the same diagnostic accuracy as [18F]FDG PET/CT, detect-
[66, 67], the same degree of evidence has not yet been ing all local and distant recurrent disease sites and actually
attained for PET/MR—due to the fact that integrated PET/ providing greater interpretation confidence than [18F]FDG
MR scanners are still not widely available as PET/CT PET/CT (P < 0.001) [73].
scanners are. A retrospective study performed by Kitajima After treatment, most patients continue to have the same
et al. [68] showed that post-acquisition software fusion of symptoms they had at the time of diagnosis (such as pain
PET imaging with MR imaging resulted in the same diag- and discharge); therefore, clinical assessment of response to
nostic performance of contrast-enhanced [18F]FDG PET/ treatment can be challenging, as physical examination does
CT for the identification of lymph node metastasis, with not necessarily reflect the presence of active disease. Since
92.3% sensitivity, 88.2% specificity, and 90.0% accuracy it has been shown that [18F]FDG PET findings after primary
for both modalities. It is reasonable to assume that newer therapy are the best predictors for tumor recurrence and life
MR imaging techniques may further enhance the detec- expectancy of cervical cancer compared with pretreatment
tion capabilities for lymph node and distant metastases. and treatment-related prognostic factors [74, 75], it is rea-
Since MR imaging has better soft tissue contrast than sonable to assume that similar—or better—results can be
CT, it can better distinguish normal anatomic structures obtained with PET/MR.
with physiologic [18F]FDG uptake from true tumor sites, Early results are still controversial, and investigations
for example, for discriminating between normal ovaries are ongoing on definition of tumor volumes for radiation
therapy using PET/MR. A study in 35 patients with cervi-

cal cancer showed excellent concordance between PET and Key Learning Points
T2-weighted imaging, a fluid-sensitive MR sequence, and • MR is the optimal imaging procedure to assess the
diffusion-weighted imaging [76]. Conversely, a similar study local invasion, while [18F]FDG PET is more sensi-
by Zhang et al. in 27 patients with cervical cancer showed tive for detecting lymph node metastasis.
disagreement between gross tumor volume as defined by • PET/MR is still not widely available in the clinical
PET and by MR [77]. practice, but it might become the gold standard to
stage endometrial cancer and to define radiotherapy
plans.
Key Learning Points
• MR per se is the optimal imaging procedure to
assess local invasion, while [18F]FDG PET/CT is 34.3.4 [18F]FDG PET/MR in Epithelial Ovarian
more sensitive for detecting lymph node Cancer
metastasis.
• PET/MR is still not widely available in the clinical The impact of integrated PET/MR to assess the local extent of
practice, but it might become the gold standard to pelvic tumors and their metastases to regional lymph nodes and
stage cervical cancer and to define radiotherapy distant sites is growing. The usefulness of this imaging tech-
plans. nique derives from the higher soft tissue contrast/resolution and
from the absence of streak artifacts secondary compared to CT
[81]. Furthermore, the use of PET/MR considerably reduces
radiation exposure, reported to be 80% lower than PET/CT
34.3.3 [18F]FDG PET/MR in Endometrial Cancer [82]. In addition, multiplanar and multimodal capabilities
(with weighted sequences and specific dynamic sequences like
[18F]FDG PET/MR is emerging as an important diagnostic DWI/MRI and DCE/MR) improve tumor characterization and
procedure in patients with endometrial cancer because MR, provide biological information such as microstructure and cell
with its contrast-enhanced T1 and contrast-enhanced T2 density, tumor oxygenation, and perfusion. Thus, MR associ-
sequences, is especially useful to evaluate tumor infiltration ated with PET may more accurately distinguish benign from
in the miometrial, cervical, and surrounding tissues [78]. A malignant ovarian masses or borderlines tumors [7] and detect
retrospective study showed better diagnostic performance of cancer recurrence better than PET/CT [83]. Furthermore, [18F]
[18F]FDG PET/MR than [18F]FDG PET/CT for evaluation of FDG PET/MR may be superior to [18F]FDG PET/CT in the
local, nodal, and distant recurrences in gynecologic cancer delineation of the primary tumor [84] and may improve the
with a diagnostic accuracy of 93.3% for PET/MR and 80% detection of small peritoneal implants as well as of distant
for PET/CT [79]. By combining better soft tissue contrast metastatic sites [85, 86]. The functional MR-derived quantita-
(typical of MR imaging) with the metabolic information tive parameters combined with the metabolic information from
given by PET, PET/MR is also used to delineate more pre- PET have the potential to act synergistically as biomarkers for
cisely target volume contours of the radiation treatment plan- assessing tumor response to treatment in patients with epithe-
ning in patients with advanced stage disease or in patients lial ovarian cancer, similarly as noted above for cervical cancer
not eligible for surgery because of medical comorbidities. and for endometrial cancer.
Furthermore, PET/MR imaging can yield a multipa-
rametric quantitative information thanks to functional
MR sequences. In patients with endometrial cancer, Shin Key Learning Points
et al. found an inverse correlation between SUV max and • MR associated with [18F]FDG PET/CT seem to pro-
ADCmin of the primary tumors derived from simultane- vide higher accuracy in characterizing ovarian
ous PET/MR imaging [61]. This inverse correlation has masses and may improve small peritoneal implants
a definite prognostic value; for example, in high-grade localization.
tumors, ADC min is significantly lower, while SUV max is • MR-derived quantitative parameters combined with
significantly higher than in low- and intermediate-grade the metabolic information from PET have the
tumors, and it has also been reported that ADC min of the potential to allow better assessment of tumor
primary tumor is the only independent predictor of dis- response to therapy.
ease recurrence [80].
34.4 Uncommon Gynecologic Tumors (including [18F]FDG PET/CT) for initial staging and for
detecting tumor recurrence, if clinically indicated.
34.4.1 Vulvar Cancer
Vulvar cancer is a relatively uncommon cancer, with an over- Key Learning Point
all 5-year survival of 71.2%. About 59% of these cancers are • [18F]FDG PET/CT is indicated for initial staging
diagnosed at the localized stage, 31% at the regional, and and for detecting recurrence.
5% with distant disease; survival rates are 86% for localized,
55% for regional, and 16% for distant metastases [87].
34.4.1.1 [18F]FDG PET/CT 34.4.2 Vaginal Cancer

Although [18F]FDG PET/CT may be useful for initial stag-
ing of patients with vulvar cancer, it cannot replace lymph- Carcinomas of the vagina are uncommon tumors, comprising
adenectomy. A Spanish study conducted in 2013 showed about 1% of the cancers that arise in the female genital sys-
100% sensitivity of [18F]FDG PET/CT for the detection of tem [90, 91]. While early-stage tumors are generally curable
local disease in squamous cell cancers and 60% sensitivity with local therapies, there is no standard treatment of proven
for local disease in non-squamous cell cancers (including efficacy for metastatic disease. About 30–50% of women
Paget, verrucous, and eccrine mucinous tumors) [88]. On with vaginal carcinomas have had a prior hysterectomy
the other hand, Kamran et al. [89] reported 50% sensitivity for benign, premalignant, or malignant disease [91, 92].
and 100% specificity for the detection of locoregional lymph According to the AJCC staging system, tumors in the vagina
node metastasis. These results are sufficient for treatment that involve the cervix of women with an intact uterus are
planning, although lymphadenectomy (or sentinel lymph classified as cervical cancers [93]. Squamous cell cancer
node biopsy) is still required for these patients, considering (SCC) accounts for approximately 85% of vaginal cancer
the low sensitivity. The NCCN guidelines suggest imaging cases [94]. The tumor initially spreads superficially within
Fig. 34.9 Vaginal cancer. [18F]FDG PET/CT performed for staging in the left side of the vaginal canal (SUVmax 8.9) up to the bladder, consis-
a 66-year-old woman with G2 SCC; hysterectomy had been performed tent with malignancy—but no other sites on increased tracer uptake
21 years earlier. [18F]FDG PET/CT shows increased tracer uptake on
Fig. 34.10 Vaginal cancer. [18F]FDG PET/CT performed for restaging in the same patient as in Fig. 34.9, about 6 months after completing chemo-
radiotherapy. The [18F]FDG PET/CT shows complete response
the vaginal wall, and later on it invades the paravaginal tis-

sues and the parametria. SCC of the vagina is associated with Key Learning Point
a high rate of infection with oncogenic strains of human pap- • Preliminary data indicate that [18F]FDG PET/CT
illomavirus (HPV) with many risk factors in common with can be useful for the detection of metastasis
SCC of the cervix [95, 96]. in locoregional lymph nodes.
34.4.2.1 [18F]FDG PET/CT

Given the rarity of this tumor, only few studies have reported
the use of [18F]FDG PET/CT in patients with vaginal cancer. References
In a prospective study of 23 patients, Lamoreaux et al. found
that posttreatment imaging had 100% sensitivity for primary 1. Howlader N, Noone AM, Krapcho M, Garshell J, Miller D,
Altekruse SF et al. SEER cancer statistics review, 1975–2012. 2015.
tumor, with more metastatic lymph nodes detected with http://seer.cancer.gov/statfacts/. Accessed 25 Dec 2014.
PET compared to CT [97]. Another study conducted in 29 2. Grant P, Sakellis C, Jacene HA. Gynecologic oncologic imaging
patients with vaginal cancer showed that [18F]FDG PET/CT with PET/CT. Semin Nucl Med. 2014;44:461–78.
imaging changed the staging in 45% of patients, in particular 3. Colombo N, Preti E, Landoni F, Carinelli S, Colombo A, Marini
C, Sessa C, et al. Endometrial cancer: ESMO clinical practice
for the detection of additional metastatic lymph nodes [98] guidelines for diagnosis, treatment and follow up. Ann Oncol.
(Figs. 34.9 and 34.10). According to the SGO guidelines, 2013;24:33–8.
[18F]FDG PET/CT can be useful in suspected recurrences 4. AIOM Linee guida 2016. Neoplasie dell’utero: endometrio e
[33], although further data are needed in order to define a cervice.
5. Ponisio MR, Fowler KJ, Dehdashti F. The emerging role of PET/
pivotal role of this imaging technique in patients with vagi- MR in gynecologic cancers. PET Clin. 2016;11:425–40.
nal cancer.
6. Pecorelli S. Revised FIGO staging of carcinoma of the vulva, cer- 25. NCCN. Cervical cancer. 2015. http://www.nccn.org/professionals/
vix and endometrium. Int J Gynaecol Obstet. 2009;105:103–4. physician_gls/pdf/cervical.pdf. Accessed 25 Dec 2015.
7. Mohaghegh P, Rockall AG. Imaging strategy for early ovarian 26. Small W Jr, Strauss JB, Jhingran A, Yashar CM, Cardenes

cancer: characterization of adnexal masses with conventional and HR, Erickson-Wittmann BA. ACR Appropriateness Criteria®
advanced imaging techniques. Radiographics. 2012;32:1751–73. definitive therapy for early-stage cervical cancer. Am J Clin Oncol.
8. Schwarz JK, Grigsby PW, Dehdashti F, Delbeke D. The role of 18F- 2012;35:399–405.
FDG PET in assessing therapy response in cancer of the cervix and 27. Choi HJ, Ju W, Myung SK, Kim Y. Diagnostic performance of
ovaries. J Nucl Med. 2009;50(Suppl 1):64S–73S. computed tomography, magnetic resonance imaging, and positron
9. Ledermann JA, Raja FA, Fotopoulou C, Gonzalez-Martin A, emission tomography or positron emission tomography/computer
Colombo N, Sessa C. Newly diagnosed and relapsed epithelial tomography for detection of metastatic lymph nodes in patients
ovarian carcinoma: ESMO Clinical Practice Guidelines for diagno- with cervical cancer: meta-analysis. Cancer Sci. 2010;101:1471–9.
sis, treatment and follow-up. Ann Oncol. 2013;24:25–32. 28. Selman TJ, Mann C, Zamora J, Appleyard TL, Khan K. Diagnostic
10. Antunovic L, Cimitan M, Borsatti E, Baresic T, Sorio R, Giorda G, accuracy of tests for lymph node status in primary cervical cancer:
et al. Revisiting the clinical value of 18F-FDG PET/TC in detection a systematic review and meta-analysis. CMAJ. 2008;178:855–62.
of recurrent epithelial ovarian carcinomas: correlation with histol- 29. Bagade S, Fowler KJ, Schwarz JK, Grigsby PW, Dehdashti F. PET/
ogy, serum CA-125 assay, and conventional radiological modali- MRI evaluation of gynecologic malignancies and prostate cancer.
ties. Clin Nucl Med. 2012;37:e184–8. Semin Nucl Med. 2015;45:293–303.
11. Bandiera E, Romani C, Specchia C, Zanotti L, Galli C, Ruggeri G, 30. Ramirez PT, Jhingran A, Macapinlac HA, Euscher ED, Munsell
et al. Serum human epididymis protein 4 and risk for ovarian malig- MF, Coleman RL, et al. Laparoscopic extraperitoneal para-aortic
nancy algorithm as new diagnostic and prognostic tool for epithe- lymphadenectomy in locally advanced cervical cancer: a prospec-
lial ovarian cancer management. Cancer Epidemiol Biomarkers tive correlation of surgical findings with positron emission tomog-
Prev. 2011;20:2496–506. raphy/computed tomography findings. Cancer. 2011;117:1928–34.
12. Holcomb K, Vucetic Z, Miller MC, Knapp RC. Human epididymis 31. Kang S, Kim SK, Chung DC, Seo SS, Kim JY, Nam BH, et al.
protein 4 offers superior specificity in the differentiation of benign Diagnostic value of 18F-FDG PET for evaluation of paraaortic nodal
and malignant adnexal masses in premenopausal woman. Am J metastasis in patients with cervical carcinoma: a metaanalysis. J
Obstet Gynecol. 2011;205:358.e1–6. Nucl Med. 2010;51:360–7.
13. Morgan RJ Jr, Alvarez RD, Armstrong DK, Burger RA, Chen LM, 32. Gadducci A, Tana R, Cosio S, Cionini L. Treatment options in
Copeland L, et al. Ovarian cancer, version 2.2013. J Natl Compr recurrent cervical cancer. Oncol Lett. 2010;1:3–11.
Cancer Netw. 2013;11:1199–209. 33. Salani R, Backes FJ, Fung MF, Holschneider CH, Parker LP,

14. Erdy YE. Limits of tumor detectability in nuclear medicine and Bristow RE, Goff BA. Posttreatment surveillance and diagnosis of
PET. Mol Imag Radionucl Ther. 2012;21:23–8. recurrence in women with gynecologic malignancies: Society of
15. Cohade C. Altered biodistribution on FDG-PET with emphasis on Gynecologic Oncologists recommendations. Am J Obstet Gynecol.
brown fat and insulin effect. Semin Nucl Med. 2010;40:283–93. 2011;204:466–78.
16. Surasi DS, Bhambhvani P, Baldwin JA, Almodovar SE, O’Malley 34. Mitchell DG, Javitt MC, Glanc P, Bennett GL, Brown DL, Dubinsky
JP. 18F-FDG PET and PET/CT patient preparation: a review of the T, et al. ACR appropriateness criteria staging and follow-up of ovar-
literature. J Nucl Med Technol. 2014;42:5–13. ian cancer. J Am Coll Radiol. 2013;10:822–7.
17. Liu Y. Benign ovarian and endometrial uptake on FDG PET-CT: 35. Chu Y, Zheng A, Wang F, Lin W, Yang X, Han L, Chen Y, Bai
patterns and pitfalls. Ann Nucl Med. 2009;23:107–12. L. Diagnostic value of 18F-FDG-PET or PET-CT in recurrent cer-
18. Yau YY, Chan WS, Tam YM, Vernon P, Wong S, Coel M, Chu vical cancer: a systematic review and meta-analysis. Nucl Med
SK. Application of intravenous contrast in PET/CT: does it really Commun. 2014;35:144–50.
introduce significant attenuation correction error? J Nucl Med. 36. Xiao Y, Wei J, Zhang Y, Xiong W. Positron emission tomogra-
2005;46:283–91. phy alone, positron emission tomography-computed tomography
19. Ozülker T, Ozülker F, Mert M, Ozpaçaci T. Clearance of the high and computed tomography in diagnosing recurrent cervical car-
intestinal 18F-FDG uptake associated with metformin after stopping cinoma: a systematic review and meta-analysis. Arch Med Sci.
the drug. Eur J Nucl Med Mol Imaging. 2010;37:1011–7. 2014;10:222–31.
20. Oldan JD, Patel PS. Positron emission tomography/computed
37. Grigsby PW, Siegel BA, Dehdashti F, Rader J, Zoberi I. Posttherapy
tomography for gynecologic malignancies. Obstet Gynecol Surv. 18
F-fluorodeoxyglucose positron emission tomography in car-
2016;71:545–56. cinoma of the cervix: response and outcome. J Clin Oncol.
21. Marnitz S, Köhler C, Roth C, Füller J, Hinkelbein W, Schneider 2004;22:2167–71.
A. Is there a benefit of pretreatment laparoscopic transperitoneal 38. Netzer I, Sobeh S, Keidar Z, Lowenstein L, Lavie O, Yosef RB,
surgical staging in patients with advanced cervical cancer? Gynecol Amit A. To operate or to radiate: the added value of the maximal
Oncol. 2005;99:536–44. standardized uptake value in PET-FDG in cervical cancer patients.
22. Hildebrandt MG, Kodahl AR, Teilmann-Jørgensen D, Mogensen Med Oncol. 2013;30:558.
O, Jensen PT. [18F] fluorodeoxyglucose PET/computed tomography 39. Sarker A, Im HJ, Cheon GJ, Chung HH, Kang KW, Chung JK,
in breast cancer and gynecologic cancers: a literature review. PET Kim EE, Lee DS. Prognostic implications of the SUVmax of primary
Clin. 2015;10:89–104. tumors and metastatic lymph node measured by 18F-FDG PET in
23. Kidd EA, Siegel BA, Dehdashti F, Rader JS, Mutch DG, Powell patients with uterine cervical cancer: a meta-analysis. Clin Nucl
MA, Grigsby PW. Lymph node staging by positron emission Med. 2016;41:34–40.
tomography in cervical cancer: relationship to prognosis. J Clin 40. Kidd EA, Thomas M, Siegel BA, Dehdashti F, Grigsby PW. Changes
Oncol. 2010;28:2108–13. in cervical cancer FDG uptake during chemoradiation and associa-
24. Yildirim Y, Sehirali S, Avci ME, Yilmaz C, Ertopcu K, Tinar
tion with response. Int J Radiat Oncol Biol Phys. 2013;85:116–22.
S, Duman Y, Sayhan S. Integrated PET/CT for the evaluation of 41. Zhao Q, Feng Y, Mao X, Qie M. Prognostic value of fluorine-
para-aortic nodal metastasis in locally advanced cervical cancer 18- fluorodeoxyglucose positron emission tomography or
patients with negative conventional CT findings. Gynecol Oncol. PET-computed tomography in cervical cancer: a meta-analysis. Int
2008;108:154–9. J Gynecol Cancer. 2013;23:1184–90.
42. Herrera FG, Prior JO. The role of PET/CT in cervical cancer. Front 58. Du XL, Jiang T, Sheng XG, Li QS, Wang C, Yu H, et al. PET/CT
Oncol. 2013;3:34. scanning guided intensity-modulated radiotherapy in treatment of
43. Kakhki VR, Shahriari S, Treglia G, Hasanzadeh M, Zakavi SR, recurrent ovarian cancer. Eur J Radiol. 2012;81:3551–6.
Yousefi Z, Kadkhodayan S, Sadeghi R. Diagnostic performance 59. Grueneisen J, Schaarschmidt BM, Beiderwellen K, Schulze-Hagen
of fluorine 18-fluorodeoxyglucose positron emission tomography A, Heubner M, Kinner S, et al. Diagnostic value of diffusion
imaging for detection of primary lesion and staging of endometrial weighted-imaging in simultaneous 18F-FDG PET/MR imaging for
cancer patients: systematic review and meta-analysis of the litera- whole-body staging of woman with pelvic malignancies. J Nucl
ture. Int J Gynecol Cancer. 2013;23:1536–43. Med. 2014;55:1930–5.
44. Melamed A, Hinchcliff EM, Clemmer JT, Bregar AJ, Uppal
60. Brandmaier P, Purz S, Bremicker K, Höckel M, Barthel H, Kluge
S, Bostock I, Schorge JO, Del Carmen MG, Rauh-Hain R, et al. 18F-FDG PET/MRI: correlation of apparent diffusion coef-
JA. Preoperative staging of cervical cancer: is 18F-FDG-PET/CT ficient (ADC) and standardized uptake value (SUV) in primary and
really effective in patients with early stage disease? Gynecol Oncol. recurrent cervical cancer. PLoS One. 2015;10(11):e0141684.
2011;123:236–40. 61. Shih IL, Yen RF, Chen CA, Chen BB, Wei SY, Chang WC, et al.
45. Kadkhodayan S, Shahriari S, Treglia G, Yousefi Z, Sadeghi
Standardized uptake value and apparent diffusion coefficient of
R. Accuracy of 18F-FDG PET imaging in the follow up of endome- endometrial cancer evaluated with integrated whole-body PET/
trial cancer patients: a systematic review and meta-analysis of the MR: correlation with pathological prognostic factors. J Magn
literature. Gynecol Oncol. 2013;128:397–404. Reson Imaging. 2015;42:1723–32.
46. Ozcan Kara P, Kara T, Kaya B, Kara Gedik G, Sari O. The value of 62. Samarin A, Burger C, Wollenweber SD, Crook DW, Burger IA,
FDG-PET/CT in the post-treatment evaluation of endometrial car- Schmid DT, et al. PET/MR imaging of bone lesions implications
cinoma: a comparison of PET/CT findings with conventional imag- for PET quantification from imperfect attenuation correction. Eur J
ing and CA 125 as a tumour marker. Rev Esp Med Nucl Imagen Nucl Med Mol Imaging. 2012;39:1154–60.
Mol. 2012;31:257–60. 63. Miller TR, Grigsby PW. Measurement of tumor volume by PET to
47. Amit A, Person O, Keidar Z. FDG PET/CT in monitoring response evaluate prognosis in patients with advanced cervical cancer treated
to treatment in gynecological malignancies. Curr Opin Obstet by radiation therapy. Int J Radiat Oncol Biol Phys. 2002;53:353–9.
Gynecol. 2013;25:17–22. 64. Grueneisen J, Schaarschmidt BM, Heubner M, Aktas B, Kinner S,
48. Nam EJ, Yun MJ, Oh YT, Kim JW, Kim JH, Kim S, Jung YW, Kim Forsting M, et al. Integrated PET/MRI for whole-body staging of
SW, Kim YT. Diagnosis and staging of primary ovarian cancer: cor- patients with primary cervical cancer: preliminary results. Eur J
relation between PET/CT, Doppler US, and CT or MRI. Gynecol Nucl Med Mol Imaging. 2015;42:1814–24.
Oncol. 2010;116:389–94. 65. Fraum TJ, Fowler KJ, McConathy J, Parent EE, Dehdashti F, Grigsby
49. Castellucci P, Perrone AM, Picchio M, Ghi T, Farsad M, Nanni C, PW, Siegel BA. PET/MRI for the body imager: abdominal and pel-
et al. Diagnostic accuracy of 18-FDG PET/CT in characterizing vic oncologic applications. Abdom Imaging. 2015;40:1387–404.
ovarian lesions and staging ovarian cancer: correlation with trans- 66. Belhocine T, Thille A, Fridman V, Albert A, Seidel L, Nickers P,
vaginal ultrasonography, computed tomography, and histology. et al. Contribution of whole-body 18FDG PET imaging in the man-
Nucl Med Commun. 2007;28:589–95. agement of cervical cancer. Gynecol Oncol. 2002;87:90–7.
50. Yuan Y, Gu ZX, Tao XF, Liu SY. Computed tomography, magnetic 67. Havrilesky LJ, Kulasingam SL, Matchar DB, Myers ER. FDG-PET
resonance imaging, and positron emission tomography/computer for management of cervical and ovarian cancer. Gynecol Oncol.
tomography for detection of metastatic lymph nodes in patients 2005;97:183–91.
with ovarian cancer: a meta-analysis. Eur J Radiol. 2012;81: 68. Kitajima K, Suenaga Y, Ueno Y, Kanda T, Maeda T, Deguchi M,
1002–6. et al. Fusion of PET and MRI for staging of uterine cervical cancer:
51. Nakamura K, Hongo A, Kodama J, Hiramatsu Y. The pretreatment comparison with contrast-enhanced 18F-FDG PET/CT and pelvic
of maximum standardized uptake value (SUVmax) of the primary MRI. Clin Imaging. 2014;38:464–9.
tumor is predictor for poor prognosis for patients with epithelial 69. Kidd EA, Spencer CR, Huettner PC, Siegel BA, Dehdashti F, Rader
ovarian cancer. Acta Med Okayama. 2012;66:53–60. JS, Grigsby PW. Cervical cancer histology and tumor differentiation
52. Chung HH, Kwon HW, Kang KW, Park NH, Song YS, Chung JK, affect 18F-fluorodeoxyglucose uptake. Cancer. 2009;115:3548–54.
et al. Prognostic value of preoperative metabolic tumor volume and 70. Kuang F, Ren J, Zhong Q, Liyuan F, Huan Y, Chen Z. The value of
total lesion glycolysis in patients with epithelial ovarian cancer. apparent diffusion coefficient in the assessment of cervical cancer.
Ann Surg Oncol. 2012;19:1966–72. Eur Radiol. 2013;23:1050–8.
53. Avril N, Gourtsoyianni S, Reznek R. Gynecological cancer.
71. Grigsby PW, Siegel BA, Dehdashti F. Lymph node staging by posi-
Methods Mol Biol. 2011;727:171–89. tron emission tomography in patients with carcinoma of the cervix.
54. Nanni C, Rubello D, Farsad M, De Iaco P, Sansovini M, Erba P, J Clin Oncol. 2001;19:3745–9.
et al. 18FDG-PET/CT in the evaluation of recurrent ovarian can- 72. Ding XP, Feng L, Ma L. Diagnosis of recurrent uterine cervical can-
cer: a prospective study on forty-one patients. Eur J Surg Oncol. cer: PET versus PET/CT: a systematic review and meta-analysis.
2005;31:792–7. Arch Gynecol Obstet. 2014;290:741–7.
55. Chen YM, Chen T, Zee CS, Shi YP, Wan LR, Tong LJ, et al. Is there 73. Beiderwellen K, Geraldo L, Ruhlmann V, Heusch P, Gomez B,
an impact on the surveillance and clinical management of recurrent Nensa F. Accuracy of [18F]FDG PET/MRI for the detection of liver
ovarian cancer? Research based on a large sample in a single PET/ metastases. PLoS One. 2015;10(9):e0137285.
CT center. Nucl Med Commun. 2014;35:347–52. 74. Grigsby PW, Siegel BA, Dehdashti F, Rader J, Zoberi I. Post
56. Hebel CB, Behrendt FF, Heinzel A, Krohn T, Mottaghy FM,
therapy [18F] fluorodeoxyglucose positron emission tomography
Bauerschlag DO, Verburg FA. Negative 18F-2-fluorodeoxyglucose in carcinoma of the cervix: response and outcome. J Clin Oncol.
PET/CT predicts good cancer specific survival in patients with a 2004;22:2167–71.
suspicion of recurrent ovarian cancer. Eur J Radiol. 2014;83:463–7. 75. Schwarz JK, Siegel BA, Dehdashti F, Grigsby PW. Association of
57. Ebina Y, Watari H, Kaneuchi M, Takeda M, Hosaka M, Kudo M, post therapy positron emission tomography with tumor response
Yamada H, Sakuragi N. Impact of FDG PET in optimizing patient and survival in cervical carcinoma. JAMA. 2007;298:2289–95.
selection for cytoreductive surgery in recurrent ovarian cancer. Eur 76. Sun H, Xin J, Zhang S, Guo Q, Lu Y, Zhai W, et al. Anatomical
J Nucl Med Mol Imaging. 2014;41:446–51. and functional volume concordance between FDG PET, and T2
and diffusion-weighted MRI for cervical cancer: a hybrid PET/MR 87. Howlader N, Noone A, Krapcho M et al. SEER cancer statis-
study. Eur J Nucl Med Mol Imaging. 2014;41:898–905. tics review, 1975–2012. 2015. http://seer.cancer.gov/statfacts/.
77. Zhang S, Xin J, Guo Q, Ma J, Ma Q, Sun H. Comparison of tumor Accessed 25 Dec 2014.
volume between PET and MRI in cervical cancer with hybrid PET/ 88. Peiró V, Chiva L, González A, Bratos R, Alonso S, Márquez R,
MR. Int J Gynecol Cancer. 2014;24:744–50. et al. Utility of the PET/CT in vulvar cancer management. Rev Esp
78. Frei KA, Kinkel K, Bonél HM, Lu Y, Zaloudek C, Hricak
Med Nucl Imagen Mol. 2014;33:87–92.
H. Prediction of myometrial invasion in patients with endometrial 89. Kamran MW, O’Toole F, Meghen K, Wahab AN, Saadeh FA,

cancer: clinical utility of contrast-enhanced MR imaging-a meta- Gleeson N, et al. Whole-body [18F]fluoro-2-deoxyglucose positron
analysis and Bayesian analysis. Radiology. 2002;216:444–9. emission tomography scan as combined PET-CT staging prior to
79. Kitajima K, Suenaga Y, Ueno Y, Kanda T, Maeda T, Takahashi planned radical vulvectomy and inguinofemoral lymphadenectomy
S. Value of fusion of PET and MRI for staging of endome- for squamous vulvar cancer: a correlation with groin node metasta-
trial cancer: comparison with 18F-FDG contrast-enhanced PET/ sis. Eur J Gynaecol Oncol. 2014;35:230–5.
CT and dynamic pelvic contrast-enhanced MRI. Eur J Radiol. 90. American Cancer Society. Cancer facts and figures 2017. Atlanta,
2013;32:1672–6. GA: American Cancer Society; 2017.
80. Nakamura K, Imafuku N, Nishida T, Niwa I, Joja I, Hongo A, 91. Eiffel PJ, Berek JS, Markman MA. Cancer of the cervix, vagina,
et al. Measurement of the minimum apparent diffusion coefficient and vulva. In: DeVita Jr VT, Lawrence TS, Rosenberg SA, editors.
(ADCmin) of the primary tumor and CA125 are predictive of disease Cancer: principles and practice of oncology. 9th ed. Philadelphia,
recurrence for patients with endometrial cancer. Gynecol Oncol. PA: Lippincott Williams & Wilkins; 2011. p. 1311–44.
2012;124:335–9. 92. Stock RG, Chen AS, Seski J. A 30-year experience in the man-
81. Partovi S, Kohan A, Rubbert C, Vercher-Conejero JL, Gaeta C, Yuh agement of primary carcinoma of the vagina: analysis of prog-
R, et al. Clinical oncologic applications of PET/MRI: a new hori- nostic factors and treatment modalities. Gynecol Oncol. 1995;56:
zon. Am J Nucl Med Mol Imaging. 2014;4:202–12. 45–52.
82. Hirsch FW, Sattler B, Sorge I, Kurch L, Viehweger A, Ritter L, 93. Compton CC. Vagina. In: Edge SB, Byrd DR, Compton CC, et al.,
et al. PET/MR in children. Initial clinical experience in paediat- editors. AJCC cancer staging manual. 7th ed. New York, NY:
ric oncology using an integrated PET/MR scanner. Pediatr Radiol. Springer; 2010. p. 387–9.
2013;43:860–75. 94. Eifel P, Berek J, Markman M. Cancer of the cervix, vagina, and
83. Beiderwellen K, Grueneisen J, Ruhlmann V, Buderath P, Aktas B, vulva. In: DeVita Jr VT, Hellman S, Rosenberg SA, editors. Cancer:
Heusch P, et al. 18F-FDG PET/MRI vs. PET/CT for whole-body principles and practice of oncology. Vols. 1 & 2. 8th ed. Philadelphia,
staging in patients with recurrent malignancies of the female pelvis: PA: Lippincott Williams & Wilkins; 2008. p. 1496–543.
initial results. Eur J Nucl Med Mol Imaging. 2015;42:56–65. 95. Daling JR, Madeleine MM, Schwartz SM, Shera KA, Carter JJ,
84. Queiroz MA, Kubik-Huch RA, Hauser N, Freiwald-Chilla B, von McKnight B, et al. A population-based study of squamous cell vag-
Schulthess G, Froehlich JM, Veit-Haibach P. PET/MRI and PET/ inal cancer: HPV and cofactors. Gynecol Oncol. 2002;84:263–70.
CT in advanced gynecological tumors: initial experience and com- 96. Ikenberg H, Runge M, Göppinger A, Pfleiderer A. Human papillo-
parison. Eur Radiol. 2015;25:2222–30. mavirus DNA in invasive carcinoma of the vagina. Obstet Gynecol.
85. Zhang P, Cui Y, Li W, Ren G, Chu C, Wu X. Diagnostic accuracy of 1990;76(3 Pt 1):432–8.
diffusion-weighted imaging with conventional MR imaging for dif- 97. Lamoreaux WT, Grigsby PW, Dehdashti F, Zoberi I, Powell MA,
ferentiating complex solid and cystic ovarian tumors at 1.5T. World Gibb RK, et al. FDG-PET evaluation of vaginal carcinoma. Int J
J Surg Oncol. 2012;10:237. Radiat Oncol Biol Phys. 2005;62:733–7.
86. Ko SM, Jung DC, Oh YT, Kim SH, Kang S. Discrepancy between 98. Robertson NL, Hricak H, Sonoda Y, Sosa RE, Benz M, Lyons G,
preoperative imaging and surgical findings in patients with et al. The impact of FDG-PET/CT in the management of patients
advanced ovarian cancer: a study of initial false-negative lesions. J with vulvar and vaginal cancer. Gynecol Oncol. 2016;140:420–4.
Comput Assist Tomogr. 2013;37:568–71.
Hybrid Imaging for Male Malignancies
35
Akram Al-Ibraheem, Abdullah S. Al Zreiqat,
Serena Chiacchio, and Abedallatif A. AlSharif
Contents
35.1 Prostate Cancer 900
35.1.1 Clinical Background 900
35.1.2 Conventional Bone Scintigraphy 900
35.1.3 18F-Fluoride PET/CT 905
35.1.4 [18F]FDG PET/CT 905
35.1.5 [11C]Choline and 18F-Choline PET/CT 905
35.1.6 68Ga-PSMA-Ligand PET/CT 909
35.1.7 Perspectives for 18F-FACBC PET/CT 912
35.2 Testicular Cancer 914
35.2.2 [18F]FDG PET/CT in Testicular Cancer 915
35.3 Penile Cancer 919
35.3.2 Bone Scintigraphy 919
35.3.3 Sentinel Lymph Node Biopsy 919
35.3.4 PET/CT with [18F]FDG 921
References 922
Learning Objectives • Understand the pathophysiologic basis and the clinical

• Summarize the main features regarding epidemiology, rationale for the residual use of bone scintigraphy with
etiology, histopathology, and most common clinical pre- 99m
Tc-bisphosphonates in patients with prostate cancer.
sentation of male malignancies, including prostate can- • Summarize the basic modalities for performing and inter-
cer, testicular cancer, and penile cancer. preting bone scintigraphy with 99mTc-bisphosphonates,
including planar imaging and SPECT/CT imaging.
• Understand the pathophysiologic basis and the clinical
rationale for the emerging use of PET/CT with 18F-NaF
A. Al-Ibraheem for detecting skeletal metastasis in patients with prostate
Department of Nuclear Medicine, King Hussein Cancer Center,
cancer.
Amman, Jordan
• Summarize the basic modalities of performing and inter-
A. S. Al Zreiqat
preting a PET/CT scan with 18F-NaF PET/CT in patients
Department of Nuclear Medicine, Jordanian Royal Medical
Services, Amman, Jordan with prostate cancer.
• Understand the pathophysiologic basis for the use of
S. Chiacchio
Regional Center of Nuclear Medicine, University Hospital of Pisa, PET/CT imaging with [18F]FDG in patients with prostate
Pisa, Italy cancer.
A. A. AlSharif (*) • Summarize the basic modalities of performing and inter-
Department of Radiology and Nuclear Medicine, School of preting a PET/CT scan with [18F]FDG PET/CT in patients
Medicine, Jordan University Hospital, University of Jordan, with prostate cancer.
Amman, Jordan
e-mail: abedsharif@ju.edu.jo

900 A. Al-Ibraheem et al.
• Understand the pathophysiologic basis and the clinical Park Cancer Institute,1 provides an early clue to the pres-
rationale for the use of PET/CT with [11C]choline or ence of prostate carcinoma. However, it is a nonspecific
18
F-choline in patients with prostate cancer. biomarker, as its serum levels increase in men as they age
• Summarize the basic modalities of performing and inter- and the prostate enlarges, as well as in patients with pros-
preting a PET/CT scan with [11C]choline or 18F-choline in tatitis. Nevertheless, an elevated PSA value should lead to
patients with prostate cancer. other diagnostic procedures such as digital rectal examina-
• Understand the pathophysiologic basis and the clinical tion, endorectal ultrasound (US), and biopsy. The ability to
rationale for the use of PET/CT with 68Ga-PSMA-ligand detect prostate cancer earlier has contributed to an apparent
in patients with prostate cancer. increase in the incidence of prostate malignancy.
• Summarize the basic modalities of performing and inter- Multiple imaging modalities are available to image the
preting a PET/CT scan with 68Ga-PSMA-ligand in prostate, including ultrasound (US), X-ray computed tomog-
patients with prostate cancer. raphy (CT), and magnetic resonance imaging (MRI). If the
• Summarize the potential clinical usefulness of PET/CT imaging studies depict findings suggestive of prostate can-
with 18F-FACBC in patients with prostate cancer. cer, further evaluation with multicore transrectal biopsy
• Understand the pathophysiologic basis and clinical ratio- should be performed to determine if cancer is present and
nale for the use of PET/CT imaging with [18F]FDG in to characterize the tumor. The most common grading system
patients with testicular cancer. for prostate cancer is the Gleason score. Under low mag-
• Summarize the basic modalities of performing and inter- nification power during histologic analysis, the glandular
preting a PET/CT scan with [18F]FDG in patients with pattern and degree of differentiation are each scored from
testicular cancer. 1 (least abnormal) to 5 (most abnormal). The score for each
• Understand the pathophysiologic basis and clinical ratio- characteristic is summed, producing a score ranging from 2
nale for the use of PET/CT imaging with [18F]FDG in to 10. If prostate cancer is identified, further imaging (usu-
patients with penile cancer. ally with contrast-enhanced CT and or MRI) is performed to
• Summarize the basic modalities of performing and inter- determine if the tumor extends beyond the prostate capsule
preting a PET/CT scan with [18F]FDG PET/CT in patients and if it involves seminal vesicles and lymph nodes. If the
with penile cancer. Gleason score is >10, a bone scan should be performed.
• Understand the pathophysiologic basis and clinical ratio-
nale of staging patients with penile cancer with the use of
radioguided sentinel lymph node biopsy. 35.1.2 Conventional Bone Scintigraphy
• Summarize the basic modalities of performing preopera-
tive lymphoscintigraphy and intraoperative gamma- Skeletal scintigraphy is very sensitive in the detection of
probe-guided detection of sentinel lymph node(s) in osseous metastatic disease. Bone scan is 50–80% more sen-
patients with penile cancer. sitive than radiographs (and CT) in detecting skeletal metas-
tases, because at least about 30–75% of the bone mineral
content must be lost before a metastasis is evident on a radio-
35.1 Prostate Cancer graph [2].
In a meta-analysis on the detection rate of bone metasta-
35.1.1 Clinical Background ses by bone scan in patients with newly diagnosed prostate
cancer, detection rates were 2.3% in patients with serum PSA
Prostate carcinoma is the most common life-threatening can- ≤10 ng/mL, 5.3% in patients with PSA between 10.1 and
cer affecting men in the Western world. Rates of detection 19.9 ng/mL, and 16.2% in patients with PSA levels between
vary, with a lower prevalence in South and East Asia than in 20 and 49.9 ng/mL [2]. The likelihood for bone metastases
Europe and the United States [1]. Prostate carcinoma is more increases to about 50% when the PSA level is greater than
common in men over the age of 50 (with ~1% occurring in 50 ng/mL [3]. Based on the Gleason score, detection rates
mid <50 years of age); it is diagnosed in 80% of men by the were 5.6% in patients with a score ≤7 and 29.9% in patients
age of 80. with a score ≥8 [2].
In the majority of cases, early-stage prostate cancer does
not cause symptoms. Sensitive diagnostic procedures are
key to improve survival rates. The prostate-specific anti- 1
https://www.roswellpark.org/cancer/prostate/prevention-early-detec-
gen (PSA) immunoassay, developed by Chu at the Roswell tion/screening/history-psa
35 Hybrid Imaging for Male Malignancies 901
Fig. 35.1 Examples of bone

scintigraphy performed for
restaging in two patients with
biochemical recurrence from
prostate cancer. Left panel shows
the bone scan in a 65-year-old
patient treated with antiandrogen
hormonal therapy after initial
diagnosis of metastatic prostate
cancer; after an initial decline
lasting several months, serum
PSA rose again, up to 1015 ng/
mL. The bone scan shows
diffusely and intensely increased
tracer uptake in virtually all the
skeletal segment, with a “super
scan” pattern (extremely low
activity in soft tissues and urinary
excretory tract). Right panel
shows the bone scan in a
68-year-old patient with
biochemical recurrence of
prostate cancer after initial
response to treatment: there are
many areas of focally increased
tracer uptake throughout several
skeletal segments (reproduced
New York: Springer, 2017)
Bone scintigraphy should be limited to patients for ini- In patients with prostate cancer after primary treatment
tial staging with elevated PSA (greater than 10 ng/mL), high (surgery or radiotherapy), bone scintigraphy is indicated
Gleason scores [4–6], locally advanced disease, elevated in two different situations: symptomatic patients or after
alkaline phosphatase, or bone symptoms [2, 3]. biochemical recurrence (increasing serum PSA levels)
The most commonly used radiopharmaceuticals (Fig. 35.1). The bone scan is also indicated in patients
to visualize bone metastases belong to the bisphos- under treatment, for monitoring response to therapy
phonates family labeled with 99mTc, such as 99mTc- (Fig. 35.2).
hydroxymethylenediphosphonic acid (99mTc-HMDP),
99m
Tc-methylene diphosphonate ( Tc-MDP), 99mTc-
99m
35.1.2.1 Patient Preparation
hydroxyethylene diphosphonate (99mTc-HDP), and Patients should be well hydrated before the study and dur-
99m
Tc-3,3-diphosphono-1,2-propanedicarboxylic acid ing the uptake time (at least two glasses of water, ~450 mL)
(99mTc-DPD). The diagnostic sensitivity of bone scin- in order to enhance renal excretion. This reduces radia-
tigraphy in prostate cancer patients varies from 75% tion exposure and contributes to achieving optimal target-
to 95%; its specificity is relatively lower (from 60% to to-background ratios. Patients do not need to fast and are
75%), because metabolic bone reactions can also occur in allowed to take all their usual medications; furthermore,
benign conditions such as osteoarthritis, inflammation, and they are also encouraged to drink more frequently for the
trauma [7, 8]. remainder of the day. Patients should be asked to empty
baseline 3-year follow-up
Fig. 35.2 Example of favorable response to hormonal therapy in a shows virtual disappearance of the metastatic lesions visualized in the
patient with skeletal metastatic prostate cancer. Left panel: in addition baseline scan, with appearance of two new small foci of increased
to bilateral arthritic changes in the first metacarpal joints, the bone scan tracer uptake, in the anterior arc of the fifth right rib and in the paraver-
shows multiple foci with markedly increased tracer uptake in the spine tebral portion of the seventh right rib; serum PSA was 2.4 ng/mL when
and ribs; serum PSA when performing this scan was 115 ng/mL. Right performing this follow-up scan
panel: the bone scan performed 3 years after starting hormonal therapy
their bladders immediately before image acquisition, and 35.1.2.3 Precautions

any metal objects should be removed to prevent attenuation Special attention must be paid to possible urine contamina-
artifacts. In selected circumstances (e.g., retention of radio- tion of the patients’ perineum, undergarments, and trousers,
active urine in the bladder because of obstruction), catheter- particularly if no bladder catheter was inserted.
ization of the urinary bladder may help to visualize pelvic
bones. 35.1.2.4 Possible Drug Interactions
No particular drug interaction is known.
35.1.2.2 Administered Radioactivity
For bone scintigraphy in adults, the average administered 35.1.2.5 Protocol/Image Acquisition
activity through a single intravenous injection should be The use of a single-headed or dual-headed gamma camera
500 MBq (300–740 MBq, 8–20 mCi). equipped with a low-energy, high-resolution parallel-hole
Fig. 35.3 Newly diagnosed high-risk prostatic cancer in an 87-year- panel, fused SPECT/CT in left upper panel). The CT component of the
old man (Gleason score 10, serum PSA 5.6 ng/mL). Planar bone scin- scan (left lower panel) demonstrates an abnormally thickened and irreg-
tigraphy performed for primary staging shows a single focus of ular left parietal bone. Final diagnosis was Paget’s disease of the bone
markedly increased tracer uptake in the parietal bone on the left side (reproduced from: Strauss HW, Mariani G, Volterrani D, Larson SM,
(right lower panel). SPECT/CT confirms the abnormal scintigraphic eds. Nuclear Oncology – From Pathophysiology to Clinical Applications,
finding as the sole abnormality (SPECT component in right upper 2nd Edition. New York: Springer, 2017)
collimator is recommended. The energy window is centered than 1.5 million counts. The image format is 1024 × 256 or
on the photon energy peak of 99mTc (140 keV), and the win- 2048 × 512. The whole-body images can be processed with a
dow width is generally set at 15% or 20%. spatial filter to reduce pixel-to-pixel variation.
Whole-body images can routinely be acquired any- The diagnostic sensitivity and specificity of bone scanning
time between 2 and 5 h after injection of the radiolabeled can be significantly increased with SPECT or, if available,
bisphosphonate. The recommended scan speed is 10–15 cm/ SPECT/CT (Fig. 35.3). Tomographic images and correlation
min. The scan speed should be adjusted so that routine ante- with CT findings improve both the sensitivity and specificity
rior and posterior whole-body images each contain more of the bone scan.
SPECT imaging should be performed as recommended by ologic tracer uptake may also be slightly nonhomogeneous,
the gamma camera manufacturer. In a typical acquisition pro- which may reflect degenerative changes.
tocol for a dual-headed gamma camera with the detector heads Special attention must be paid to avoid misinterpretation
set in a 180° geometry, a total of 60 or 64 frames per detector of “flare phenomenon” which occurs frequently after che-
head, each with duration of 10–30 s, are acquired over 360° motherapy and usually reflects a favorable response of bone
into a 128 × 128 matrix. An equivalent total number of counts metastases to treatment [9]. This phenomenon is typically
should be acquired if continuous acquisition is used. seen between 2 weeks and 3 months following a change in
SPECT/CT images are acquired using a multimodality hormonal or chemotherapy, but can rarely be seen as late as
camera that combines a gamma camera and a multi-slice spi- 6 months after treatment. Continued increase in the number
ral or flat panel/cone beam CT scanner. The CT scan is usu- and intensity of lesions beyond 6 months usually indicates
ally performed immediately before the SPECT acquisition; disease progression, up to reach the extreme pattern of the
acquisition protocols are specific to each type of machine. so-called super scan (Fig. 35.4) [9]. Patients on androgen
The use of intravenous iodinated contrast material is gener- deprivation therapy usually demonstrate excellent response
ally not required. on bone scintigraphy, with complete or almost complete res-
olution of lesions (Fig. 35.3). It should, however, be noted
35.1.2.6 Interpretation that androgen deprivation therapy for prostate cancer (bicalu-
In general, a normal bone scan demonstrates uniform tracer tamide, estrogens) may result in gynecomastia, which leads
distribution throughout the skeleton. The pattern of physi- to increased tracer uptake in the mammary tissue.
Fig. 35.4 Typical appearance of “super scan” in a patient with diffuse so evident in the low intensity/contrast display (left panels), the “super
skeletal metastatic disease from prostate cancer. The anterior and poste- scan” pattern is obvious especially in the high intensity/contrast dis-
rior whole-body scans are displayed both with low intensity/contrast play, where soft tissues exhibit very low activity and there is virtually
(left panel) and with high intensity/contrast (right panels). Although not no detectable radioactivity in the urinary excretion tract
A bone scan index (BSI) has been developed to quantify To obtain high-quality skeletal images on current devices,
the load of bone metastases from prostate cancer. This index is emission scans with acquisition times of 1–2 min per bed
calculated by considering the average weight of bones derived position in 3D mode are performed, depending on the above-
from a reference man and the fractional contribution of each mentioned parameters.
bone expressed as a percentage of the entire skeleton [9]. Usually a 128 × 128 matrix is applied for image acqui-
Therefore, BSI represents the percentage of the skeletal mass sition, although a 256 × 256 matrix may be advantageous
occupied by metastases, a value that indicates the extent of bone if processing times are reasonable. Commercially available
involvement. BSI was demonstrated to represent a reproducible software packages for iterative reconstruction are widely
prognostic tool and can be used both for stratifying patients in available. The optimal number of iterations and subsets,
treatment protocols and for assessing treatment response [10]. filters, and other reconstruction parameters will depend on
patient and camera factors. In general, the same reconstruc-
tion protocols as are used for imaging [18F]FDG PET may be
35.1.3 18F-Fluoride PET/CT used for 18F-NaF. Maximum intensity projection images can
be generated to help facilitate lesion detection.
PET/CT with 18F-sodium fluoride (18F-NaF) detects skeletal
metastatic disease by a mechanism similar to that of 99mTc- 35.1.3.4 Interpretation
labeled bisphosphonates. Clinical indications are the same In general, a normal 18F-NaF PET shows uniform tracer dis-
as for bone scintigraphy with bisphosphonates. According to tribution throughout the skeleton with a pattern similar to
Langsteger et al., 18F-NaF PET/CT has sensitivity, specific- that seen on conventional bone scintigraphy, also regarding
ity, and accuracy values of 91%, 83%, and 88%, respectively, distribution and variants (Fig. 35.5). Similarly to conven-
for detecting skeletal metastases [10]. tional bone scintigraphy, 18F-NaF PET/CT findings may be
18
F-NaF PET/CT can also be used in patients with bone affected by the “flare phenomenon.”
metastases from prostate cancer for evaluating the response
to therapy, with better performance than conventional bone
scintigraphy [4]. 35.1.4 [18F]FDG PET/CT
35.1.3.1 Patient Preparation Although not strictly tumor-specific, [18F]FDG PET/CT is cur-
Patients should be fully informed about the concept and rently the most widely available and widely used nuclear med-
technical performance of the examination. Overall, 18FNaF icine procedure to identify primary and metastatic cancers.
PET/CT has the same preparation as bone scintigraphy. The usefulness of [18F]FDG is based on the increased
anaerobic glucose metabolism present in most tumors. In the
35.1.3.2 Administered Radioactivity early stages of tumor growth, prostate cancer is not charac-
18
F-NaF is injected intravenously by direct venipuncture or terized by a significant increase in the metabolic activity on
intravenous catheter as follows: adults, 1.5–3.7 MBq/kg (40– [18F]FDG PET/CT imaging, thus reducing the sensitivity of
100 μCi/kg). A maximum recommended activity (370 MBq) [18F]FDG imaging in this phase of the disease [5].
should be considered for obese patients. Low-grade prostate carcinoma has a low metabolic rate
and low levels of the glucose transport protein (Glut-1) and,
35.1.3.3 Image Acquisition consequently, relatively low levels of anaerobic glucose
In patients with normal renal function, acquisition of the metabolism and [18F]FDG uptake. Conversely, when the
axial skeleton may begin about 30–45 min after administra- tumor is more aggressive (higher Gleason scores), glucose
tion of the radiopharmaceutical, due to the fast blood clear- utilization is increased, and [18F]FDG uptake is increased
ance and rapid skeletal uptake of 18F-NaF. However, it is [6], thus indicating a poor prognosis.
necessary to wait longer to obtain high-quality images of the In patients with the so-called biochemical failure or PSA
extremities, with a start time of 90–120 min for imaging of relapse, that is, rising PSA after surgical or radiation pros-
the upper and lower extremities. tatectomy, the tumor is more likely to increase glycolysis,
PET images may be acquired in 2D or 3D mode. The 3D allowing [18F]FDG PET/CT to identify sites of recurrence
mode (which is becoming the only modality that manufactur- with a reasonable degree of accuracy [6].
ers of PET/CT scanners are selling) is recommended for whole-
body acquisition, because the higher count rates compensate for
the shorter acquisition times required for imaging a large area. 35.1.5 [11C]Choline and 18F-Choline PET/CT
Acquisition time per bed position varies depending on
the amount of injected radioactivity, uptake time, body mass In initial investigations, choline was labeled with 11C
index, body habitus of the patient, and camera factors. ([11C]CHO), therefore making this radiolabeled choline
Fig. 35.5 Comparison between

PET with 18F-fluoride (MIP view
in left panel) and bone
scintigraphy with 99mTc-HMDP
(anterior and posterior whole-
body scans in right panel) in a
patient with recurrent prostate
cancer after radiotherapy, with
serum PSA rising from 2 to
38 ng/mL. There is a definite
mismatch between the two scans;
in fact, bone scintigraphy is
apparently negative for metastatic
lesions, while 18F-fluoride PET
shows multiple sites of increased
uptake in the skeleton
(reproduced from: Strauss HW,
SM, eds. Nuclear Oncology –
Ant Post
indistinguishable from natural choline. In vitro experiments The influence of androgen deprivation therapy (ADT) on
demonstrated that radiolabeled choline is incorporated into choline uptake in patients with prostate cancer disease has not
tumor cells by active transport and subsequently phosphorylated been clarified yet. Some studies suggest that ADT significantly
in the cells and incorporated into membrane phospholipids. reduces [11C]choline uptake in androgen-sensitive prostate
Both [11C]choline and 18F-choline (18F-FCH) enter phos- cancer patients [11]. Nevertheless, most patients continue their
pholipid and membrane metabolism; choline transport, antiandrogen therapy prior to choline-PET/CT acquisition.
phosphorylation, and membrane metabolism are upregulated
in prostate cancer. Choline is phosphorylated to phosphoryl- 35.1.5.2 Administered Radioactivity
choline by the enzyme choline kinase; in this chemical form, [11C]CHO PET/CT imaging protocols are based on the i.v.
it is trapped within the cancer cells. bolus injection of 370–740 MBq. The effective radiation dose
(ED) is estimated to be approximately 0.3 Sv, and the critical
35.1.5.1 Patient Preparation organ is the liver. In fact, [11C]CHO is primarily metabolized
For both of these tracers, fasting is not strictly necessary, and excreted through the liver, which is advantageous since
although many studies suggest fasting for at least 4–6 h there is little radioactivity in the urinary tract. The 20-min
before tracer injection and PET scanning. Patients should be half-life of 11C limits its use to sites with a cyclotron and PET
well hydrated. radiochemistry facilities.
For 18F-CHO PET/CT imaging, patients receive an i.v.

bolus injection of 4.1 MBq/kg.
35.1.5.3 Protocol/Image Acquisition

The PET/CT scan for [11C]CHO PET/CT starts 3–5 min
after intravenous tracer injection, and images are acquired
for a total of 10–20 min with 2–5 min per bed posi-
tion, depending on the PET camera. For 18F-FCH PET/
CT, whole-body imaging starts 60–90 min after tracer
administration.
Occasionally, for prostate bed evaluation, early static or
dynamic pelvic images may be obtained, starting 2 min after
tracer administration.
The patient is positioned with the arms elevated and sup-
ported above the head. If the patient cannot tolerate this
position, one arm can be kept above the head with the other
positioned alongside the body, or both arms can be posi-
tioned alongside and close to the body. Axial anatomical
scan coverage range from the base of the skull to the mid-
thigh is sufficient.
35.1.5.4 Interpretation
Physiologic uptake of radiolabeled choline is noted in the
salivary glands, liver, pancreas, as well as renal parenchyma
and urinary bladder; faint uptake is seen in the spleen, bone
marrow, and muscles, whereas bowel activity is variable
(Fig. 35.6).
35.1.5.5 Clinical Applications of 18F-FCH PET/CT Fig. 35.6 Three-dimensional surface volume rendering of [11C]choline
The clinical use of 18F-FCH PET/CT imaging has been inves- PET/CT, showing an inhomogeneous hypermetabolic mass in the pel-
tigated for a variety of indications in prostate cancer patients: vis and markedly increased [11C]choline uptake in pelvic and abdomi-
nal lymph nodes. The image also serves to illustrate the pattern of
physiologic biodistribution of radiolabeled choline in the body (repro-
• Initial staging of prostate cancer disease (high-risk pros- duced from: Strauss HW, Mariani G, Volterrani D, Larson SM, eds.
tate cancer patients) Nuclear Oncology – From Pathophysiology to Clinical Applications,
18
F-FCH PET/CT is a useful imaging modality to char- 2nd Edition. New York: Springer, 2017)
acterize the extent of disease in the entire body, both soft
tissue and the skeleton, especially in high-risk prostate patients with higher PSA, higher initial Gleason score,
cancer patients. Nevertheless, some choline uptake in and shorter PSA doubling time [12].
inflamed tissue in addition to abnormal uptake in cancer • Localization of prostate cancer (elevated PSA level and
cells results in suboptimal specificity. Similarly as with negative biopsy)
the use of [11C]choline (Fig. 35.7), 18F-FCH PET/CT is The false-negative biopsy rate in patients with prostate
useful in the initial staging of patients with biopsy-proven cancer is 20% [13]. In addition to prostate carcinoma,
high-risk prostate cancer. radiolabeled choline uptake can be increased also in
• Restaging of prostate cancer disease (biochemical evi- benign changes such as prostatitis and prostatic hypertro-
dence of recurrence) phy. A study of 20 patients with elevated PSA level and
At present, the best indication for [11C]CHO or 18F- negative biopsy showed that 18F-FCH PET/CT correctly
FCH PET/CT consists in detecting recurrence/metastasis identified prostate malignancy in 25% of the patients [14].
of prostate carcinoma in patients with biochemical recur- This low specificity is the main limitation for the use of
rence, since this imaging modality provides information radiolabeled choline to identify primary prostate
about locoregional as well as distant sites of recurrence carcinoma.
(Fig. 35.8). • Treatment monitoring of prostate cancer disease
• In patients with recurrent prostate cancer disease, the In patients receiving antiandrogen treatment, conven-
overall sensitivity of 18F-FCH PET/CT is higher for those tional imaging modalities very often do not show any
a b
c d
Fig. 35.7 Fused axial images from PET/CT with [11C]choline per- nent also shows grossly enlarged right iliac lymph nodes. (d) Markedly
formed for staging in a patient with high-risk prostate cancer (Gleason increased [11C]choline uptake in para-aortic lymph node (reproduced
score 4 + 5 = 9 and serum PSA >600 ng/mL). (a) Enlarged prostate with from: Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear
inhomogeneous tracer uptake within the mass. (b) Cranial portion of Oncology – From Pathophysiology to Clinical Applications, 2nd
the prostate lesion. (c) The left internal iliac and right pararectal lymph Edition. New York: Springer, 2017)
nodes are enlarged and show increased tracer uptake; the CT compo-
Fig. 35.8 PET/CT with 18F-CHO in a patient with biochemical recur- nal area and at more distant site (mediastinum and left supraclavicular);
rence of prostate cancer after radical prostatectomy and pelvic lymph a metastatic lesion is also observed in a thoracic vertebra (reproduced
node dissection plus radiotherapy (fused PET/CT images in left panel, from: Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear
MIP image in right panel). About 16 months posttreatment, serum PSA Oncology – From Pathophysiology to Clinical Applications, 2nd
rose from undetectable to 5.7 ng/mL. Progression of disease is shown Edition. New York: Springer, 2017)
by increased tracer uptake in multiple lymph nodes both in the abdomi-
October 2012 November 2014 May 2015 October 2015 Jan 2016
Radical
prostatectomy
(GS: 8; pT4N1)
Therapy Serum PSA (ng/mL)
3.6 27 1.01 0.23
Bicalutamide Enantone Abiraterone
Fig. 35.9 Sequential 18F-choline PET/CT scans (MIPI images) per- lymph nodes can be observed particularly after the start of treatment
formed to assess response in a patient with prostate cancer showing with abiraterone acetate (reproduced from: Strauss HW, Mariani G,
progressively increasing serum PSA levels during androgen deprivation Volterrani D, Larson SM, eds. Nuclear Oncology – From
therapy (bicalutamide). The patient was then switched to therapy with Pathophysiology to Clinical Applications, 2nd Edition. New York:
enantone and then again to abiraterone acetate for further increasing Springer, 2017)
serum PSA levels. A significant reduction in 18F-choline uptake in the
significant morphologic changes in metastatic lymph especially in the stroma adjacent to neovasculature of solid
nodes or other sites of tumor metastasis. In these patients tumors. Due to its selective overexpression in 90–100% of
18
F-FCH PET/CT has the potential to demonstrate meta- prostate cancer lesions (either the primary tumor or metas-
bolic response to hormonal therapy. tases in the lymph nodes and bone), PSMA is a reliable tis-
18
F-FCH PET/CT is useful for the detection of pros- sue marker for prostate carcinoma and is considered an ideal
tatic recurrence, lymph node, and skeletal lesions espe- target for theranostic applications. Increased PSMA expres-
cially in high-risk prostate cancer patients with PSA value sion is correlated with high tumor grade, pathological stage,
>4 ng/mL during antiandrogen therapy [15] (Fig. 35.9). aneuploidy, and biochemical recurrence. Most importantly,
• Radiotherapy planning of prostate cancer disease PSMA expression is upregulated when tumors become
Most patients with prostate cancer have one or two androgen-independent, in up to 100% of the cases [17].
dominant intraprostatic lesions. Due to regional toxicity One of the earliest molecular imaging probes specifically
of external beam radiation therapy, accurate definition of targeting PSMA was 111In-capromab pendetide (Prostascint®),
volumes is crucial for focal radiotherapy, in primary or an 111In-labeled anti-PSMA antibody. However, capromab
recurrent prostate cancer disease; in this scenario radiola- pendetide suffers from an important limitation, i.e., it binds
beled choline PET/CT may play a role. to an intracellular domain of PSMA; therefore, capromab
pendetide either binds to viable tumor cells following inter-
nalization or to dying cells with disrupted cell membranes.
35.1.6 68Ga-PSMA-Ligand PET/CT High-affinity antibodies against extracellular epitopes of
PSMA have also been developed, such as J415, J533, and
The prostate-specific membrane antigen (PSMA, also known J591. The major disadvantages that limit the use of these
as glutamate carboxypeptidase II, or NAALADase) is an radiolabeled monoclonal antibodies as theranostic radio-
integral membrane glycoprotein present in all prostatic tis- pharmaceuticals include their relatively long half-life in cir-
sues. Although increased PSMA expression can be seen in culating blood (3–4 days), poor tumor penetration, and low
a variety of malignancies, it most notably occurs in prostate tumor-to-normal tissue ratios, especially at early time points.
cancer [16]. It has been suggested that PSMA is involved These drawbacks would be overcome by the use of small
in angiogenesis, as increased PSMA expression was found molecules that, contrary to antibodies, exhibit rapid diffusion
in the extravascular space and faster blood clearance, thus The majority of published data is based on an injected activ-
resulting in high tumor-to-normal tissue contrast early after ity of approximately 1.8–2.2 MBq/kg of body weight.
tracer injection. Flushing of the administration syringe should be done
In search for PSMA tracers with such favorable character- with at least the same volume of physiological saline, and
istics, modified forms of NAALAdase inhibitors have been subsequent emptying into the i.v. access is recommended to
evaluated for their potential to diagnose and treat prostate maximize use of dispensed activity.
cancer. A series of preclinical studies explored the potential
of radiolabeled small-molecule PSMA-inhibiting ligands for 35.1.6.3 Protocol/Image Acquisition
imaging of human prostate carcinoma using various radio- A 60-min interval is recommended for uptake time, with
nuclides such as 11C, 18F, 123I, 99mTc, and 68Ga. Among these an acceptable range of 50–100 min. Increased lesion
radionuclides, 68Ga-labeled peptides have attracted consid- detection has been reported with delayed imaging up to
erable interest for cancer imaging because of the favorable 3–4 h after injection [18]. The patient is positioned with
physical characteristics of 68Ga and the availability of reli- the arms elevated and supported above the head to avoid
able GMP-compliant 68Ge/68Ga generators. Moreover, the beam-hardening artifacts in the abdominal and pelvic
half-life of 68Ga matches the pharmacokinetics of the small regions, as well as artifacts caused by truncation of the
PSMA-inhibiting peptides. 68Ga decays with a half-life of measured FOV. If the patient cannot tolerate this posi-
67.63 min, with 89% yield by positron emission. Several tion, one arm can be kept above the head with the other
low-molecular-weight ligands for human PSMA, coupled positioned alongside the body, or both arms can be posi-
with a chelator for conjugation to 68Ga, are clinically avail- tioned alongside and close to the body. PET acquisition
able for PET/CT imaging. should start from the mid-thigh to the base of the skull, in
order to exploit the reduced 68Ga-PSMA-ligand uptake in
35.1.6.1 Patient Preparation the urinary system after pre-scan voiding; PET scans are
Patients do not need to fast and are allowed to take all acquired in 3D mode, usually with an acquisition time of
their medications; patients should be well hydrated before 2–4 min per bed position.
the study and during the uptake time (e.g., oral intake of
500 mL of water during a 2-h period prior to acquisition). 35.1.6.4 Interpretation
Bladder voiding immediately before image acquisition is Physiologic PSMA-ligand uptake of variable degree can
recommended. be observed in the following tissues: lacrimal glands, sali-
vary glands, liver, spleen, small intestine, colon, and kidney.
35.1.6.2 Administered Radioactivity Usually, tumor lesions inside and outside the prostate gland
The 68Ga-PSMA ligand is injected as an intravenous bolus. show a very high tumor-to-background ratio compared to the
Currently, the optimal injected activity is still under debate. surrounding tissues (Fig. 35.10).
Fig. 35.10 Axial image of the

pelvis (left panels) and MIP
image (right panel) from PET/CT
with 68Ga-PSMA-ligand uptake
in a patient with prostate cancer
previously treated with
radiotherapy, but then showing
progressively increasing serum
PSA levels. There is a distinct
focus of increased tracer uptake
at the right posterolateral
peripheral zone of the prostate
extending from the mid-gland
down to the apex (reproduced
Immunohistochemistry and 68Ga-PSMA-ligand PET data • Primary staging in high-risk disease before surgical pro-
demonstrate that increased PSMA expression can also be cedures or planning of external beam radiation therapy
found in the neovasculature of non-prostate cancers such as In patients with high-risk disease (Gleason score >7,
colon cancer, esophageal cancer, thyroid cancer, lung cancer, PSA >20 ng/mL, clinical stage T2c–3a), the likelihood of
renal cell carcinoma, brain tumors, as well as in benign tis- lymph node and bone metastases is particularly high.
sue [19, 20]. Several studies demonstrate the superiority of
An important pitfall is a relevant 68Ga-PSMA-ligand 68
Ga-PSMA-ligand PET/CT as compared to CT, mag-
uptake in celiac ganglia of the autonomic nervous system, netic resonance imaging (MRI), or bone scintigraphy for
which can be misinterpreted as retroperitoneal lymph node detection of metastases for initial staging after primary
metastases [21]. Moreover, uptake in Paget’s bone disease, diagnosis [24, 25].
in stellate and other ganglia as well as in the ribs (without Nevertheless, the 68Ga-PSMA-ligand scan cannot replace
morphologic correlation on CT), may cause misinterpreta- pelvic MRI for local tumor delineation (as needed by the
tion [22]. surgeon). It is still debated whether additional functional
imaging with bone-seeking agents (e.g., bone scintigraphy,
35.1.6.5 Appropriateness of Use Criteria 18
F-NaF PET/CT) has a relevant added value after performing
• Localization of recurrent prostate cancer a 68Ga-PSMA-ligand PET/CT, e.g., in patients with PSMA-
68
Ga-PSMA-ligand PET/CT is especially recom- negative tumors or densely sclerotic bone lesions (Fig. 35.11).
mended in patients with measurable but low PSA (values
between 0.2 and 10 ng/mL) to identify the site of recur- 35.1.6.6 E merging Clinical Applications
rence and to potentially guide salvage therapy. Higher sen- • Staging before and during PSMA-directed radiotherapy
sitivities are noted in patients with shorter PSA doubling (mainly in metastatic castration-resistant prostate
times and those with higher initial Gleason scores [23]. cancer)
Fig. 35.11 A 73-year-old patient with newly diagnosed prostate can- suspicious pelvic lymph nodes (T3aN0M1). 68Ga-PSMA-ligand PET/
cer: Gleason score 4 + 5 = 9, serum PSA 20 ng/mL. The staging CT CT detects a metastatic left pelvic lymph node and excludes liver
scan shows a hypodense liver lesion suspicious for metastasis and no metastasis (T3aN1M0)
Imaging before PSMA-directed therapy (e.g., radioli- tion; the main safety issues are standard radiation safety
gand therapy) is crucial to determine the presence and precautions that are inherent in all intravenously injected,
intensity of expression of the target molecule [26–28], as renally excreted radionuclide therapies.
low PSMA expression in target lesions contraindicates Injected activities have ranged from 3 to 8 GBq per sin-
radioligand therapy. gle administration with up to six injections given to men,
• Targeted biopsy after previous negative biopsy in patients generally at a minimum 6-week intervals; dose calculations
with high suspicion of prostate cancer depend on a combination of disease burden, patient weight,
Preliminary data indicate that 68Ga-PSMA-ligand and renal function. 177Lu-PSMA ligand is administered
PET/CT is of value for guidance of repeat biopsy in by a slow intravenous injection (30–60 s) in a volume of
patients with high suspicion of prostate cancer and prior 5 mL (diluted with 0.9% sterile sodium chloride solution),
negative biopsies. In fact, the scan provides useful infor- followed by a flush of sterile 0.9% sodium chloride. It is
mation to localize the primary prostate cancer [29], there- recommended that the patients are hydrated pre- and post-
fore to direct biopsies in the prostate cancer surveillance administration of 177Lu-PSMA ligand with 1–1.5 L of water
population who undergo repeat biopsies. and encouraged to void as frequently as possible.
• Monitoring of systemic treatment in metastatic prostate 177
Lu PSMA is excreted in the urine in the first 48 h
cancer following injection, and instructions need to be given to
Conventional monitoring systems such as RECIST 1.1 patients, staff, and families on managing potential radioac-
are limited by the high prevalence of non-measurable tive spills. Given the rapid renal excretion of 177Lu-PSMA
lesions, e.g., lymph node and bone metastases. Bone scin- ligand, patients are required to remain in the nuclear medi-
tigraphy is limited by a potential flare phenomenon. cine department for 2–4 h for observation and for measured
68
Ga-PSMA-ligand PET/CT has the potential to monitor radiation levels to decrease.
systemic disease, although it has not yet been definitely There are currently a limited number of published tri-
demonstrated that this imaging modality overcomes the als in the use of 177Lu-PSMA ligand for the treatment of
limitations of other modalities. metastatic castrate-resistant prostate cancer (mCRPC) [30].
Although these are generally small, retrospective trials, they
35.1.6.7 Emerging Therapeutic Applications have shown almost uniformly that most men treated with
Prostate cancer is an ideal target for the development of tar- 177
Lu-PSMA ligand have a significant response to treatment.
geted radionuclide therapy because of the frequent occurrence 177
Lu-PSMA ligand is showing promising responses in
of multifocal disseminated disease. In particular, recurrence men with mCRPC and almost certainly has an important
of a prostate carcinoma initially classified as confined to the future role in the treatment of prostate cancer. Preliminary
prostate gland and frequent presentation of the cancer in an publications suggest it has a low toxicity profile and appears
advanced stage preclude curative surgery and radiation therapy. generally well tolerated in men with end-stage metastatic
Furthermore, chemotherapy does not have long-term efficacy disease. Prospective randomized trials are needed to deter-
in patients with metastatic castration-resistant prostate cancer. mine its impact on survival and to rigorously assess its clini-
Radiolabeled molecules with high binding affinity for cal benefit compared to other treatments of prostate cancer,
PSMA constitute the basis for targeted radionuclide therapy, including chemotherapy, external beam radiotherapy, and
with the aim of delivering sufficient radiation doses to tumor androgen blockade.
lesions while sparing the adjacent normal tissues.
Lutetium-177 (177Lu) has gained popularity as the thera-
peutic radionuclide of choice due to its desirable physical 35.1.7 Perspectives for 18F-FACBC PET/CT
properties; 177Lu is a β− emitter (78% abundance of 490 keV)
with a maximum tissue penetration of about 2 mm; this The US FDA has recently approved a synthetic l-leucine ana-
somewhat short β− range of 177Lu provides better irradiation logue (trans-1-amino-3-18F-fluorocyclobutane-1-carboxylic
of small tumors. 177Lu is a reactor-produced radiometal that acid or 18F-FACBC) for PET imaging in patients with pros-
emits gamma rays at 208 and 113 keV with 11% and 6% tatic cancer. Uptake in prostatic cancer cells of this agent
abundance, respectively. The gamma emission from 177Lu (also denominated 18F-fluciclovine) is mediated by neutral
allows for external gamma camera imaging and consequently amino acid sodium-independent systems, through transport-
the collection of information pertaining to tumor localiza- ers associated with proliferation and angiogenesis [31–33].
tion and dosimetry. Furthermore, 177Lu has a relatively long High physiologic uptake of 18F-FACBC is noted in the liver,
physical half-life of 6.73 days. It is these physical properties red bone marrow, lung, and pancreas [34–36]. 18F-FACBC
that allow for the delivery of high activities of 177Lu-PSMA shows a somewhat favorable biodistribution versus radiola-
ligand to prostate cancer cells. beled choline [37] (Fig. 35.12). In fact, this agent undergoes
177
Lu-PSMA ligand is an easily administered targeted minimal excretion through the kidneys and urinary tract, thus
therapy with no significant symptoms at the time of injec- resulting in improved detection of small lesions in the prostatic
Fig. 35.12 MIP images from

PET/CT with 18F-FACBC (left)
SUV = 9,7 SUV = 19,5
and with [11C]choline (right),
showing comparative
biodistribution patterns of the
two tracers (reproduced from:
Nanni C, Schiavina R, Boschi S,
Ambrosini V, Pettinato C,
Brunocilla E, et al. Comparison
of 18F-FACBC and 11C-choline
PET/CT in patients with radically
treated prostate cancer and
biochemical relapse: preliminary
results. Eur J Nucl Med Mol
Imaging. 2013;40(Suppl
1):S11–7)
SUV = 0 SUV = 0
bed and pelvis. Furthermore, PET images can be acquired as

early as 4–5 min post-administration, due to its very fast blood
clearance and very rapid uptake in prostatic cancer cells. bisphosphonates, with the advantage of better sen-
Sensitivity of PET with 18F-FACBC in patients with sitivity and specificity.
prostate cancer is directly correlated to PSA concentrations • The role of [18F]FDG PET/CT is limited to the most
and inversely correlated to PSA doubling time; however, aggressive prostate cancers, with higher Gleason
the uptake of 18F-FACBC is similar in prostatic cancer as in scores and biochemical failure or PSA relapse.
benign prostatic hyperplasia, thus resulting in low specificity • PET/CT with [11C]CHO or 18F-FCH is most useful
[38]. Nevertheless, the combination of 18F-FACBC PET/CT for detecting prostate cancer recurrence in patients
with MRI appears to be optimal for accurate localization of with biochemical recurrence and, to a lesser extent,
cancer [37], either in the prostatic bed or at distant metastatic for the detection of prostatic recurrence, lymph
sites [39]. node, and skeletal lesions during staging, especially
Studies comparing 18F-FACBC PET/CT with [11C]choline in patients with high-risk prostate cancer.
PET/CT found that the detection rate of tumor lesions with • 68
Ga-PSMA-ligand PET/CT is one of the most
18
F-FACBC PET/CT was superior to that of [11C]choline promising imaging procedures for primary staging
PET/CT, regardless of the PSA levels [40, 41] (Fig. 35.13). in high-risk disease and for monitoring the efficacy
Further, larger-scale studies are necessary to confirm the of systemic treatment. It is currently considered the
potential clinical usefulness of PET/CT with 18F-FACBC best available radionuclide imaging modality for
for the primary diagnosis of prostatic cancer, as well as for the evaluation of patients with biochemical failure.
staging and restaging prostate cancer patients after poten- • In patients with a positive 68Ga-PSMA-ligand PET/
tially curative treatment. CT scan, targeted radionuclide therapy with
177
Lu-PSMA ligand ensues promising responses in
the treatment of metastatic castration-resistant pros-
tate cancer.
Key Learning Points • Promising results have been obtained with the syn-
• Bone scintigraphy with 99mTc-bisphosphonates is thetic l-leucine analogue (trans-1-amino-3-18F-flu-
still considered the cornerstone for detection of orocyclobutane-1-carboxylic acid or 18F-FACBC)
bone metastases in prostate cancer patients due to for PET/CT imaging in patients with prostatic
wide availability, good sensitivity, low radiation cancer.
dose, and low cost. The specificity of bone scintig- • Although characterized by low specificity because
raphy can be dramatically improved by using of 18F-FACBC uptake also in benign prostatic
SPECT or, better, SPECT/CT. hyperplasia, this agent displays biodistribution and
• Clinical indications and interpretation criteria for uptake characteristics that constitute potential
18
F-sodium fluoride (18F-NaF) PET/CT are similar advantages versus imaging with radiolabeled cho-
to those for bone scintigraphy with 99mTc- line-based PET tracers.
Fig. 35.13 Axial PET and fused PET/CT images in patients with cor- C, Schiavina R, Boschi S, Ambrosini V, Pettinato C, Brunocilla E, et al.
responding positive findings (indicated by white arrows) on the [11C] Comparison of 18F-FACBC and 11C-choline PET/CT in patients with
choline scans (left panels) and on the 18F-FACBC scans (right panels), radically treated prostate cancer and biochemical relapse: preliminary
respectively. TBR target-to-background ratio (reproduced from: Nanni results. Eur J Nucl Med Mol Imaging. 2013;40(Suppl 1):S11–7)
35.2 Testicular Cancer Seminomas, which represent approximately 50% of all

germ cell tumors, are very radiosensitive. The cure rate for
35.2.1 Clinical Background patients with seminoma (all stages combined) exceeds 90%.
NSGCT tend to metastasize early to the retroperitoneal
Testicular cancer is a relatively rare malignancy comprising lymph nodes and to the lungs. Tumors with histologic mix-
1% of all the diagnosed cancer cases. Although rare, it is the tures of seminoma and non-seminomatous components are
most frequent solid non-hematological cancer diagnosed in treated as NSGCT. Initial work-up is testicular ultrasound
young adult men, and its incidence is increasing. The most followed by measurements of serum tumor-associated bio-
common presentation is a painless, hard, testicular lump. markers, such as alpha-fetoprotein (AFP). MRI is especially
Histologically, testicular tumors can be of germ cell origin useful as a problem solver in equivocal cases. Trans-scrotal
or non-germ cell origin. Germ cell tumors (GCT) are the needle biopsy or orchiectomy is contraindicated because
most common testicular cancers, further divided into semi- these procedures can potentially lead to nonperitoneal lym-
nomas and non-seminomatous germ cell tumors (NSGCT). phatic dissemination of the tumor cells. CT is the primary
imaging modality for assessing lymph node status and for management plan for surveillance versus adjuvant che-
evaluation of distant metastasis. Since bone metastases are motherapy [43]. Eight of the 11 patients with equivocal
rare in testicular cancers, bone scintigraphy is not indicated. CT findings had a true-negative [18F]FDG PET/CT scan,
TNM staging is conventionally used for testicular cancer. while 3 patients had a true-positive [18F]FDG PET/CT
Accurate initial staging is very important to classify patients scan. These findings are consistent with previous reports
into low- or high-risk groups, because management differs leading to the conclusion that [18F]FDG PET/CT is most
between the two groups. Accurate restaging post-chemo- useful in patients with equivocal CT scans [44, 45]. The
therapy is also important to stratify patients for further man- five high-risk patients with a negative CT scan had nega-
agement plans. tive [18F]FDG PET/CT scan, but two of them subsequently
relapsed. The authors concluded that [18F]FDG PET/CT
is helpful in primary staging when the primary staging CT
35.2.2 [18F]FDG PET/CT in Testicular Cancer scan is equivocal, but it is not sensitive enough to predict
occult metastases in high-risk patients with a negative CT
35.2.2.1 Indications scan [43].
• Diagnosis and staging • Elevated serum tumor markers
There is at present no sufficient evidence justifying to [18F]FDG PET/CT plays a clinically relevant role in
recommend the use of [ F]FDG PET/CT for primary
18
GCT patients who exhibit elevated serum levels of tumor
diagnosis or staging of testicular cancer, although the markers associated with a negative CT scan, since in a
results of the scan are frequently accurate and useful for large percentage of these patients PET imaging detects
planning the most adequate treatment strategy. In a recent sites of the disease [46]. In a series of 47 [18F]FDG PET/
meta-analysis of 16 articles including an overall 957 [ F]
18
CT scans obtained for assessing residual masses (18 with
FDG PET scans in 807 patients, pooled sensitivity and raised serum tumor markers) and 23 scans for investigat-
specificity of diagnostic [18F]FDG PET/CT were 75% and ing patients with raised serum markers associated with
87%, respectively [42]. This sensitivity, with a negative normal CT findings, it was found that all but one of the
likelihood ratio of 0.31, is not sufficient to confidently [18F]FDG PET/CT scans identified the site of disease
exclude testicular malignancy and metastases. False- [46]. In this series, however, the NPV was as low as 50%,
negative studies are mostly due to the partial volume thus indicating that a negative scan did not reliably pre-
effect of small lesions and microscopic tumor foci in sem- dict the absence of disease. On the other hand, subsequent
inoma and NSGCT; in addition, teratoma components of [18F]FDG PET/CT imaging in those patients with nega-
NSGCT do not demonstrate significant metabolic activity tive scans was frequently the first imaging modality that
on [18F]FDG PET/CT [42]. Furthermore, false-positive identified the site of disease. Therefore, in the presence of
findings have occasionally been reported, due to infec- raised serum tumor marker levels and negative imaging
tive/inflammatory processes or to inadequate timing post- findings, including a negative [18F]FDG PET/CT, the
chemotherapy [42]. most appropriate follow-up imaging may be to repeat
• Equivocal findings on the CT scan [18F]FDG PET/CT [46].
As a functional/metabolic imaging procedure, [18F] • Post-chemotherapy residual seminoma GCT masses
FDG PET/CT offers in principle several advantages over Post-chemotherapy residual masses and metastatic sem-
CT. Since detection of metastatic lymph node disease at inoma occur in approximately 25% of the cases [47]. [18F]
CT is based on the size criterion (>1 cm diameter of short FDG PET/CT is indicated for evaluation of post-
axis) and morphology (round lymph nodes >0.8 cm), this chemotherapy residual masses and metastatic seminoma.
imaging modality lacks the desired accuracy for charac- This is currently the only indication for [18F]FDG PET/CT
terizing lymph nodes, and sensitivity can be low as 36%. in testicular cancer endorsed by NCCN guidelines, particu-
In a recent study by Cook et al., [18F]FDG PET/CT was larly when the residual mass exceeds 3 cm in maximal
used at initial staging of 16 patients (10 with seminoma, 5 diameter. The prospective SEMPET study shows that [18F]
with NCGCTs, 1 with mixed GCT). Eleven patients FDG PET/CT can confidently distinguish residual disease
underwent [18F]FDG PET/CT to clarify equivocal find- from posttreatment fibrosis [48]. Several other studies have
ings on the staging CT scan, usually consisting of para- shown the superiority of [18F]FDG PET/CT compared with
aortic lymph nodes that were in the expected drainage CT in predicting viable tumor in seminoma residual masses
from the primary tumor (such as the retroperitoneal after chemotherapy, with accuracy >90% [49, 50].
basin), but were less than the 1 cm cutoff size value A retrospective validation of the SEMPET trial includ-
required to be classified as metastasis on CT; in addition, ing 127 patients yielded sensitivity, specificity, negative
five high-risk patients with a negative primary staging CT predictive value, and positive predictive values of [18F]
scan underwent [ F]FDG PET/CT to aid decisions on
18
FDG PET/CT at 50%, 77%, 91%, and 25%, respectively
Column A Column B Column C
Fig. 35.14 Trans-axial sections at the abdominal level of a patient with metastatic lymphadenopathy. Column B: post-chemotherapy scan dem-
metastatic seminoma submitted to repeat [18F]FDG PET/CT scans for onstrating residual lymph node without significant [18F]FDG uptake,
assessment of response to chemotherapy (CT images in top row, PET indicating complete response to treatment. Column C: follow-up scan
images in middle row, fused PET/CT images in bottom row). Column at 12 months confirming the absence of any significant metabolic activ-
A: baseline scan demonstrating a bulky hypermetabolic retroperitoneal ity, indicating long-lasting remission of the disease
[51]. Correct timing of PET scans is critical and may sub- • Post-chemotherapy residual NSGCT masses
stantially reduce false-positive results. [18F]FDG PET/CT [18F]FDG PET/CT is not routinely indicated for post-
should be performed not earlier than 6 weeks after the end chemotherapy restaging in NSGCT patients [52]. Residual
(day 21) of the last chemotherapy cycle [52]. masses after completion of first-line chemotherapy are
In order to reduce the post-chemotherapy false-posi- found in approximately 40% of patients, even after nor-
tive cases, it is currently recommended to repeat the [18F] malization of serum tumor markers [54]. Histology of the
FDG PET/CT scan after at least 6 additional weeks. resected lesions reveals necrosis in 40%, vital carcinoma
Repeating [18F]FDG PET/CT is a helpful strategy particu- in 20%, and mature teratoma in 40% of cases [55]. Mature
larly if residual disease is considered clinically unlikely; teratomas are completely chemoresistant; in addition,
it is also helpful to guide biopsy in those patients who they carry the risk of subsequent malignant transforma-
otherwise can also be falsely classified as negative. If the tion, and in that case they usually become [18F]FDG-avid.
repeat scan is not performed, the residual mass must then Therefore, surgical dissection is indicated for all residual
be biopsied [52, 53] (Fig. 35.14). NSGCT masses larger than 1 cm [52].
In a prospective trial including 121 stage IIC or III methods may be applied. The CT attenuation-corrected scan
NSGCT patients, [18F]FDG PET/CT was performed after data are reconstructed by filtered back projection or iterative
completion of chemotherapy, and the results were corre- algorithms available and reformatted in coronal, sagittal, and
lated with histopathology, CT scan, and serum tumor trans-axial slices.
markers [56]. [18F]FDG PET/CT predicted tumor viabil-
ity in only 56% of the cases, with approximately the same 35.2.2.6 Interpretation
accuracy as CT (55%) or tumor markers (56%). This pro- Images must be evaluated using software capable to display
spective, histology-controlled study demonstrated that fused PET and CT data and use an SUV scale. PET images
[18F]FDG PET/CT does not yield a clear additional clini- should be displayed with and without attenuation correction
cal benefit with respect to the standard diagnostic proce- to identify artifacts possibly caused by contrast agents, metal
dures (CT and serum tumor markers) for predicting tumor implants, and/or patient motion. The presence or absence
viability in residual masses [56]. of abnormal [18F]FDG accumulation on the PET images, in
On the other hand, [18F]FDG PET/CT can be used to combination with level of uptake and morphologic correla-
localize post-chemotherapy hypermetabolic tissue, there- tion, should be evaluated. The absence of tracer accumula-
fore serving as a guide for biopsy or surgical tion in anatomical abnormalities seen on the CT scan or other
intervention. imaging may be particularly significant. It is important to
• Active surveillance review these images on a workstation that has the capacity to
There is still no evidence to support the use of [18F] triangulate findings in axial, coronal, and sagittal planes. The
FDG PET/CT in the follow-up of testicular cancers dur- PET scan is deemed positive if there is increased [18F]FDG
ing active surveillance. uptake above the normal surrounding tissue in an area where
metastases could be found. Any lesions identified on the PET
35.2.2.2 Contraindications component are then correlated with the CT images, review-
None. ing soft tissue, lung, and bone windows as appropriate to the
location of the abnormality (Fig. 35.15). As a final step, the
35.2.2.3 Procedure CT images are sequentially reviewed using soft tissue, lung,
After a fast for at least 4 h and checking blood glucose and bone windows to identify structural abnormalities not
level to be less than 200 mg/dL, the patient is injected with previously identified on PET imaging.
2–4 MBq/kg of [18F]FDG intravenously. After the 60–90 min If the [18F]FDG PET/CT scan is negative in patients with
uptake period, the patient is positioned with the arms elevated seminomas and residual mass larger than 3 cm, no further
and supported above the head to avoid beam-hardening arti- treatment is needed. On the other hand, caution must be paid
facts in the abdominal and pelvic regions, as well as artifacts to NSGCT patients with residual masses detected on the CT
caused by truncation of the measured FOV. If the patient can- scan; lesions less than 1.0 cm may still harbor residual dis-
not tolerate this position, one arm can be kept above the head ease and therefore must be interpreted with caution. [18F]
with the other positioned alongside the body, or both arms FDG PET/CT has a limited NPV in NSGCT patients with
can be positioned alongside and close to the body. First, a residual masses, but still might be useful for guiding multi-
CT topogram is used to define the coaxial imaging range. modality treatment [57].
Axial anatomical scan coverage range from the base of the
skull to the mid-thigh is sufficient. A standard diagnostic CT 35.2.2.7 Reporting
scan with intravenous contrast agent may, if appropriate, be The PET/CT report should include the study identification,
performed. clinical information, procedure description, description of
the findings, and brief impression. When appropriate, the
35.2.2.4 Side Effects report should correlate PET/CT findings with those of other
None. diagnostic tests, interpret them in that context, and consider
them in relation to the clinical data.
35.2.2.5 Data Processing About 20% of patients with seminoma tumors have
The PET emission data must be corrected for geometrical metastases at the time of diagnosis, most often lymph node
response and detector efficiency (normalization), system metastases in the retroperitoneum and/or the supraclavicular
dead time, random coincidences, scatter, and attenuation. region. NSGCT often displays rapid spread, and about 50%
All the corrections necessary to obtain quantitative image of patients have metastases at the time of diagnosis [58].
data should be applied during the reconstruction process. The PET scan is deemed positive if there is increased [18F]
When available, time-of-flight information should be used FDG uptake above the normal surrounding tissue in an area
during reconstruction. Resolution modeling during recon- where metastases could be found. For both seminoma and
struction or other new reconstruction or image processing non-seminoma testicular cancer, metastasis occurs primarily
Fig. 35.15 MIP and axial

images of [18F]FDG PET/CT
scans performed in a young man
with history of metastatic
seminoma. In the baseline scan
(upper panels), there is a large
left mass with markedly
increased tracer uptake (arrows in
upper panels); the mass extends
from the lower border of T12 to
the lower border of L4, measures
8 × 7 × 12 cm, and has SUVmax
20 at the level of renal hilum.
The [18F]FDG PET/CT
performed after chemotherapy
(lover panels) demonstrates
excellent response to treatment,
with a residual hypometabolic
mass (arrow in lower panel)
(reproduced from: Strauss HW,
SM, eds. Nuclear Oncology –
to the retroperitoneal lymph nodes and subsequently further or subtropical regions of Latin America, Asia, and Africa.
along the thoracic duct. Metastasis can also occur through the Incidence of penile cancer is also affected by race and eth-
hematogenous route, especially in patients with NSGCT [59]. nicity, with the highest incidence (1.01 per 100,000) in North
American white Hispanics [61].
35.2.2.8 Future Perspectives Risk factors include phimosis, chronic inflammatory con-
• Initial reports indicate a predictive and prognostic role of ditions such as balanoposthitis and lichen sclerosus et atro-
[18F]FDG PET/CT for early interim metabolic assessment phicus, sexual history (multiple partners, early age at first
in patients with metastatic seminoma who are receiving intercourse), history of condylomata, and smoking. Human
chemotherapy [60]. These results imply the possibility to papillomavirus DNA has been identified in 70–100% of
personalize the number of chemotherapy cycles. intraepithelial neoplasms and in 40–50% of invasive penile
• [18F]FDG PET/CT has an established role in seminoma cancers. The virus plays an important role in oncogenesis
for evaluating whether further treatment is necessary after through interaction with oncogenes and tumor suppressor
chemotherapy for residual masses >3 cm. However, its genes [61].
impact on long-term disease outcome and cost- Ultrasonography (US) is useful for evaluating suspected
effectiveness still needs to be established. penile masses and is recommended as the first-line imaging
• PET probe-guided surgery might be useful in surgical modality for primary penile cancer deemed to invade the
oncology for recurrent testicular cancer in difficult-to- corpora cavernosa [62]. Squamous cell carcinoma (SCC)
access locations. usually presents as a hypoechoic lesion with heterogeneous
• The development of newer radiotracers with different appearance. However, US cannot reliably identify the pres-
metabolic pathways may be promising in the evaluation ence of microscopic invasion. On MR imaging, penile car-
of NSGCT. cinoma is often a solitary, ill-defined, and infiltrating mass
hypointense with respect to the adjacent corpora on both
T1- and T2-weighted images [63]. The local extent, depth
Key Learning Points of tumor invasion, and extension to adjacent structures can
• In patients with seminoma, [ F]FDG PET/CT has
18 be accurately depicted using MR, thus providing valuable
high specificity, sensitivity, and NPV for evaluating information for surgical planning [63].
post-chemotherapy residual masses, as PET imag- However, both clinical examination and conventional
ing is very effective in distinguishing residual tumor imaging techniques, including US, CT, and MRI, are unre-
from fibrosis. liable in accurately detecting lymph node metastases [62].
• Correct timing of the PET scan for assessing tumor Hence, prophylactic inguinal lymphadenectomy is often
response to therapy is crucial, as it may substan- recommended to improve survival, particularly in high-risk
tially reduce the false-positive results. [ F]FDG
18 patients. Nevertheless, the morbidity of groin dissection is
PET/CT should be carried out not earlier than high, with complication rates between 30% and 50%. This
6 weeks after the end (day 21) of the last chemo- has stimulated interest in the search for other techniques able
therapy cycle. to noninvasively assess the lymph node status and possibly
• [ F]FDG PET/CT can be helpful in patients with
18 reduce the need for groin dissection [64].
elevated serum levels of tumor markers, whether or
not a residual mass is seen on CT.
• In non-seminomatous tumors, sensitivity of [18F] 35.3.2 Bone Scintigraphy
FDG PET is hampered by a high rate of false-nega-
tive results, particularly in mature teratomas. Since bone metastases are rare in patients with penile cancer,
bone scintigraphy is not commonly used.
35.3 Penile Cancer 35.3.3 Sentinel Lymph Node Biopsy
35.3.1 Clinical Background 35.3.3.1 Indications

Sentinel lymph node biopsy for lymph node staging in penile
Penile carcinomas, rare in most developed nations includ- cancer is indicated in patients with non-palpable lymph
ing Europe and the United States (with an incidence <1 per nodes. The sensitivity of SLNB in patients with penile can-
100,000 men per year), are not infrequent in the tropical cer with non-palpable lymph nodes ranges between 70% and
86% [65, 66]. According to some authors, SLNB is indicated patient’s body is helpful for better anatomical localization.
only in G3 tumors in stages T2–T4 [67]. The value of lymphoscintigraphy can be greatly enhanced by
acquiring SPECT/CT imaging, usually 2 h postinjection; in
35.3.3.2 Contraindications this way the lymphoscintigraphic findings can be correlated
SLNB is not indicated patients with clinically palpable with regional morphology as assessed with the CT compo-
nodes, because of the possibility of false-negative results as nent of the scan. Such combined functional and morphologic
due to re-routing of lymphatic drainage when the primary imaging results in precise anatomo-topographic localization
sentinel lymph nodes are massively invaded by cancer cells. of sentinel lymph nodes and provides the surgeon with a
road map leading to faster harvesting of the sentinel nodes.
35.3.3.3 Technique SPECT/CT also helps to minimize false-positive findings in
Although the procedure can be performed in a single- or 2-day the event of contamination and can identify superimposed
protocol, the most recent trend favors the single-day proto- lymph nodes or lymph nodes adjacent to the injection site.
col, due to a lower false-negative rate than with the 2-day The sensitivity of SPECT/CT may be as high as 95% [69].
protocol [68]. The radiocolloid can be injected intradermally
around the tumor, or it can be injected at three sites along 35.3.3.4 Data Processing
the penile shaft circumferentially at the 2, 6, and 10 o’clock Skin projection of the sentinel lymph nodes visualized dur-
positions (20 MBq per aliquot for the single-day protocol ing lymphoscintigraphy can be identified with the aid of a
or 40 MBq for the 2-day protocol). Scintigraphic imaging 57
Co point source and marked with a felt-tip pen on the skin
is then acquired in different views starting at approximately surface. The SPECT/CT data can be utilized to reconstruct
1 h postinjection. When the 2-day protocol is adopted, addi- 3D surface volume rendering, using different filters in order
tional imaging is recommended just 1 h before surgery on the to emphasize location of the radioactive lymph nodes rela-
second day (Fig. 35.16). Body contouring with a transmis- tive to anatomic structures such as the skin, skeletal muscles,
sion scan using a 57Co planar source positioned beneath the or bones.
a b c
d e f
Fig. 35.16 Main phases of lymphoscintigraphy and intraoperative images. (e) Intraoperative use of the handheld gamma probe for sentinel
sentinel lymph node detection in a patient with penile cancer, involving lymph node localization. (f) Localization of the “hot” sentinel lymph
both gamma-probe guidance and optical guidance with blue dye. (a) node (white arrow) and visualization of in-transit blue dye (reproduced
Interstitial injection of 99mTc-nanocolloidal albumin proximally to the from: Strauss HW, Mariani G, Volterrani D, Larson SM, eds. Nuclear
tumor. (b) Imaging with dual-head gamma camera. (c) Planar imaging: Oncology – From Pathophysiology to Clinical Applications, 2nd
early (top) and delayed (bottom), showing lymphatic drainage to both Edition. New York: Springer, 2017)
right and left inguinal lymph nodes. (d) Fused SPECT/CT and MIP
35.3.3.5 Data Interpretation and Reporting 35.3.4.2 Contraindications

Interpretation of SPECT/CT images requires expertise in None.
both nuclear medicine and radiology, and special attention
must be paid to the interpretation of hybrid images. 35.3.4.3 Technique
Special precautions must be taken to avoid contamination
35.3.3.6 Future Perspectives of the field of interest with radioactive urine. In this regard,
A combined multimodal approach using lymphoscintigra- insertion of an indwelling bladder catheter is advised, espe-
phy with SPECT/CT and ultrasonography has recently been cially if evaluation of the primary tumor is required. After
suggested for use in patients with palpable lymph nodes. By the i.v. administration of [18F]FDG, some authors advocate
enabling to identify also massively metastatic lymph nodes, the i.v. administration of 20 mg furosemide to accelerate the
this procedure markedly reduces the rate of false-negative renal clearance of [18F]FDG and 20 mg butylscopolamine to
SLNB and results in less morbidity than de novo inguinal decrease intestinal activity. During acquisition, the catheter
lymphadenectomy [65]. is fixed with adhesive tape so that the penis is oriented in the
upright position.
35.3.4 PET/CT with [18F]FDG 35.3.4.4 Data Interpretation and Reporting

The diagnosis of penile carcinoma is usually established
35.3.4.1 Indications by visual inspection and tissue biopsy. Therefore, [18F]
[18F]FDG PET/CT, which is recommended for staging penile FDG PET/CT does not play a clinical role for the primary
cancer patients with palpable lymph nodes, can replace diagnosis of penile cancer. Cross-sectional anatomic imag-
abdominopelvic CT or MRI and chest imaging [65, 66, 70, ing as provided by CT has proven to be disappointing for
71]. Distant metastasis is detected in 20% of the patients evaluating patients with possible regional lymph node
with cytologically positive inguinal disease. Integration of metastasis, because of the frequent false-positive results
the PET/CT findings into preoperative staging algorithms [73]. In fact, such lymph nodes frequently exceed 2 cm in
may avoid surgical staging in selected patients. size under physiological conditions or can undergo reactive
It has also been shown that [18F]FDG PET can be used changes secondary to the malignant disease of the penis or
for assessing the metabolic response to induction chemo- to chronic inflammation [73]. [18F]FDG PET/CT has dem-
therapy in patients with positive inguinal lymph nodes; in onstrated high sensitivity (88%) and high specificity (98%)
fact, changes in glucose metabolism precede morphological in detecting inguinal lymph node metastasis, being particu-
changes, thus confirming that functional assessment is more larly useful for evaluating deep inguinal and obturator lymph
sensitive than morphological assessment [72]. nodes [65, 66, 70, 71] (Fig. 35.17).
a b
Fig. 35.17 [18F]FDG PET/CT performed for staging purposes in a CT (left panel), coronal PET (center panel), and fused PET/CT image
patient with newly diagnosed penile carcinoma. (a) Fused PET/CT (right panel) demonstrating a pulmonary metastasis in the right anterior
images at various levels showing increased tracer uptake at the primary costophrenic space (reproduced from: Strauss HW, Mariani G,
tumor site (lower right panel) and metastatic disease in left common Volterrani D, Larson SM, eds. Nuclear Oncology – From
iliac lymph nodes. (b) Coronal CT of (left panel), coronal PET (center Pathophysiology to Clinical Applications, 2nd Edition. New York:
panel), and fused PET/CT image (right panel) demonstrating metastatic Springer, 2017)
disease in bilateral mediastinal and para-hilar lymph nodes. (c) Coronal
4. Zukotynski KA, Kim CK, Gerbaudo VH, Hainer J, Taplin ME,

35.3.4.5 Future Perspectives
Kantoff P, et al. 18F-FDG-PET/CT and 18F-NaF-PET/CT in men
Since penile cancer is a rare tumor in most medium- and with castrate-resistant prostate cancer. Am J Nucl Med Mol
high-income countries, there are a limited number of studies Imaging. 2015;5:72–82.
that evaluate the role of SLNB and of [18F]FDG PET/CT in 5. Effert PJ, Bares R, Handt S, Wolff JM, Bull U, Jakse G. Metabolic
imaging of untreated prostate cancer by positron emission tomogra-
the clinical management of this tumor. Future investigations
phy with 18fluorine-labeled deoxyglucose. J Urol. 1996;155:994–8.
will most likely explore the potential role of PET/MRI in the 6. Effert P, Beniers AJ, Tamimi Y, Handt S, Jakse G. Expression
management of this rare neoplasm. of glucose transporter 1 (Glut-1) in cell lines and clinical speci-
mens from human prostate adenocarcinoma. Anticancer Res.
2004;24:3057–63.
Key Learning Points 7. Bombardieri E, Aktolun C, Baum RP, Bishof-Delaloye A,
Buscombe J, Chatal JF, et al. Bone scintigraphy: procedure
• Sentinel lymph node biopsy is indicated in patients guidelines for tumour imaging. Eur J Nucl Med Mol Imaging.
with non-palpable lymph nodes. 2003;30:BP99–106.
• [18F]FDG PET/CT has no role in evaluating primary 8. Gnanasegaran G, Cook G, Adamson K, Fogelman I. Patterns, vari-
neoplasm. ants, artifacts, and pitfalls in conventional radionuclide bone imag-
ing and SPECT/CT. Semin Nucl Med. 2009;39:380–95.
• [18F]FDG PET/CT is recommended for staging 9. Love C, Din AS, Tomas MB, Kalapparambath TP, Palestro
penile cancer patients with palpable lymph nodes, CJ. Radionuclide bone imaging: an illustrative review.
where it can replace conventional imaging proce- Radiographics. 2003;23:341–58.
dures (CT and MRI). 10. Langsteger W, Heinisch M, Fogelman I. The role of fluorodeoxy-
glucose, 18F-dihydroxyphenylalanine, 18F-choline, and 18F-fluoride
• [18F]FDG PET/CT may play a potential role for in bone imaging with emphasis on prostate and breast. Semin Nucl
monitoring the efficacy of treatment. Med. 2006;36:73–92.
11. Fuccio C, Schiavina R, Castellucci P, Rubello D, Martorana G,
Celli M, et al. Androgen deprivation therapy influences the uptake
of 11C-choline in patients with recurrent prostate cancer: the pre-
References liminary results of a sequential PET/CT study. Eur J Nucl Med Mol
Imaging. 2011;38:1985–9.
1. Siegel R, Ward E, Brawley O, Jemal A. Cancer statistics, 2011: the 12. Cimitan M, Bortolus R, Morassut S, Canzonieri V, Garbeglio

impact of eliminating socioeconomic and racial disparities on pre- A, Baresic T, et al. [18F]fluorocholine PET/CT imaging for the
mature cancer deaths. CA Cancer J Clin. 2011;61:212–36. detection of recurrent prostate cancer at PSA relapse: experi-
2. Even-Sapir E. Imaging of malignant bone involvement by mor- ence in 100 consecutive patients. Eur J Nucl Med Mol Imaging.
phologic, scintigraphic, and hybrid modalities. J Nucl Med. 2006;33:1387–98.
2005;46:1356–67. 13. Rabbani F, Stroumbakis N, Kava BR, Cookson MS, Fair

3. Hricak H, Choyke PL, Eberhardt SC, Leibel SA, Scardino WR. Incidence and clinical significance of false-negative sextant
PT. Imaging prostate cancer: a multidisciplinary perspective. prostate biopsies. J Urol. 1998;159:1247–50.
Radiology. 2007;243:28–53.
14. Igerc I, Kohlfurst S, Gallowitsch HJ, Matschnig S, Kresnik E, 30. Heck MM, Retz M, Tauber R, Knorr K, Kratochwil C, Eiber M.
Gomez-Segovia I, et al. The value of 18F-choline PET/CT in [PSMA-targeted radioligand therapy in prostate cancer]. Der
patients with elevated PSA-level and negative prostate needle Urologe Ausg A. 2017;56:32–39
biopsy for localisation of prostate cancer. Eur J Nucl Med Mol 31. Oka S, Hattori R, Kurosaki F, Toyama M, Williams LA, Yu W,
Imaging. 2008;35:976–83. et al. A preliminary study of anti-1-amino-3 18F-fluorocyclobutyl-
15. Kwee SA, Coel MN, Ly BH, Lim J. 18F-Choline PET/CT imaging 1-carboxylic acid for the detection of prostate cancer. J Nucl Med.
of RECIST measurable lesions in hormone refractory prostate can- 2007;48:46–55.
cer. Ann Nucl Med. 2009;23:541–8. 32. Okudaira H, Shikano N, Nishii R, Miyagi T, Yoshimoto M,

16. Silver DA, Pellicer I, Fair WR, Heston WD, Cordon-Cardo
Kobayashi M, et al. Putative transport mechanism and intracellular
C. Prostate-specific membrane antigen expression in normal and fate of trans-1-amino-3-18F-fluorocyclobutane carboxylic acid in
malignant human tissues. Clin Cancer Res. 1997;3:81–5. human prostate cancer. J Nucl Med. 2011;52:822–9.
17. Wright GL Jr, Grob BM, Haley C, Grossman K, Newhall K,
33. Sorensen J, Owenius R, Lax M, Johansson S. Regional distribution
Petrylak D, et al. Upregulation of prostate-specific membrane and kinetics of 18F-fluciclovine (anti-18F-FABC) a tracer of amino
antigen after androgen-deprivation therapy. Urology. 1996;48: acid transport, in subject with primary prostate cancer. Eur J Nucl
326–34. Med Mol Imaging. 2013;40:394–402.
18. Afshar-Oromieh A, Malcher A, Eder M, Eisenhut M, Linhart HG, 34. Nye JA, Schuster DM, Yu W, Camp VM, Goodman MM, Votaw
Hadaschik BA, et al. PET imaging with a [68Ga]gallium-labelled JR. Biodistribution and radiation dosimetry of the synthetic nonme-
PSMA ligand for the diagnosis of prostate cancer: biodistribution tabolized amino acid analogue anti-18F-FACBC in humans. J Nucl
in humans and first evaluation of tumour lesions. Eur J Nucl Med Med. 2007;48:1017–20.
Mol Imaging. 2013;40:486–95. 35. Asano Y, Inoue Y, Ikeda Y, Kikuchi K, Hara T, Taguchi C, et al.
19. Verburg FA, Krohn T, Heinzel A, Mottaghy FM, Behrendt FF. First Phase I clinical study of NMK36: a new PET tracer with the syn-
evidence of PSMA expression in differentiated thyroid cancer using thetic amino acid analogue anti-[18F]FACBC. Ann Nucl Med.
[68Ga]PSMA-HBED-CC PET/CT. Eur J Nucl Med Mol Imaging. 2011;25:414–8.
2015;42:1622–3. 36. McParland BJ, Wall A, Johansson S, Sørensen J. The clinical safety,
20. Demirci E, Ocak M, Kabasakal L, Decristoforo C, Talat Z, Halac biodistribution and internal radiation dosimetry of [18F]fluciclo-
M, et al. 68Ga-PSMA PET/CT imaging of metastatic clear cell renal vine in healthy adult volunteers. Eur J Nucl Med Mol Imaging.
cell carcinoma. Eur J Nucl Med Mol Imaging. 2014;41:1461–2. 2013;40:1256–64.
21. Krohn T, Verburg FA, Pufe T, Neuhuber W, Vogg A, Heinzel A, 37. Turkbey B, Mena E, Shih J, Pinto PA, Merino MJ, Lindenberg ML,
et al. [68Ga]PSMA-HBED uptake mimicking lymph node metasta- et al. Localized prostate cancer detection with 18F-FACBC PET/
sis in coeliac ganglia: an important pitfall in clinical practice. Eur J CT: comparison with MR imaging and histopathologic analysis.
Nucl Med Mol Imaging. 2015;42:210–4. Radiology. 2014;(3):849–56.
22. Artigas C, Alexiou J, Garcia C, Wimana Z, Otte FX, Gil T, et al. 38. Kairemo K, Rasulova N, Partanen K, et al. Preliminary clinical expe-
Paget bone disease demonstrated on 68Ga-PSMA ligand PET/ rience of trans-1-amino-3-18Ffluocyclobutanecarboxylic acid (anti-
CT. Eur J Nucl Med Mol Imaging. 2016;43:195–6. 18
F-FACBC)-PET/CT image in prostate cancer patients. Biomed
23. Verburg FA, Pfister D, Heidenreich A, Vogg A, Drude NI, Voo S, Res Int. 2014;2014:305182. https://doi.org/10.1155/2014/30
et al. Extent of disease in recurrent prostate cancer determined by 5182.
[68Ga]PSMA-HBED-CC PET/CT in relation to PSA levels, PSA 39. Odewole OA, Tade FI, Nieh PT, et al. Recurrent prostate cancer
doubling time and Gleason score. Eur J Nucl Med Mol Imaging. detection with anti-3 18F-FACB PET/CT: comparison with CT. Eur
2016;43:397–403. J Nucl Med Mol Imaging. 2016;43:1773–83.
24. Herlemann A, Wenter V, Kretschmer A, Thierfelder KM,
40. Nanni C, Schiavina R, Brunocilla E, Boschi S, Borghesi M, Zanoni
Bartenstein P, Faber C, et al. 68Ga-PSMA positron emission L, et al. 18F-fluciclovine PET/CT for the detection of prostate can-
tomography/computed tomography provides accurate staging of cer relapse: a comparison to 11C-choline PET/CT. Clin Nucl Med.
lymph node regions prior to lymph node dissection in patients 2015;40:386–91.
with prostate cancer. Eur Urol. 2016;70:553–7. 41. Nanni C, Zanoni L, Pultrone C, Schiavina R, Brunocilla E, Lodi
25. Pyka T, Okamoto S, Dahlbender M, Tauber R, Retz M, Heck M, F, et al. 18F-FACBC (anti-1-amino-3-18F-fluorocyclobutane-
et al. Comparison of bone scintigraphy and 68Ga-PSMA PET for 1-carboxylic acid) versus 11C-choline PET/CT in prostate cancer
skeletal staging in prostate cancer. Eur J Nucl Med Mol Imaging. relapse: results of a prospective trial. Eur J Nucl Med Mol Imaging.
2016;43:2114–21. 2016;43:1601–10.
26. Rahbar K, Schmidt M, Heinzel A, Eppard E, Bode A,
42. Zhao JY, Ma XL, Li YY, Zhang BL, Li MM, Ma XL, et al. Diagnostic
Yordanova A, et al. Response and tolerability of a single dose of accuracy of 18F-FDG-PET in patients with testicular cancer: a meta-
177
Lu-PSMA-617 in patients with metastatic castration-resistant analysis. Asian Pac J Cancer Prev. 2014;15:3525–31.
prostate cancer: a multicenter retrospective analysis. J Nucl Med. 43. Cook GJ, Sohaib A, Huddart RA, Dearnaley DP, Horwich A, Chua
2016;57:1334–8. S. The role of 18F-FDG PET/CT in the management of testicular
27. Baum RP, Kulkarni HR, Schuchardt C, Singh A, Wirtz M, Wiessalla cancers. Nucl Med Commun. 2015;36:702–8.
S, et al. 177Lu-Labeled prostate-specific membrane antigen radio- 44. Tsatalpas P, Beuthien-Baumann B, Kropp J, Manseck A, Tiepolt C,
ligand therapy of metastatic castration-resistant prostate cancer: Hakenberg OW, et al. Diagnostic value of 18F-FDG positron emis-
safety and efficacy. J Nucl Med. 2016;57:1006–13. sion tomography for detection and treatment control of malignant
28. Ahmadzadehfar H, Eppard E, Kurpig S, Fimmers R, Yordanova germ cell tumors. Urol Int. 2002;68:157–63.
A, Schlenkhoff CD, et al. Therapeutic response and side effects 45. de Wit M, Brenner W, Hartmann M, Kotzerke J, Hellwig D,

of repeated radioligand therapy with 177Lu-PSMA-DKFZ-617 Lehmann J, et al. [18F]-FDG-PET in clinical stage I/II non-
of castrate-resistant metastatic prostate cancer. Oncotarget. seminomatous germ cell tumours: results of the German multicen-
2016;7:12477–88. tre trial. Ann Oncol. 2008;19:1619–23.
29. Eiber M, Weirich G, Holzapfel K, Souvatzoglou M, Haller B, 46. Hain SF, O’Doherty MJ, Timothy AR, Leslie MD, Harper PG,
Rauscher I, et al. Simultaneous 68Ga-PSMA HBED-CC PET/MRI Huddart RA. Fluorodeoxyglucose positron emission tomography
improves the localization of primary prostate cancer. Eur Urol. in the evaluation of germ cell tumours at relapse. Br J Cancer.
2016;70:829–36. 2000;83:863–9.
47.
Rajpert-De Meyts E, McGlynn KA, Okamoto K, Jewett etoposide, and bleomycin chemotherapy for metastatic seminoma:
MA, Bokemeyer C. Testicular germ cell tumours. Lancet. clinical value and future directions. J Urol. 2017;198:106.
2016;387:1762–74. 61. Hakenberg OW, Comperat EM, Minhas S, Necchi A, Protzel C,
48. De Santis M, Becherer A, Bokemeyer C, Stoiber F, Oechsle K, Watkin N, et al. EAU guidelines on penile cancer: 2014 update. Eur
Sellner F, et al. 2-18fluoro-deoxy-D-glucose positron emission Urol. 2015;67:142–50.
tomography is a reliable predictor for viable tumor in postchemo- 62. Kochhar R, Taylor B, Sangar V. Imaging in primary penile cancer:
therapy seminoma: an update of the prospective multicentric current status and future directions. Eur Radiol. 2010;20:36–47.
SEMPET trial. J Clin Oncol. 2004;22:1034–9. 63. Mueller-Lisse UG, Scher B, Scherr MK, Seitz M. Functional imag-
49. Becherer A, De Santis M, Karanikas G, Szabo M, Bokemeyer C, ing in penile cancer: PET/computed tomography, MRI, and sentinel
Dohmen BM, et al. FDG PET is superior to CT in the prediction lymph node biopsy. Curr Opin Urol. 2008;18:105–10.
of viable tumour in post-chemotherapy seminoma residuals. Eur J 64. Spiess PE, Izawa JI, Bassett R, Kedar D, Busby JE, Wong F, et al.
Radiol. 2005;54:284–8. Preoperative lymphoscintigraphy and dynamic sentinel node
50. Secil M, Altay C, Basara I. State of the art in germ cell tumor imag- biopsy for staging penile cancer: results with pathological correla-
ing. Urol Oncol. 2016;34:156–64. tion. J Urol. 2007;177:2157–61.
51. Bachner M, Loriot Y, Gross-Goupil M, Zucali PA, Horwich A, 65. Scher B, Seitz M, Reiser M, Hungerhuber E, Hahn K, Tiling R,
Germa-Lluch JR, et al. 2-18fluoro-deoxy-D-glucose positron emis- et al. 18F-FDG PET/CT for staging of penile cancer. J Nucl Med.
sion tomography (FDG-PET) for postchemotherapy seminoma 2005;46:1460–5.
residual lesions: a retrospective validation of the SEMPET trial. 66. Graafland NM, Leijte JA, Valdes Olmos RA, Hoefnagel CA,

Ann Oncol. 2012;23:59–64. Teertstra HJ, Horenblas S. Scanning with 18F-FDG-PET/CT for
52. Albers P, Albrecht W, Algaba F, Bokemeyer C, Cohn-Cedermark detection of pelvic nodal involvement in inguinal node-positive
G, Fizazi K, et al. Guidelines on testicular cancer: 2015 update. Eur penile carcinoma. Eur Urol. 2009;56:339–45.
Urol. 2015;68:1054–68. 67. Mobilio G, Ficarra V. Genital treatment of penile carcinoma. Curr
53. Hinz S, Schrader M, Kempkensteffen C, Bares R, Brenner W, Opin Urol. 2001;11:299–304.
Krege S, et al. The role of positron emission tomography in the 68. Dimopoulos P, Christopoulos P, Shilito S, Gall Z, Murby B, Ashworth
evaluation of residual masses after chemotherapy for advanced D, et al. Dynamic sentinel lymph node biopsy for penile cancer: a com-
stage seminoma. J Urol. 2008;179:936–940; discussion 940. parison between 1- and 2-day protocols. BJU Int. 2016;117:890–6.
54. Gerl A, Clemm C, Schmeller N, Dienemann H, Weiss M, Kriegmair 69. Lutzen U, Zuhayra M, Marx M, Zhao Y, Colberg C, Knupfer S,
M, et al. Sequential resection of residual abdominal and thoracic et al. Value and efficiency of sentinel lymph node diagnostics in
masses after chemotherapy for metastatic non-seminomatous germ patients with penile carcinoma with palpable inguinal lymph nodes
cell tumours. Br J Cancer. 1994;70:960–5. as a new multimodal, minimally invasive approach. Eur J Nucl Med
55. Hartmann JT, Schmoll HJ, Kuczyk MA, Candelaria M, Bokemeyer Mol Imaging. 2016;43:2313–23.
C. Postchemotherapy resections of residual masses from meta- 70. Schlenker B, Scher B, Tiling R, Siegert S, Hungerhuber E, Gratzke
static non-seminomatous testicular germ cell tumors. Ann Oncol. C, et al. Detection of inguinal lymph node involvement in penile
1997;8:531–8. squamous cell carcinoma by 18F-fluorodeoxyglucose PET/CT: a
56. Oechsle K, Hartmann M, Brenner W, Venz S, Weissbach L, Franzius prospective single-center study. Urol Oncol. 2012;30:55–9.
C, et al. [18F]Fluorodeoxyglucose positron emission tomography 71. Graafland NM, Lam W, Leijte JA, Yap T, Gallee MP, Corbishley
in nonseminomatous germ cell tumors after chemotherapy: the C, et al. Prognostic factors for occult inguinal lymph node involve-
German multicenter positron emission tomography study group. J ment in penile carcinoma and assessment of the high-risk EAU
Clin Oncol. 2008;26:5930–5. subgroup: a two-institution analysis of 342 clinically node-negative
57. Quak E, Kovacs I, Oyen WJ, van der Graaf WT. FDG-PET/CT in patients. Eur Urol. 2010;58:742–7.
a patient with poor-risk non-seminoma testis with mature teratoma 72. Graafland NM, Valdes Olmos RA, Teertstra HJ, Kerst JM,

and secondary gliosarcoma: multimodality imaging for guiding Bergman AM, Horenblas S. 18F-FDG PET/CT for monitoring
multimodality treatment. Nucl Med Mol Imaging. 2015;49:237–40. induction chemotherapy in patients with primary inoperable penile
58. Coleman RE, Hillner BE, Shields AF, Duan F, Merlino DA, Hanna carcinoma: first clinical results. Eur J Nucl Med Mol Imaging.
LG, et al. PET and PET/CT reports: observations from the National 2010;37:1474–80.
Oncologic PET Registry. J Nucl Med. 2010;51:158–63. 73. Lont AP, Kroon BK, Gallee MP, van Tinteren H, Moonen LM,
59. Disibio G, French SW. Metastatic patterns of cancers: results from Horenblas S. Pelvic lymph node dissection for penile carcinoma:
a large autopsy study. Arch Pathol Lab Med. 2008;132:931–9. extent of inguinal lymph node involvement as an indicator for pel-
60. Richie JP. Re: Interim 18F-Fluorodeoxyglucose positron emission vic lymph node involvement and survival. J Urol. 2007;177:947–
tomography for early metabolic assessment of response to cisplatin, 952; discussion 952.
Nuclear Medicine in Pediatrics
36
Pietro Zucchetta and Diego De Palma
Contents
36.2 Patient Management 926
36.3 Bone Scintigraphy 927
36.3.1 Benign Musculoskeletal Disease 927
36.3.2 Musculoskeletal Malignancies 932
36.4 Thyroid Scintigraphy 932
36.4.1 Thyroid Disease 933
36.5 Nephro-urology 934
36.5.1 Dynamic Renal Scintigraphy 934
36.5.2 Static Renal Scintigraphy 935
36.5.3 Radionuclide Cystography 937
36.6 Radionuclide Imaging for Neuroblastoma 938
36.6.1 MIBG Scintigraphy 938
36.7 PET/CT in Oncology 940
36.8 Cardiopulmonary Imaging 942
36.8.1 Lung Perfusion Scintigraphy 942
36.8.2 Myocardial Perfusion Scintigraphy................................................................................................. 944
36.9 Gastroenterology 944
36.9.1 Gastroesophageal Reflux 944
36.9.2 Gastric Emptying Scintigraphy 945
36.9.3 Pulmonary Aspiration 946
36.9.4 Ectopic Gastric Mucosa (Meckel Diverticulum) 946
References 948
Learning Objectives
• Illustrate the wide spectrum of diseases affecting sub-
jects in the pediatric age range and the different pre-
sentations and peculiarities of such condition with
respect to similar conditions occurring in adults.
• Illustrate the need for keeping the radiation burden
P. Zucchetta (*)
linked to diagnostic radionuclide-based procedures
Nuclear Medicine Department, University Hospital of Padua, performed in children as low as possible.
Padua, Italy • Understand how some peculiarities of the pediatric age
e-mail: pietro.zucchetta@unipd.it influence the modalities of performing diagnostic
D. De Palma radionuclide-based procedures, with special attention
Nuclear Medicine Department, “H. Circolo” Varese, Varese, Italy to management of patients during the examination.

926 P. Zucchetta and D. De Palma
• Briefly summarize the main mechanisms of radiophar- imaging (i.e., most frequently without intravenous contrast
maceutical uptake carrying diagnostic information. media) with the main purpose to provide data for attenuation
• Describe the relevant modalities for performing, pro- correction and anatomic reference, data should be acquired
cessing, and interpreting the scintigraphic findings in with a pitch ≥1, a slice increment equal to slice thickness (no
children/adolescents. overlap), activation of the dose modulation systems (DMS),
• Describe the use and clinical value of nuclear medicine and setting, whenever possible, a maximum value for both kV
diagnostic examinations in the spectrum of diseases and mA independent from and lower than those suggested by
affecting the pediatric age range. the DMS (Fig. 36.1), according to the body segment scanned
• List the indications for performing the following nuclear and the RF employed. Moreover, in SPECT/CT particular
medicine diagnostic procedures in children/adolescents: care should be paid in limiting the scan length; in this regard,
bone scintigraphy, thyroid scintigraphy, dynamic and image findings observed on the planar whole-body scan are
static renal scintigraphy, radionuclide cystography, very important (Fig. 36.2).
MIBG scintigraphy, [18F]FDG PET/CT, lung perfusion
scintigraphy, myocardial perfusion scintigraphy, gastro-
esophageal reflux scintigraphy, gastric emptying scin- 36.2 Patient Management
tigraphy, and Meckel diverticulum scintigraphy.
• For each of the above investigations, illustrate the advan- Children and adolescents above about 8 years of age are nor-
tages of hybrid imaging versus other imaging modalities. mally able to fully understand the importance of the proce-
dure, and, once they are adequately informed and the
scanning procedure has been clearly explained, they usually
36.1 Introduction remain still for the time required. On the other hand, if a
teenager refuses to undergo to a procedure, this fact becomes
Many distinctive aspects make pediatric nuclear medicine a a matter of local legal requirements about informed consent;
well-defined sub-specialty. Disease spectrum is largely dif- in the great majority of countries, the scan cannot be made
ferent from adults and children need careful personalization without a legal enforcement.
of the technique, taking into account physical and psycho- For those children below this age range, complete infor-
logical development. An accurate evaluation of mation must be given to parents or legal caregivers.
caregiver-child interactions is crucial to reduce exam-related Management of parental anxiety is also very important to
stress on patient and family [1, 2]. reduce stress in the child.
Pediatric nuclear medicine is a well-defined field of inter- Applying an anesthetic cream on the venipuncture site
est, because of the age range and radioprotection issues. Due reduces anxiety due to fear of the pain. For babies, if fasting is
to the small size of many patients, instrumentation that pro- not necessary, it is wise to feed them and perform the scan once
vides better resolution and higher contrast in areas of disease asleep, using Velcro straps, sandbags, and/or pillows to mini-
is very important. The development of hybrid PET/CT and mize motion. Positioning must be comfortable and handling
SPECT/CT systems has allowed radionuclide imaging to careful, especially in children with painful conditions.
make a major step forward. Sometimes it is advisable to image the patient in a nonstandard
During the last 15 years, PET scanners evolved from the position to address the clinical question. Older noncooperative
three-ring BGO crystal and 2-D acquisition system to the children can be entertained with videos or toys. When sedation
modern four-ring LSO crystal and 3-D acquisition. This fact is required [7], the anesthesiology team should be available to
means that we can perform high-quality scans injecting administer the medications and monitor the patient.
lower activities of [18F]FDG [3], as low as 2.5 MBq/kg body
weight. This is of paramount importance in reducing up to
50% the radiation burden due to PET radiopharmaceuticals Key Learning Points
(RFs) in children and adolescents. Increasing the urinary • Pediatric nuclear medicine requires a dedicated
excretion with small amounts (0.2 mg/kg) of furosemide can approach.
also help to reduce the radiation burden, while at the same • There is a wide spectrum of different disease to be
time avoiding a high RF concentration into the urinary tract. explored (or treated) with radionuclide-based
The CT component requires careful setup, since [4] the procedures.
absorbed dose is inversely proportional to body diameters • Practical and emotional issues must be dealt with.
with the same tube setting [5], a concept of paramount impor- • An integrated approach to the management of child
tance in pediatrics. Additional acquisition parameters such as AND family is strongly recommended.
tube mA, kV, pitch, and slice thickness are also important for • There is a stringent need for optimizing radiation
keeping to a minimum the radiation burden from the CT scan. exposure during hybrid imaging (SPECT/CT and
The dose-length product can be converted into effective dose PET/CT).
values [6]. When performing a CT scan for pediatric hybrid
36 Nuclear Medicine in Pediatrics 927
a b c
d e
Fig. 36.1 MIP images from [18F]FDG PET/CT scan of a 14-year-old modified CT parameters for reducing the radiation burden. The interim
girl with Hodgkin’s lymphoma. (a) Baseline study, (b) interim study. study shows quality comparable to the baseline, and there was no ham-
Both scans performed with a four-ring detector scanner allowing reduc- pering in detecting the residual uptake in the subcarinal node (red arrow
tion in injected activity to 2.5 MBq/kg (150 MBq) and dose modulation on MIP, fused images) (c). Dosimetry reports (d, baseline; e, interim)
system for the CT component. The interim scan was acquired with show a 60% reduction in DLP
36.3 Bone Scintigraphy times have an abrupt onset, like acute osteomyelitis (OM), or
arise as a localized pain slowly leading to a functional impair-
Bone scintigraphy is performed most often with 99mTc-labeled ment (e.g., avascular necrosis, osteochondritis, osteoid oste-
bone-seeking tracers (99mTc-phosphonates), which produce a oma). Benign bone lesions (non-ossifying fibromas, bone
functional image of bone remodeling. The standard acquisi- cysts, chondroma, etc.) are almost always serendipitously
tion technique requires waiting for 2–3 h after intravenous found at an X-ray exam performed for different reasons, such
injection (no specific preparation required, besides good as pain. It can be localized or ill defined; its intensity can vary
hydration), before acquiring a whole-body scan in the ante- and may be associated with regularly planned physical activi-
rior and posterior projection; this is then complemented by ties (i.e., stress fractures) or the time of day or night. It may or
spot images and/or SPECT or SPECT/CT, when necessary. may not respond to nonsteroidal anti-inflammatory drugs. It
The delay in acquisition after RF administration allows the is often difficult to localize, especially in nonverbal or nonco-
urinary excretion of the unbound tracer, thus decreasing operative children; the correct anatomical localization of the
background activity and increasing image contrast. source of pain may be hampered by the presence of referred
Three-phase scintigraphy includes a dynamic acquisition pain. In these clinical settings, the value of bone scintigraphy
immediately after injection (60 s, perfusion phase) centered on (BS) remains valuable, due to its high negative predictive
the site of patient symptoms, followed by early equilibrium value and fast whole-body evaluation. Although MRI is obvi-
images (blood pool phase). The third phase is the standard ously the cornerstone of morphologic evaluation of musculo-
whole-body acquisition plus spot and/or tomographic images. skeletal diseases in children, BS only seldom requires
sedation and can be performed safely in children with metal-
lic implants or impaired renal function.
36.3.1 Benign Musculoskeletal Disease Hybrid SPECT/CT systems with state-of-the-art gamma
camera technology, together with the use of the EANM/
Children can be affected by various types of benign diseases SNMMI reference tables/web calculators for pediatric RF
involving the musculoskeletal system. These illnesses some- activities, keep the radiation dose to the patient below 3 mSv.
a b
c e
Fig. 36.2 Three-phase bone scan of a 9-year-old girl with fever, limp- showed bone involvement (e, fused images, top row), with a clear
ing, and pain at the left hip. Elevated WBC count (9.900 mm−3) and ischiopubic metaphyseal plate abnormality (red arrows) (f, CT bone
ESR (65). Hip X-ray and US were unremarkable. Blood flow (a) and window). With CT soft tissue window (g) an abnormality of the adduc-
blood pool phases (b) showed increase at the left groin (arrows). The tory muscle was clearly visible (violet arrow), indicating the inflamma-
delayed whole-body scan (c) showed only intense uptake at the left tory origin of the findings. MRI (H, T2 weighted) confirmed this
ischiopubic branch, with extension to ileopubic branch and to the ace- diagnosis. Dosimetric CT report shows the very low radiation burden
tabulum (arrows). SPECT/CT with a very limited scan length (d) linked to the additional CT scan
The fusion of simultaneously acquired SPECT and CT allows ings obtained during a three-phase BS are abnormal very
a precise anatomical localization of the scintigraphic findings early in the course of disease, as images depict increased
[8–10], thus making it possible to identify not only if a lesion blood flow, increased capillary permeability (blood pool
is hypermetabolic but also which part of the lesion is more or image), and increased bone uptake in the late scan (3P-BS).
less hypermetabolic [11]. The availability of 18F-fluoride as a The bone scan data can help to identify sites for detailed
PET bone-seeking agent offers the superior spatial resolution morphological studies [15] (Fig. 36.3).
of PET/CT than either planar and/or SPECT imaging; shorter Although MRI is necessary for complete characterization,
duration of the study (final images are obtained in a maxi- SPECT/CT can rule out the majority of malignant bone
mum time of 90 min) is an additional definite advantage over lesions; it also serves to evaluate soft tissue involvement and
gamma camera imaging with 99mTc-phosphonates [12, 13]. helps in guiding the MRI scan [16].
Reflex sympathetic dystrophy (RSD, also called “com-
36.3.1.1 Clinical Indications plex regional pain syndrome”) is characterized by severe
Acute osteomyelitis, an acute illness affecting non-violated pain, often coupled with reddening, swelling, and changes
bone, is due to hematogenous spread of bacteria colonizing in the skin; it normally affects one limb. It can be crypto-
the bone marrow. In children the active metabolism of genic but often arises after an injury [17]. Three-phase
growing plates requires a relatively large blood flow, a factor BS normally allows a quick and reliable diagnosis, while
that facilitates bone localization of bacteria [14]. The find- SPECT/CT can be very useful for defining the character-
b c d
Fig. 36.3 Three-phase bone scan of an 11-year-old girl with fever, bone (red arrows, fused images, top row), with no clear bone abnor-
limping, and pain at the left hip. Elevated WBC count and ESR. Hip Rx malities (CT bone window, middle row). With CT soft tissue window
and US were unremarkable. Blood flow (a) and blood pool phases (b) (bottom row), an abnormality of the pectineus muscle was clearly visi-
showed increase at the left groin (black arrows). The delayed whole- ble. MRI (e, T2 weighted) confirmed the primary involvement of soft
body scan (c) showed only mild increase of uptake at the pubic bone tissues with a mild bone reaction
(black arrow). SPECT/CT (d) confirmed the mild involvement of pubic
istic of foci of more intense uptake, sometime the occult which is nevertheless very low. Chronic recurrent multi-
cause of RSD (Fig. 36.4). It must be emphasized that focal osteomyelitis (CRMO) is an uncommon condition
hybrid imaging of the extremities (hands, feet) implies an in which the bones have multiple inflammatory, painful
additional radiation burden due to the CT component, lesions [18, 19]. It is characterized by multiple areas of
a b e
c d
Fig. 36.4 Young boy with slow onset of swelling and pain at the right cuneiform (c) with no major bone damage at either hybrid CT (d) or
foot; RDS suspected. Blood flow (a) and blood pool (b) phases of a radiological CT (e). RDS was ruled out, leaving aseptic necrosis as the
three-phase bone scan showed only a focal increase in the midfoot most likely diagnosis
(arrows); SPECT/CT showed focally increased uptake at the second
a b
c d
Fig. 36.5 SPECT/CT of an 18-year-old girl with pain at the left leg images). Bone window CT images (c) demonstrate a less dense area
and X-ray femoral hyperostosis. SPECT/CT MIP image (a) shows a in between, the nidus of an osteoid osteoma, later confirmed by a tar-
focal high uptake, localized at the hyperostotic focus (b, fused geted CT scan (d)
bone inflammatory involvement; the fast, low-dose less than 2 cm in size [20]; if larger, it is normally an
whole-body survey allowed by BS represents an impor- osteoblastoma, still a benign condition. It is quite pain-
tant asset in its diagnosis, and a carefully targeted SPECT/ ful, with characteristic increase during nighttime.
CT can be helpful to select sites for bone biopsy. Osteoid SPECT/CT BS can be considered as “the one-stop shop”
osteoma (OO) is an osteoblastic benign tumor generally [21] for the diagnosis of OO; the typical appearance
d
b
Fig. 36.6 A 17-year-old girl, professional runner. She complained pain an uptake at the insertion of adductory muscles. SPECT/CT better dem-
at the left leg severely hampering her daily practice. A previous MRI onstrates the findings (b, MIP image), involving the cortical bone (c,
showed marrow edema due to a likely stress fracture. The whole-body fused images), with no bone abnormalities (d, CT bone window)
bone scan (a) showed the bilateral (double-headed arrow) presence of
a b
Fig. 36.7 A 16-year-old young volley player with persisting pain after dow) shows tiny bone damage, consistent with residual lesion of the
left ankle sprain. SPECT/CT shows focal uptake at the tibial malleolus deltoid ligament insertion
(a, MIP image), well shown by the fused images (b). CT (c, bone win-
shows the translucent nidus with the surrounding bone

sclerosis on CT, with a nidus-corresponding very “hot” Key Learning Points
spot on the BS and surrounding diffuse mildly increased • 99m
Tc-labeled bone-seeking agents (phosphonates)
tracer uptake (Fig. 36.5). are employed for conventional bone scintigraphy
Increased participation in competitive sports has increased with single-photon imaging.
the number of patients evaluated for traumatic injuries. • There are specific indications for standard delayed
Although these lesions are usually benign, they are often acquisition versus the three-phase acquisition
painful. Occasionally, subtle bone lesions like spondylolysis protocol.
[22] or stress fractures [23] (Fig. 36.6) are identified (Fig. • For benign musculoskeletal disorders, bone scintig-
36.7) with BS. raphy allows fast whole-body exploration and has a
Another teenage health problem in which hybrid bone high negative predictive value.
scan helps is condylar hyperplasia, an asymmetric growth of • Hybrid imaging with SPECT/CT better defines
the mandibular condylar metaphysis that causes a distortion site(s) and intra-lesion distribution of tracer uptake.
of the mandibular bone and both gnawing and esthetic prob- • Benign diseases that most benefit from diagnostic
lems [24]. The right timing for surgical correction is at the characterization with bone scintigraphy include
end of the developing age growth, before the arising of osteomyelitis, reflex sympathetic dystrophy/com-
degenerative alteration; it is the BS only that can reveal when plex regional pain syndrome, chronic recurrent
the metabolic activity of the metaphysis is declining com- multifocal osteomyelitis, osteoid osteoma, and
pared to reference bone. sport-related injuries.
36.3.2 Musculoskeletal Malignancies 36.4 Thyroid Scintigraphy
Osteosarcoma and Ewing’s sarcoma have been typical indi- Thyroid scintigraphy is performed by intravenous injection
cations for skeletal scintigraphy, both at initial staging and of 99mTc-pertechnetate (99mTcO4−) or ingestion of 123I-Iodide.
for evaluating the response to treatment. Nevertheless, [18F] It requires withdrawal of drugs interfering with radiotracer
FDG and 18F-Fluoride PET/CT have more recently been pro- uptake in the thyroid, such as antithyroid drugs, iodinated
posed as valuable alternatives to BS. compounds, etc.
Static images are acquired 15–30 min after 99mTcO4− SPECT/CT is helpful for evaluating both the target volume
injection (anterior view, lateral if required, e.g., for a lingual to be treated with 131I-iodide or, in the posttreatment scan,
thyroid), using high-resolution parallel-hole collimators or anatomical localization of the residual tissue (Fig. 36.8).
pinhole collimators. In therapy planning the small additional radiation burden
due to the CT scan is irrelevant, although of course it should
not be totally neglected. As a rule of thumb, the total cumula-
36.4.1 Thyroid Disease tive 131I-iodide administered activity must not exceed
18.5 GBq, because no clinical advantage has been demon-
Although rare, hyperthyroidism (HT) and differentiated thy- strated with greater activities. Such a limit helps avoid the
roid carcinoma (DTC) may affect children and adolescents. detrimental effect of chronic radiation-induced sialoadenitis.
In children and adolescents, HT is nearly always due to The radiation dose to the salivary glands can be reduced by
Graves’ disease, and its management follows basically the administering lemon juice or candies following administra-
same steps and protocols as in adults. Clinical protocols tion of 131I-iodide for ablation (or therapy); the acidic compo-
commonly start with medical therapy with antithyroid drugs, nent of lemon juice stimulates salivation, thus reducing the
lasting for at least 6–12 months. More radical approaches, residence time of radioiodide in the salivary glands. Another
surgery and radioiodine treatment, are considered for those concern is prevention of radiation-induced lung fibrosis in
affected by severe side effects caused by antithyroid drugs, case of diffuse pulmonary micro-metastasis.
for nonresponders, or for patients relapsing after TSH nor- Following treatment, residual moderately high levels of
malization and subsequent withdrawal of antithyroid drugs serum thyroglobulin usually remain stable or have a slow
[25]. It must be emphasized that the percentage of children decline over time, thus signaling a stable, non-life-threatening
falling into these three categories is higher than for adults. residual disease. For both diagnostic and therapeutic pur-
Radioiodine treatment can be considered unreservedly in poses, recombinant human TSH (rhTSH) can be used instead
children, considering the technical difficulties and the occa- of thyroid hormone replacement withdrawal [29]. Although
sional complications of surgery. The administered activity rhTSH is not approved for use in pediatrics, there is robust
must be carefully calculated on the basis of a radioiodine evidence in the literature that its use is safe and certainly less
uptake test including measurements 4–6, 24, and at least 96 h potentially dangerous on children development than hor-
after oral administration of the tracer amount of radioiodide mone withdrawal. It is nevertheless advisable to obtain a
(0.4 MBq); this prolonged uptake test serves to accurately local approval for this off-label use.
define the effective residence time of radioiodine into the Thyroid scintigraphy is a pillar also in many clinical pro-
thyroid parenchyma. Volume of the gland is calculated on the tocols for congenital hypothyroidism, particularly for the
basis of ultrasound imaging, with an approximation that con- diagnosis of thyroid ectopy, where 123I-iodide offers some
siders the thyroid lobes as rotational ellipsoid to derive the advantages over 99mTc-pertechnetate [30]; SPECT/CT pro-
volume from the three axes. Sometimes the thyroid lobes are vides the most useful information for localizing the ectopic
not so quite ellipsoid; in these cases, a 99mTc-pertechnetate thyroid tissue.
thyroid scan complemented with a SPECT/CT acquisition
(with a very short scan field including only the thyroid) can
be used to calculate volume of the gland. Although “normal- Key Learning Points
ization” of thyroid function can be attempted with radioio- • Thyroid scintigraphy can be performed in children
dide therapy, the final outcome of radiometabolic therapy in with either 99mTc-pertechnetate (99mTcO4−) or
children with Graves’ disease unresponsive to medical treat- 123
I-Iodide.
ment is generally ablation of the gland. • The use of 131I-iodide for diagnostic purposes is
DTC in pediatrics is relatively rare, but its incidence is highly discouraged in children, because of the high
expected to increase, being one of the most frequent second- radiation burden of this radionuclide.
ary cancers in survivors of a first malignancy (such as neuro- • Radiometabolic therapy with 131I-iodide is com-
blastoma or Hodgkin’s disease) often implying a radiotherapy monly employed in pediatrics for two different
treatment involving the neck. The smaller size of children’s indications: Graves’ disease unresponsive to medi-
neck and gland justifies the fact that the primary cancer infil- cal therapy and differentiated thyroid carcinoma.
trates the capsule (thus becoming a T3 tumor) more often • For therapy of Graves’ disease, activity of therapeu-
than in adults. tic 131I-iodide is based on dose calculation using
After surgery, radioiodide ablation is normally required data from an extended radioiodine uptake test and
[26–28]. To optimize calculation of the ablation activity, a volume of the gland derived from ultrasound.
pre-therapy whole-body scan with 123I-iodide should be per-
formed. When noteworthy focal uptake is encountered,
tomography (CT), provides clinicians with a full assessment

• In children with Graves’ disease, a first radioiodide of the morpho-functional conditions of the urinary tract.
therapy may aim “normalizing” thyroid function,
resorting to an ablative therapy if the disease relapses.
• There is a maximum limit of 18.5 GBq to be admin- 36.5.1 Dynamic Renal Scintigraphy
istered for postsurgical treatment of children with
differentiated thyroid carcinoma. Dynamic renal scintigraphy is performed using tubular extrac-
• In children with congenital hypothyroidism, the tion tracers, as 99mTc-mercaptoacetyltriglycine (99mTc-MAG3),
most important diagnostic imaging information is 123
I-orthoiodohippurate (123I-OIH) and 99m
Tc-
provided by thyroid scintigraphy with 123I-iodide, ethylenedicysteine ( Tc-EC), or
99m 99m
Tc-diethylenetriamine-
possibly acquiring SPECT/CT images. pentaacetic acid (99mTc-DTPA). The latter RF is excreted by
glomerular filtration only.
99m
Tc-labeled tubular tracers (99mTc-MAG3, 99mTc-EC)
are preferable in children, particularly in newborns and
36.5 Nephro-urology infants, because they have greater renal extraction allowing
better visualization of the excretory apparatus, even in hypo-
Nuclear medicine has been playing a key role in pediatric functioning kidneys and/or in high degree hydronephrosis;
nephro-urology for decades and retains a central position in furthermore, they are more readily available than 123I-OIH.
the workup of many frequent conditions, such as urinary For dose calculation the EANM Dosage Card Calculator
tract infections or congenital hydronephrosis [31]. or the SNM Pediatric Injected Activity Tool should be used
The functional information (particularly split renal func- [32, 33].
tion) provided by radionuclide techniques is crucial for the Good hydration is crucial for good-quality images, and it
clinical management of many patients, and it is not easily is obtained prior to the examination by oral hydration or by
obtained by other means, despite progress in radiology. intravenous fluid administration.
The integration of scintigraphy with other imaging tech- Images are acquired as a dynamic sequence (15–20 s/frame)
niques, mainly with ultrasonography (US), micturating cys- for 30–45 min in the posterior view. At least one static image
togram (MCUG), or, less frequently, MRI or X-ray computed after micturition and upright positioning (gravity- assisted
a b
Fig. 36.8 (a) Post-ablation (18.5 GBq 131I-iodide) whole-body scan of residual lymph node involvement. SPECT/CT (b, coronal slices; c,
an 18-year-old girl who underwent total thyroidectomy and central transaxial slices) shows that the activity is localized into the thyroid bed
compartment lymphadenectomy because of a papillary thyroid carci- and along the thyroglossal duct path and is then due to thyroid
noma, pT1 pN1a. Activity in the left side of the neck was suspect for remnant
drainage) is mandatory to evaluate kidney washout with- Less frequent indications are renovascular hypertension,
out the reflex interference of a full bladder on ureteral peri- kidney stones, flank pain, or transplanted kidney follow-up.
stalsis [34].
Intravenous furosemide administration (1 mg/kg, up to a
maximum of 40 mg) can be used to obtain maximal urine Key Learning Points
flow rates, making it easier to identify obstruction. • Integrated protocols combine radionuclide methods
Traditionally furosemide has been injected around 20 min with other imaging modalities, such as either US,
after tracer injection (F + 20 protocol), but it may be more MCUG, MR, and/or CT.
practical to inject it immediately after the radiotracer (F + 0 • Dynamic renal scintigraphy can be performed with
or F + 2 protocol), to shorten the acquisition time and the different radiopharmaceuticals characterized by
need for immobility. different parameters of excretion.
Tracer washout can be evaluated qualitatively or by calcu- • Tubular agents (99mTc-MAG3, 99m
Tc-EC,
lating a numerical index, taking into account kidney func- 123
I-hippuran) are preferred for use in children.
tion, as normalized residual activity (NORA) or output • Careful hydration and post-void imaging (gravity-
efficiency (OE). Time measurements of tracer washout assisted drainage) are crucial for highest diagnostic
(post-furosemide t50, etc.) are not reproducible and cannot accuracy.
be relied upon to identify obstruction [35]. • Different protocols have been proposed for the
Radiotracer uptake and excretion are visually assessed by furosemide test (1 mg/kg i.v.), with variable timing
inspecting the whole sequence of images and by evaluating between injection of the renal excretion radiophar-
background-subtracted activity/time curves, derived from maceutical and furosemide injection.
two background corrected regions of interest—one covering • Evaluation with numerical washout parameters
the cortex only and the other encompassing the whole kidney includes NORA or OE.
(including the cortex and medulla). • The most common clinical conditions for use of
The most common indication for dynamic kidney scin- dynamic renal scintigraphy in children include con-
tigraphy in children is congenital hydronephrosis/hydro- genital hydronephrosis and complex malformations.
ureteronephrosis, usually after antenatal detection by fetal
US screening. It has been demonstrated that congenital
kidney dilatation does not imply obstruction, since con-
genital hydronephrosis is a self-limiting condition that in 36.5.2 Static Renal Scintigraphy
the majority of cases does not lead to loss of kidney func-
tion or complications if left untreated. There are, neverthe- 99m
Tc-dimercaptosuccinic acid (99mTc-DMSA) is currently
less, a significant number of patients developing the tracer of choice for static renal scintigraphy. About half
progressive kidney damage and irreversible loss of kidney of the injected dose (see EANM Dosage Card Calculator or
function. In this group, prompt surgical correction (pyelo- the SNM Pediatric Injected Activity Tool [32, 33]) is trapped
ureteroplasty or ureteral reimplant, depending on the site in the kidney cortex 3–4 h after injection, which is the rec-
of obstruction) is required to prevent permanent loss of ommended waiting time to obtain high-quality, low-
renal function (Fig. 36.9). background images. Normal hydration before injection and
No single test can identify those “surgical” cases, but the during the waiting time is the only preparation required, and
combination of ultrasound and dynamic diuretic scintigraphy sedation is seldom needed.
offers the best balance between clinical outcome, invasive- Static images are acquired in posterior, left posterior
ness, and costs [35]. Although several approaches have been oblique, and right posterior oblique views; additional views
proposed, sequentially combined kidney ultrasonograms and are acquired if needed. In smaller children, particularly in
radionuclide scans probably offer the best opportunity of infants, pinhole collimators (2–3 mm insert) can be used for
evaluating all critical aspects, including dilatation severity, image magnification. SPECT imaging may be useful in case
cortical thickness, split renal function, and urine washout. of ectopic or transplanted kidney, but it should be kept in
Diuretic renal scintigraphy plays a pivotal role in this mind that SPECT data are more susceptible to artifacts than
multimodality approach, allowing a fast, repeatable, nonin- planar acquisitions, and often the longer acquisition time
vasive assessment of split renal function and urine drainage, involved by the procedure requires sedation.
with very low radiation exposure and no need for routine 99m
Tc-DMSA uptake depicts the distribution of function-
sedation. ing renal parenchyma and images are visually evaluated for
Dynamic renal scintigraphy offers information in com- cortical defects: typical postinfection scars are seen as “cold”
plex renal malformations, as ectopic or duplex kidney, where defects of the kidney profile, often triangular in shape with
it is possible to evaluate each renal moiety. the vertex toward the calico-pyelic cavities. Furthermore,
936
a 0.6
Furosemide
0.5
0.4
1 min 2' 4' 6' 8'
0.3 LK
Kcounts
0.2
0.1
RK
0.0
10' 12' 14' 16'
–0.1
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600
image
b
L 18' 20' 22' H
2 3 4
26' 28' 30'
6 7 8
L R
Post–Void
34' 36' 38' 10 11 12
Fig. 36.9 A 2-month-old boy with congenital left hydronephrosis. Dynamic renal scintigraphy (99mTc-MAG3) showing (a) symmetric uptake, progressive radiotracer retention in the left renal
pelvis, normal drainage on the right. (b) Negligible drainage from the left pelvis after voiding and upright position
P. Zucchetta and D. De Palma
tory toxic factors. They progress to typical scar lesion when

reparative processes, often lasting over 6 months, fail.
Other useful indications for static renal scintigraphy are
kidney ectopy, when data on urine excretion are not needed,
or to confirm the diagnosis of nonfunctioning multicystic
kidney.
Key Learning Points

• Static renal scintigraphy with 99mTc-DMSA depicts
the functioning renal parenchyma.
• Cortical defects and split renal uptake (L/R %) are
easily evaluated with static renal scintigraphy.
• The scan is highly sensitive and accurate for detect-
POST ing post-UTI scarring and for confirming acute
pyelonephritis.
Fig. 36.10 A 3-year-old girl with history of febrile urinary tract infec-
tion before (last episode 7 months before). Static renal scintigraphy
(99mTc-DMSA) shows a scar in the lower pole of the right kidney
36.5.3 Radionuclide Cystography
Radionuclide cystography is indicated to detect and quantify

split renal function is easily calculated, by drawing regions vesicoureteral reflux (VUR).
of interest on both kidneys in the posterior view image. After voiding, the bladder is slowly filled through a #6
Geometric mean calculation and/or background subtraction French no-balloon urethral catheter, using warmed (37 °C)
are necessary only for very small or ectopic kidneys. saline solution mixed with a 99mTc-labeled tracer (usually
Typical indication for 99mTc-DMSA scintigraphy is the 99m
Tc-pertechnetate or technetium sulfur colloid in the amount
search for kidney scarring (Fig. 36.10) after febrile urinary tract of 15–30 MBq). When the bladder reaches its maximum capac-
infections (UTI) [36]. In order to avoid false-positive studies, ity, whenever possible the catheter is gently removed (small
imaging has to be delayed at least 6 months after the last UTI. babies often void suddenly giving no time for this maneuver),
Acute pyelonephritis and renal scarring have been considered and the patient voids in a bedpan or in a urinary bag, taking
for long time a direct consequence of vesicoureteral reflux adequate precautions to avoid spilling and contamination.
(VUR). Better insight into the relationships between VUR, Dynamic images (10 s/frame) are acquired in the poste-
UTI, renal scarring, and kidney function loss has currently rior view throughout the filling phase, including the whole
shifted the focus on renal parenchyma, thus confirming the abdomen in the field of view. During the voiding phase,
important role of 99mTc-DMSA scintigraphy in the workup of faster images (2–5 s/frame) are acquired, again in the poste-
children with recurrent urinary infections. Sensitivity of static rior view. The same imaging protocol may be used, without
renal scintigraphy for renal scars is far superior to ultrasonog- the need for bladder catheter, after completion of a dynamic
raphy, with a limited exposure to ionizing radiations and with renography, to obtain an indirect radionuclide cystography.
no need for sedation, even in young children. This procedure requires more patient cooperation and is bet-
It has been demonstrated that a negative 99mTc-DMSA ter suited to toilet-trained children.
scan is highly predictive of absent or very low-grade reflux, Vesicoureteral reflux is diagnosed when tracer appears in the
thus making MCUG unnecessary in a significant number of ureter(s) and renal pelvis. Approximate grading of reflux sever-
patients. ity is possible, based on the level reached by the tracer (lower
99m
Tc-DMSA scanning can be very useful in confirming ureter, mid-ureter, etc.) and on the dilatation degree (Fig. 36.11).
or excluding the clinical diagnosis of acute pyelonephritis, Radionuclide cystography is the most sensitive diagnostic
which is sometimes controversial, especially when empiric procedure for detecting VUR and has a very favorable dosi-
antibiotic treatment has already been started [37]. The typi- metric profile when compared with conventional radiographic
cal uptake pattern in acute pyelonephritis consists of hypoac- technique, but it cannot provide anatomical information on
tive (“cold”) areas, which are of variable size and usually do the bladder and urethra. Therefore, traditional MCUG is bet-
not reach the kidney profile. These hypoactive areas depend ter suited for the first diagnostic evaluation of suspected
on a combination of vascular stasis, interstitial edema, and VUR. There is some debate on the role of ultrasonographic
tubular cells hypo-function secondary to local and circula- cystography, which completely avoids ionizing radiations,
Fig. 36.11 A 2-year-old girl

with history of febrile urinary
tract infection and right-sided
VUR. Direct radionuclide
cystogram, voiding phase:
VUR reaching the right renal
pelvis
but requires a very skilled operator and is therefore not always

readily available. 36.6 Radionuclide Imaging
Radionuclide cystography is usually recommended in the for Neuroblastoma
follow-up of VUR patients or as first examination in selected
cases, when bladder dysfunction and urethral anomalies can The norepinephrine analog metaiodobenzylguanidine
be excluded on a clinical basis. (MIBG) is a substrate for the catecholamine reuptake system
I, which is located in the cells derived from the neural crest.
Therefore, the radioiodinated agent MIBG has a specific
Key Learning Points affinity for some neuroendocrine tumors, particularly for
• Direct radionuclide cystography consists in the neuroblastoma and for pheochromocytoma [38].
intravesical administration of a 99mTc-labeled agent. Neuroblastoma is the most common extracranial solid
• This radionuclide procedure is indicated in patients tumor of childhood and arises most commonly in the adrenal
with recurrent UTI, to ascertain the presence of gland or in the retroperitoneum. The vast majority of cases
vesicoureteral reflux. are detected during infancy or early childhood, and a greater
• Fast dynamic images are acquired both during fill- proportion of patients diagnosed after 1 year of age have
ing and during the voiding phase. advanced disease and worse prognosis.
• This very sensitive examination involves a very low
radiation burden to patients.
• Direct radionuclide cystography does not provide 36.6.1 MIBG Scintigraphy
anatomical information (on the bladder wall, ure-
thra, etc.) that is important for planning surgery. [123I]MIBG is preferred for diagnostic imaging over [121I]
• It is best indicated for follow-up after the diagnosis MIBG, because it offers better image quality and lower radi-
(with or without surgical treatment) of vesicoure- ation exposure because of the more favorable physical prop-
teral reflux. erties of 123I versus 131I (159 keV versus 364 keV peak of
Fig. 36.12 A 14-month-old

girl recently diagnosed with
neuroblastoma. Whole-body
[123I]MIBG scan showing
diffuse skeletal involvement
ANTERIOR POSTERIOR
γ-ray energy, absent emission of β− particles, and 13-h half- Most studies have evaluated 131I-MIBG treatment in refrac-
life versus 8-d half-life). tory or recurrent neuroblastoma, either as a single agent or in
Patient preparation includes protective thyroid blockade combination with radiation sensitizers and cytotoxic drugs,
(e.g., saturated solution of potassium iodide, or Lugol’s solu- with encouraging results. Recent data seem to support also the
tion) and careful screening for interfering medications, such use of [131I]MIBG therapy in newly diagnosed high-risk
as phenylephrine and its analogs (as in nasal decongestants), patients.
tricyclic antidepressants, or labetalol. [18F]FDG PET/CT has been used as a complementary
Images are typically acquired 24 h after the i.v. injection imaging technique in MIBG-negative neuroblastomas and in
of [123I]MIBG (see EANM Dosage Card Calculator [32] or cases presenting discordant or nonconclusive findings during
the SNM Pediatric Injected Activity Tool [33]), using prefer- follow-up.
ably medium-energy parallel-hole collimators, to reduce Very promising results have also been reported with the
image blurring due to septal penetration by high-energy pho- use of 18F-DOPA PET/CT, showing higher sensitivity and
tons (≈400 keV), which represent around 3% of the 123I γ-ray accuracy for 18F-DOPA compared with [123I]MIBG [38, 40].
emission.
Planar imaging includes anterior and posterior whole-
body images (Fig. 36.12), which are usually complemented Key Learning Points
by lateral views of the skull and high-count spot images of • In children with neuroblastoma, [123I]MIBG is pre-
the region of the primary tumor (most frequently the abdo- ferred over [131I]MIBG because of more favorable
men). In recent years SPECT (and especially SPECT/CT) physical characteristics resulting in better quality
of the primary tumor has gained popularity, because it imaging and lower radiation burden to patients.
increases the rate of lesions’ detection and localization • Whole-body imaging is usually acquired 24 h post-
(Fig. 36.13). injection, SPECT/CT often supplying the most use-
Both planar and tomographic images are visually ful information.
inspected for abnormal foci of MIBG uptake. A standardized • A standardized scoring system (SIOPEN) has been
scoring system developed by the International Society of developed to evaluate the findings of MIBG
Paediatric Oncology Europe Neuroblastoma (SIOPEN) may scintigraphy.
be used to increase reproducibility, prognostic value, and • The SIOPEN score is crucial for defining extension
patient stratification [39]. of skeletal disease.
The combination of [123I]MIBG imaging and of targeted • There is an important opportunity for theragnostics,
therapy with [131I]MIBG offers a possible theragnostic based on the use of [123I]MIBG scintigraphy for
approach to neuroblastoma. assessment pre-therapy with [131I]MIBG.
Fig. 36.13 A 14-month-old girl recently diagnosed with neuroblastoma. [123I]MIBG SPECT/CT of the pelvis. Fused images
36.7 PET/CT in Oncology adults: tracer uptake in the Waldeyer’s ring is frequently
observed, as well as diffuse thymic uptake, and, as already
The introduction of PET/CT into clinical practice has fos- mentioned, variable uptake in brown fat tissue. Maturation
tered a revolution in nuclear medicine, involving also pedi- and development of the central nervous system cause dra-
atric applications, particularly in the oncology field. matic changes in the pattern of [18F]FDG accumulation in the
Similarly as in adults, [18F]FDG is the most widely used brain, especially during the first 2 years of life; therefore,
PET radiotracer in children, and patient preparation has the careful evaluation is required to distinguish normal patterns
same requisites (4–6 h fasting, avoidance of intense physi- from disease.
cal activity during the prior 24 h, measurement of blood Image interpretation is based on qualitative evaluation of
glucose before injection). Furthermore, special care must lesion uptake, in comparison with surrounding normal struc-
be taken to keep the patient warmed before (30–60 min) tures or liver parenchyma, and on calculation of the stan-
and after [18F]FDG administration, to prevent tracer uptake dardized uptake value (SUV).
by activated brown adipose tissue, which could interfere The most frequent indications for [18F]FDG PET/CT in
with image interpretation, particularly in the neck and children are malignant lymphoma (Fig. 36.14) (most com-
upper thorax. monly Hodgkin’s disease in the pediatric age range), Ewing’s
Acquisition usually starts 50–60 min after injection and sarcoma, soft tissue sarcomas, and osteosarcomas. [18F]FDG
includes at least the torso, from skull base to thighs. Whole- scanning has a clinical impact in up to 60% of patients, mod-
body imaging (top of the skull to feet) is often performed in ifying the management both in the diagnostic phase and dur-
children, depending on the clinical suspicion/diagnosis. ing follow-up [41, 42].
Integrated PET/CT scanners offer the possibility of low- Integrated PET/MR has been introduced into the clinical
dose CT, aimed at attenuation correction and anatomical local- practice in the last few years and is being considered with
ization, or full-dose diagnostic CT, depending on the clinical great interest for pediatric applications, because MRI is often
question and on practical and regulatory considerations. required in the workup of many pediatric cancers and the
The normal patterns of [18F]FDG biodistribution in chil- absence of the CT component significantly reduces radiation
dren change during development and differ from those of exposure (Figs. 36.15 and 36.16).
Fig. 36.14 A 15-year-old boy. Hodgkin’s lymphoma. [18F]FDG PET/CT. Coronal fused images
Fig. 36.15 A 13-year-old girl. Recurring non-Hodgkin’s lymphoma. [18F]FDG PET/MR. Coronal fused images
Patients treated for complex malformations often present

Key Learning Points a unique combination of anatomic anomalies and conse-
• [18F]FDG is the radiopharmaceutical most com- quences of multiple procedures. In this regard, MRI is
monly employed for PET/CT scanning. emerging as a powerful imaging technique for diagnosis/
• [18F]FDG uptake due to brown adipose tissue activa- characterizing congenital heart disease [43–47]. Therefore,
tion should be prevented by keeping the patient warm. extra-care is required to obtain a clear picture of the anatomo-
• CT protocols optimized to keep radiation exposure functional status, in order to tailor the scintigraphic study to
as low as possible must be adopted. the specific diagnostic needs [48].
• There are important age-dependent changes in the
physiologic biodistribution patterns of [18F]FDG.
• The most common indications for [18F]FDG PET/ 36.8.1 Lung Perfusion Scintigraphy
CT in pediatrics are Hodgkin’s lymphoma, Ewing’s
sarcoma, soft-tissue sarcomas, and osteosarcomas. Lung perfusion scintigraphy (LPS) is performed using 99mTc-
MAA (macroaggregates of albumin) or 99mTc-albumin
microspheres. It is advisable to reduce not only the injected
activity (see EANM Dosage Card Calculator [32] or the
36.8 Cardiopulmonary Imaging SNM Pediatric Injected Activity Tool [33]) but also the num-
ber of injected particles, in order to avoid a significant
Implementation of radionuclide techniques in pediatric car- increase in pulmonary vascular resistance. Further reduction
diology requires some modifications of the protocols for is appropriate when a right-to-left shunt is present, to
adults, even if the radiotracers employed and the procedures decrease the chance of systemic dissemination.
are the same as for adult patients. In fact, changes are There is no need for a specific preparation, and sedation is
required to take into account different needs at different ages very rarely used, because tomographic acquisition is usually
and to deal with a completely different disease spectrum, not performed.
where atherosclerosis is virtually absent and, conversely, Detailed clinical history is of paramount importance for
congenital malformations are prominent. precise planning of the exam and interpretation of the results.
Fig. 36.16 A 13-year-old girl. Recurring non-Hodgkin’s lymphoma. [18F]FDG PET/MR. Pelvic and leg lesions
All relevant anatomical features of the original malformation anterior views. In infants and uncooperative children, it is
have to be considered and correlated with the effects deriv- possible to avoid the acquisition of the anterior view, because
ing from surgical/percutaneous procedures, because they are there is little difference between the results obtained by the
major determinants in particle distribution. Lung perfusion geometric mean and those gained from the posterior view
scintigraphy is not contraindicated when right-to-left shunt- only. If no right-to-left shunt is present, the injected particles
ing is present, even in the case of complete mixing, as is the will be completely trapped inside the pulmonary capillaries,
case for univentricular heart. A substantial decrease in parti- and background subtraction is not necessary.
cle numbers and the preferential use of 99mTc-MAAs, which It is possible to quantify the magnitude of a right-to-left
have a shorter biological half-life, reduce the risk of systemic shunt by comparing total lung counts with counts derived
embolism. The number of particles can be decreased to from a ROI drawn on the whole brain: brain activity should
10,000, when required, without losing diagnostic power. be below 1% of the pulmonary activity in normal children.
Concomitant disease, particularly bronchopulmonary Pulmonary perfusion percentages obtained by lung perfu-
infections, may influence tracer distribution, making image sion scintigraphy lie in the 45%–55% range for the single
interpretation difficult. It is recommended to postpone scan- lung and are highly reproducible, representing a very useful
ning when wheezing is present, because even low-grade and practical tool for serial imaging during follow-up (Fig.
bronchial constriction may produce a significant redistribu- 36.17). It should be emphasized that lung perfusion scintig-
tion of pulmonary blood flow, most often resulting in a raphy is not sensitive to bilateral blood flow disturbances, as
“patchy” tracer distribution, which has no relationship with in main pulmonary trunk stenosis. Therefore, it is required to
congenital heart disease and is completely reversible. integrate 99mTc-MAA scanning with echocardiography,
The radiopharmaceutical is slowly injected into a periph- which can offer a detailed evaluation of the central pulmo-
eral vein, avoiding whenever possible central lines, which nary bed, but is less sensitive to peripheral abnormalities of
may lead to “hot spot” artifacts. The injection site has to be pulmonary blood flow.
carefully chosen in some complex malformation or after Besides diffuse hypoperfusion related to congenital heart
some type of surgery, when shunts are present between sys- disease, it is possible to observe also focal perfusion defects.
temic circulation and pulmonary bed, as is the case for cavo- Very infrequently they are expression of pulmonary embo-
pulmonary anastomosis. In this case it is required to split the lism, which is relatively rare in the pediatric population and
dose and to inject in multiple selected sites (i.e., upper and most often associated with a central venous catheter. The
lower limbs). most frequent causes of focal lung hypoperfusion in children
Planar images are acquired, few minutes after injection, with congenital heart disease are peripheral vascular abnor-
in the anterior and posterior projection. Oblique views may malities (distal stenosis, systemic collaterals, etc.) or surgery,
be added when dubious findings are observed. as in some cases of Blalock-Taussig shunt, where isolated
Relative split lung perfusion is calculated by geometric apical hypoperfusion can be observed as a consequence of
mean, drawing regions of interest (ROI) on the posterior and vascular distortion during the procedure.
Fig. 36.17 A 10-year-old

girl. Correction of Fallot a b
tetralogy at the age of
8 months. (a) Lung perfusion
scintigraphy (99mTc-MAA)
showing marked
hypoperfusion of the left lung.
(b) Follow-up scan
demonstrates significant
improvement after left
pulmonary artery dilatation
L= 8% R= 92%
L= 23% R= 77%
POSTERIOR
in deep modifications of ventricular chambers morphology

Key Learning Points (hypertrophic or systemic right ventricle, etc.).
• Tc-MAA particle number must be scaled down,
99m
The most frequent indications for myocardial perfusion
taking into account right-to-left shunting when imaging in children are anomalies of the coronary arteries, as
present. the anomalous origin of the left coronary artery from the pul-
• Knowledge of initial anatomy and surgery is essen- monary artery (ALCAPA); in these patients, evaluation of
tial for correct interpretation of the images. myocardial perfusion is sometimes required in the postoper-
• Relative distribution (L/R %) is the most significant ative follow-up, to ascertain recovery of perfusion. Surgical
information. manipulation of the coronary arteries, as in the arterial switch
• Results must be interpreted integrating anatomical operation for the transposition of the great arteries (TGA),
information from US and/or cardiac catheterization. often necessitates a follow-up evaluation of myocardial
perfusion.
Chest pain is common in children, but in the vast majority
36.8.2 Myocardial Perfusion Scintigraphy of cases, it is noncardiac in origin and is most frequently
related to musculoskeletal or respiratory causes. Nevertheless,
Myocardial perfusion scintigraphy (MPS) is performed when the pain is exertion-related or coronary abnormalities
using 99mTc-sestamibi or 99mTc-tetrofosmin, which are pre- are present or suspected, myocardial perfusion SPECT can
ferred to 201Tl-chloride because of their more favorable phys- usefully integrate echocardiography [49].
ical properties resulting both in better image quality
(including suitability for SPECT imaging) and in lower radi-
ation burden to patients. Key Learning Points
Activity to be injected is usually calculated by weight- • 99m
Tc-labeled perfusion tracers offer better image
based normalization of the standard adult dose (see EANM quality and radiation safety profile than
Dosage Card Calculator or the SNM Pediatric Injected 201
Tl-chloride.
Activity Tool [32, 33]). • In younger children the adenosine stress test is pre-
Myocardial perfusion SPECT may require sedation in chil- ferred over physical stress.
dren up to 5–6 years of age; in this case, fasting is prescribed. • Acquisition parameters (zoom!) must be adapted to
Drugs interfering with heart rate (as typically beta- heart size (age-dependent).
blockers) should be discontinued before stress testing, • Application of iterative reconstruction and “resolu-
whereas xanthine-containing medications and beverages tion recovery” algorithms are strongly
should be avoided in the days preceding a pharmacological recommended.
stress using adenosine or regadenoson. • Indications for myocardial perfusion scintigraphy
A physical stressor is preferable, because it is more physi- include coronary artery anomalies and/or follow-up
ologic, but in younger children the lack of suitable equip- after certain surgical techniques employed to cor-
ment and the reduced compliance make pharmacological rect complex congenital anomalies.
stress more practical. Adenosine (0.14 mg/kg/min i.v. over • Myocardial perfusion scintigraphy is also indicated
6 min) is usually preferred because it has a very short bio- in patients with exercise-induced chest pain,
logical half-life (around 10 s) and very mild, transient side although this condition is rather rare in children
effects. Main contraindications for both drugs are atrioven- without congenital heart anomalies.
tricular block, asthma, and wheezing. Radiopharmaceutical
injection is performed after 3 min of adenosine infusion, thus
leaving 3 additional minutes of pharmacological stimulation
on the coronary circulation, which assures an optimal tracer 36.9 Gastroenterology
distribution.
Image acquisition requires a dual-head SPECT camera in 36.9.1 Gastroesophageal Reflux
90° configuration (180° orbit), using high-resolution or
ultrahigh-resolution collimators with adequate zooming Gastroesophageal reflux (GER), i.e., leaking of gastric
factors, to compensate for the small size of the heart. Small contents back into the esophagus, is very common in
size of the heart must be taken into account also in the recon- infants, probably as a consequence of incomplete matura-
struction process, using iterative algorithms and resolution tion of the lower esophageal sphincter (LES). As the LES
recovery whenever possible. reaches functional normalization, reflux episodes become
SPECT/CT acquisition, integrating low-dose CT, can be less frequent and mostly disappear between 18 and
particularly helpful when cardiac congenital anomalies result 24 months of age.
Gastroesophageal reflux disease (GERD) usually arises It is possible to acquire a fast dynamic study after the
when recurring reflux episodes result in symptoms and com- first single swallowing act (2–3 frames/s for 1 min) to
plications, such as regurgitation/vomiting and dysphagia or investigate esophageal transit. This phase is better imaged
wheezing; in most severe cases, GERD causes recurrent with the child sitting upright, leaning against the gamma
pneumonia, which is particularly frequent in neurologically camera.
disabled children. Symptoms are often vague and nonspe- Late images are usually acquired 2–4 h after the end of
cific, leading to diagnostic delay and uncertainty. the dynamic study, to evaluate gastric emptying and to detect
Gastroesophageal reflux scintigraphy (“milk scan”) is a pulmonary aspiration. The value of 24-h images is debated:
noninvasive technique for the detection of gastroesopha- they could increase the chance of detecting pulmonary aspi-
geal reflux. Mixing a small amount of radioactivity with a ration during sleep, but count statistics is often too low to
standard meal (usually milk or milk formula) allows the yield reliable diagnostic images.
scintigraphic evaluation of reflux and gastric emptying Images are reviewed in cine loop, and gastroesophageal
(when required), without interfering with physiological reflux is detected when activity appears in the esophagus.
activity [50]. Reporting should include the number of episodes, the esoph-
Fasting for 4–6 h is necessary and the study should be ageal level reached by the tracer and their duration. It is pos-
timed to fit in the normal feeding schedule of the child. sible to obtain an activity/time curve by drawing a region of
Sedation is contraindicated post-feeding, as it could alter interest on the esophagus; nevertheless, accurate assessment
motility of the upper digestive tract. for motion artifacts is recommended.
Volume and composition of the meal should correspond,
as much as possible, to those of the daily routine. An insuf-
ficient volume will decrease sensitivity for reflux and for
gastric emptying disturbances. The same holds true for Key Learning Points
administration through nasogastric feeding or a permanent • A standard meal (volume and composition) con-
gastric feeding tube. taining a trace amount of either 99mTc-sulfur colloid
It is advisable to split the meal volume into two volumes or 99mTc-DTPA.
(1/3 and 2/3). The smaller volume, containing the tracer • The early dynamic phase can be combined with late
radionuclide, will be administered first, followed by the acquisitions (90–180 min), to evaluate gastric
remaining two thirds, to wash down activity left in the mouth, emptying.
oropharynx, and esophagus or from the feeding tube. The • It is the most physiologic method for evaluating GER.
meal should not last over 10–15 min, and utmost care is
required to avoid any spillage, positioning absorbent sheaths
to prevent superimposed activity, particularly on the lung 36.9.2 Gastric Emptying Scintigraphy
fields.
The preferred radiopharmaceutical for GER scintigraphy Various disturbances in the delicate interrelated mecha-
is 99mTc-sulfur colloid (where it is available, either commer- nisms controlling gastric emptying may lead to symptoms
cially or as an in-house preparation), because it can with- ranging from early satiety to nausea, abdominal pain and/or
stand the low pH of the stomach without breaking and its vomiting. Delayed gastric emptying can contribute to gas-
absorption from the GI or respiratory mucosa is almost troesophageal reflux, particularly in neurologically disabled
absent; these features keep the background activity at a mini- patients.
mum, thus increasing sensitivity for reflux and/or pulmonary Meal labeling offers the possibility of a truly physiologic
aspiration. A good alternative is 99mTc-DTPA, which is also study, but the number of factors involved in gastric emptying
very stable in gastric juice. (density, calories, fat and protein content, etc.) makes stan-
A very low amount of radioactivity is administered; many dardization very difficult. Solid meals yield more reproduc-
centers administer a 10–20 MBq fixed activity in children ible results, but they are not an option in infants and smaller
under 5 years, thus keeping radiation exposure at a mini- children or when the patient requires enteral tube feeding. In
mum, in the order of 0.8–1.2 mSv. infants the study is usually part of a gastroesophageal reflux
Dynamic images (30 s/frame) are acquired over 60 min in scintigraphy, using colloid-labeled milk.
the posterior view, including the mouth and the upper abdo- In older children 99mTc-sulfur colloid is added to a beaten
men in the field of view, adjusting the zoom to size of the egg or egg white, and the mix is cooked. An interesting alter-
child. When evaluation of the gastric emptying is required, native is labeling a standardized industrial semisolid meal
the entire stomach has to be included. Anterior images are with 99mTc-sulfur colloid or 99mTc-DTPA.
commonly omitted, because it is difficult to position and Fasting is required (4–6 h), and drugs interfering with
control the child under the detector or in the prone position. gastric motility should be discontinued, except when per-
forming the scan for evaluating therapeutic activity. Duration The radionuclide salivagram has no significant contrain-
of the meal should be kept below a fixed threshold (e.g., dications, does not require patient cooperation, and results in
10 min), to increase reproducibility. very low radiation exposure (0.7–1.3 mSv). Therefore, it can
A gastric activity/time curve is obtained from decay- be safely performed in debilitated noncooperating children
corrected images (30 s/frame, dynamic for 60 min, with late without sedation (which would interfere with swallowing). It
acquisitions at 2 and 3–4 h), calculating the emptying per- is indicated whenever salivary aspiration is suspected, as is
centage at 1, 2, and 3–4 h. The wide variability in meal com- the case in many neurological disabilities, where it can com-
position, preparation, and administration has hampered the plement gastroesophageal reflux scintigraphy and gastric
creation of a reference standard for gastric emptying time in emptying scintigraphy, providing a full physiological nonin-
children, although many centers have developed internal ref- vasive assessment.
erence values that are usefully implemented in the clinical
practice.
36.9.4 Ectopic Gastric Mucosa (Meckel
Diverticulum)
36.9.3 Pulmonary Aspiration
Meckel diverticulum, the most common (2–4%) congenital
Dysfunction of the swallowing mechanism often leads to anomaly of the small intestine, is due to incomplete oblitera-
aspiration of foreign materials (liquid or solid foods, saliva) tion of the omphalomesenteric duct (also known as vitelline
in the respiratory tract; this occurrence is often associated duct or yolk stalk). Most cases remain asymptomatic, but
with neurological disabilities. Symptoms vary widely with complications may develop when the Meckel diverticulum
age and neurological status of the child, from choking epi- contains ectopic gastric mucosa. These complications
sodes during feeding to recurrent wheezing, asthma, and include obstruction, inflammation (diverticulitis), and perfo-
recurrent pneumonia, which, if untreated, may lead to per- ration. A frequent complication is lower gastrointestinal
manent lung damage. bleeding, leading to repeated episodes of melena or hemato-
Gastroesophageal reflux scintigraphy may identify aspi- chezia (most often painless bleeding). All such complica-
ration secondary to a reflux episode or during swallowing, if tions are caused by acid secretion by the ectopic gastric
the esophageal transit is evaluated. Late images (up to 24 h) mucosa, which causes mucosal ulceration since this portion
have been proposed, but the global sensitivity of a “milk of the intestine is not protected by the normal defense mech-
scan” for pulmonary aspiration remains low, even if it more anism that protects the normal gastric mucosa from its own
sensitive than radiographic barium upper GI swallows. acid secretions.
Beyond alimentary aspiration, defective swallowing often Since 99mTc-pertechnetate (99mTcO4−) accumulates in sur-
causes salivary aspiration, which may also cause recurrent face cells of gastric mucosa, scintigraphy with 99mTcO4−
pulmonary infections, by transporting the normal oropharyn- allows the identification of Meckel diverticulum when
geal flora to the lower respiratory tract. ectopic gastric mucosa is present.
The radionuclide salivagram has the unique capability of Fasting (4–6 h) is recommended, and premedication with
documenting aspiration of saliva in a simple, cheap, and non- gastric secretion inhibitors (H2-antagonists or proton-pump
invasive way. The test requires the placement of a single drop inhibitors) during the 12–24 h before injection helps to
(0.1 mL or less) of a 99mTc-labeled tracer (99mTc-sulfur col- reduce radiotracer release from the gastric mucosa, thus
loid or 99mTc-DTPA, 10 MBq) under the tongue. No prepara- increasing image contrast.
tion is required, but absorbent sheaths should be positioned A weight-scaled activity of 99mTcO4− (see EANM Dosage
to prevent spillage and contamination. Imaging starts imme- Card Calculator or the SNM Pediatric Injected Activity Tool
diately after tracer administration, in the posterior view [32, 33]) is injected i.v., and dynamic images (128 × 128
(child lying supine), using a low-energy parallel-hole colli- matrix, 1 min/frame, low-energy high-resolution parallel-
mator. Dynamic images (128 × 128 matrix, 10–30 s/frame) hole collimator) are acquired for 60 min in the anterior view,
are acquired for 60 min, including the oropharynx and thorax including the whole abdomen in the field of view.
in the field of view. Static images are acquired after the Meckel diverticulum is visualized as a focal spot of
dynamic phase, positioning radioactive markers on the uptake, which can be localized throughout the whole abdo-
xiphoid process, suprasternal notch and axillae, to assist men. Ectopic mucosa usually appears simultaneously with
in localizing tracer distribution. the stomach, but delayed appearance, up to 40–50 min, is
Images are reviewed in cine loop and inspected for activ- possible, particularly in younger children (Fig. 36.18).
ity appearing in the trachea and/or bronchi. The level of Particular care is needed to avoid misinterpretation of
tracer penetration should be noted, because distal aspiration physiologic activity accumulation in the abdomen, espe-
is probably more relevant than aspiration limited to the tra- cially in the urinary system, by acquiring additional views
chea [50]. (lateral, post-void, etc.) [51, 52].
ANT
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
16 17 18 19 20
Fig. 36.18 A 4-year-old boy. Hematochezia over the last 72 h. Focal uptake in the dynamic 99mTc-pertechnetate scan, confirming the diagnostic
hypothesis of ectopic gastric mucosa in Meckel diverticulum
Key Learning Points • Beware of urinary activity (lateral views, post-void,

• The 99mTc-pertechnetate scan detects and localizes etc., as required).
Meckel diverticula containing ectopic gastric • Prolonged acquisition (>30–45 min) recommended
mucosa. if early pathological uptake is absent (possible
• Premedication with gastric secretion inhibitors delayed visualization).
enhances image quality.
References 23. Matcuk GR Jr, Mahanty SR, Skalski MR, Patel DB, White EA,
Gottsegen CJ. Stress fractures: pathophysiology, clinical presen-
tation, imaging features, and treatment options. Emerg Radiol.
1. Bielsa IR. Pediatric nuclear medicine and its development as a spe-
2016;23:365–75.
cialty. Semin Nucl Med. 2017;47:102–9.
24. Carrol PS, Pieter GH, Raijmakers M, Tuinzing DB. Comparison
2. Gelfand MJ, Clements C, MacLean JR. Nuclear medicine pro-
of planar bone scintigraphy and single photon emission computed
cedures in children: special considerations. Semin Nucl Med.
tomography in patients suspected of having unilateral condylar
2017;47:110–7.
hyperactivity. Oral Surg Oral Med Oral Pathol Oral Radiol Endod.
3. Accorsi R, Karp JS, Surti S. Improved dose regimen in pediatric
2008;106:426–32.
PET. J Nucl Med. 2010;51:293–300.
25. Williams JL, Paul D, Bisset G III. Thyroid disease in children: part
4. Fahey FH. Dosimetry of pediatric PET/CT. J Nucl Med.
2. State-of-the-art imaging in pediatric hyperthyroidism. Pediatr
2009;50:1483–91.
Radiol. 2013;43:1254–64.
5. Strauss KJ, Goske MJ. Estimated pediatric radiation dose during
26. Francis GL, Waguespack SG, Bauer AJ, Angelos P, Benvenga
CT. Pediatr Radiol. 2011;41(Suppl 2):472–81.
S, Cerutti JM, et al. Management guidelines for children with
6. Huda W, Ogden KM, Khorasani MR. Converting dose length prod-
thyroid nodules and differentiated thyroid cancer. The ATA
uct to effective dose at CT. Radiology. 2008;248:995–1003.
Guidelines Task Force on Pediatric Thyroid Cancer. Thyroid.
7. Cravero JP, Havidich JE. Pediatric sedation – evolution and revolu-
2015;25:716–59.
tion. Paediatr Anaesth. 2011;21:800–9.
27. Machac J. Thyroid cancer in pediatrics. Endocrinol Metab Clin N
8. Even-Sapir E, Flusser G, Lerman H, Lievshitz G, Metser
Am. 2016;45:359–404.
U. SPECT/multislice low-dose CT: a clinically relevant constituent
28. Luster M, Pfestroff A, Hänscheid H, Verburg FA. Radioiodine ther-
in the imaging algorithm of nononcologic patients referred for bone
apy. Semin Nucl Med. 2017;47:126–34.
scintigraphy. J Nucl Med. 2007;48:319–24.
29. Luster M, Handkiewicz-Junak D, Grossi A, Zacharin M, Taïeb D,
9. Andersen JB, Mortensen J, Bech BH, Hoigaard L, Borgwardt
Cruz O, et al. Recombinant thyrotropin use in children and adoles-
L. First experiences from Copenhagen with single photon emission
cents with differentiated thyroid cancer: a multicenter retrospective
computed tomography/computed tomography. Nucl Med Commun.
study. J Clin Endocrinol Metab. 2009;94:3948–53.
2011;32:356–62.
30. Keller-Petrot I, Leger J, Sergent-Alaoui A, de Labriolle-Vaylet

10. De Palma D, Nadel HR, Bar-Sever Z. Skeletal scintigraphy with
C. Congenital hypothyroidism: role of nuclear medicine. Semin
SPECT/CT in benign pediatric bone conditions. Clin Transl
Nucl Med. 2017;47:135–42.
Imaging. https://doi.org/10.1007/s40336-016-0169-8.
31. Mendichovszky I, Solar BT, Smeulders N, Easty M, Biassoni

11. Nadel HR. Pediatric bone scintigraphy update. Semin Nucl Med.
L. Nuclear medicine in pediatric nephro-urology: an overview.
2010;40:31–40.
Semin Nucl Med. 2017;47:204–28.
12. Segall G, Delbeke D, Stabin MG, Even-Sapir E, Fair J, Sajdak R,
32. http://www.eanm.org/publications/dosage-calculator/ Accessed 27
Smith GT. SNM practice guideline for sodium 18F-fluoride PET/CT
July 2017.
bone scans 1.0. J Nucl Med. 2010;51:1813–20.
33. http://www.snmmi.org/ClinicalPractice/PediatricTool.aspx
13. Drubach LA. Nuclear medicine techniques in pediatric bone imag-
Accessed 27 July 2017.
ing. Semin Nucl Med. 2017;47:190–203.
34. Biassoni L. Pitfalls and limitations of radionuclide renal imaging in
14. Chiappini E, Camposampiero C, Lazzeri S, Indolfi G, De Martino
pediatrics. Semin Nucl Med. 2015;45:411–27.
M, Galli L. Epidemiology and management of acute haematog-
35. Ismaili K, Piepsz A. The antenatally detected pelvi-ureteric junc-
enous osteomyelitis in a tertiary paediatric center. Int J Environ
tion stenosis: advances in renography and strategy of management.
Res Public Health. 2017;14(5):pii:E477. https://doi.org/10.3390/
Pediatr Radiol. 2013;43:428–35.
ijerph14050477.
36. La Scola C, De Mutiis C, Hewitt IK, Puccio G, Toffolo A, Zucchetta
15. Signore A, Glaudemans AW, Gheysens O, Lauri C, Catalano

P, et al. Different guidelines for imaging after first UTI in febrile
OA. Nuclear medicine imaging in pediatric infection or chronic
infants: yield, cost, and radiation. Pediatrics. 2013;131:e665–71.
inflammatory diseases. Semin Nucl Med. 2017;47:286–303.
37. Herz DB. The top-down approach: an expanded methodology. J
16. Jaramillo D, Dormans JP, Delgado J, Laor T, St Geme JW

Urol. 2010;183:856–7.
III. Haematogenous osteomyelitis in infants and children: imaging
38. Pfluger T, Piccardo A. Neuroblastoma: MIBG imaging and new
of a changing disease. Radiology. 2017;283:629–43.
tracers. Semin Nucl Med. 2017;47:143–57.
17. Weissman R, Uziel Y. Pediatric complex regional pain syndrome: a
39. Lewington V, Lambert B, Poetschger U, Sever ZB, Giammarile F,
review. Pediatr Rheumatol Online J. 2016;14:29.
McEwan AJ, et al. 123I-mIBG scintigraphy in neuroblastoma: devel-
18. Ariza Jiménez AB, Núñez Cuadros E, Galindo Zavala R, Núñez
opment of a SIOPEN semi-quantitative reporting method by an
Caro L, Díaz-Cordobés Soriano G, et al. Recurrent multifo-
international panel. Eur J Nucl Med Mol Imaging. 2017;44:234–41.
cal osteomyelitis in children: experience in a tertiary care center.
40. Lopci E, Piccardo A, Nanni C, Altrinetti V, Garaventa A, Pession A,
Reumatol Clin. 2017;pii:S1699–258.
et al. 18F-DOPA PET/CT in neuroblastoma: comparison of conven-
19. Khanna G, Sato TS, Ferguson P. Imaging of chronic recurrent mul-
tional imaging with CT/MR. Clin Nucl Med. 2012;37:e73–8.
tifocal osteomyelitis. Radiographics. 2009;29:1159–77.
41. Harrison DJ, Parisi MT, Shulkin BL. The role of 18F-FDG-PET/CT
20. http://emedicine.medscape.com/article/392850-overview Accessed
in pediatric sarcoma. Semin Nucl Med. 2017;47:229–41.
27 July 2017.
42. Kluge R, Kurch L, Georgi T, Metzger M. Current role of FDG-
21. Sharma P, Mukherjee A, Karunanithi S, Nadarajah J, Gamanagatti
PET in pediatric Hodgkin’s lymphoma. Semin Nucl Med.
S, Khan SA, et al. 99mTc-Methylenediphosphonate SPECT/CT as
2017;47:242–57.
the one-stop imaging modality for the diagnosis of osteoid oste-
43. Ntsinjana HN, Tann O, Taylor AM. Trends in pediatric cardiovascu-
oma. Nucl Med Commun. 2014;35:876–83.
lar magnetic resonance imaging. Acta Radiol. 2013;54:1063–74.
22. Alkhawaldeh K, Al Ghuweri A, Kawar J, Jaafreh A. Back pain in
44. Vasanawala SS, Hanneman K, Alley MT, Hsiao A. Congenital heart
children and diagnostic value of 99mTc-MDP bone scintigraphy.
disease assessment with 4D flow MRI. J Magn Reson Imaging.
Acta Inform Med. 2014;22:297–301.
2015;42:870–86.
45. Kathiria NN, Higgins CB, Ordovas KG. Advances in MR imaging Further Reading
assessment of adults with congenital heart disease. Magn Reson
Imaging Clin N Am. 2015;23:35–40.
Grant FD. 18F-fluoride PET and PET/CT in children and young adults.
46. Chan FP, Hanneman K. Computed tomography and magnetic reso-
PET Clin. 2014;9:287–97.
nance imaging in neonates with congenital cardiovascular disease.
Kwatra NS, Sarma A, Lee EY. Practical indication-based pediatric
Semin Ultrasound CT MR. 2015;36:146–60.
nuclear medicine studies: update and review. Radiol Clin North Am.
47. Dacher JN, Barre E, Durand I, Hazelzet T, Brasseur-Daudruy M,
2017;55:803–44.
Blondiaux É, et al. CT and MR imaging in congenital cardiac
Lassmann M, Treves ST. EANM/SNMMI Paediatric Dosage
malformations: where do we come from and where are we going?
Harmonization Working Group. Paediatric radiopharmaceutical
Diagn Interv Imaging. 2016;97:505–12.
administration: harmonization of the 2007 EANM paediatric dos-
48. Milanesi O, Stellin G, Zucchetta P. Nuclear medicine in pediatric
age card (version 1.5.2008) and the 2010 North American consen-
cardiology. Semin Nucl Med. 2017;47:158–69.
sus guidelines. Eur J Nucl Med Mol Imaging. 2014;41:1036.
49. Robinson B, Goudie B, Remmert J, Gidding SS. Usefulness of
Nadel HR. SPECT/CT in pediatric patient management. Eur J Nucl
myocardial perfusion imaging with exercise testing in children.
Med Mol Imaging. 2014;41(Suppl 1):S104–14.
Pediatr Cardiol. 2012;33:1061–8.
Parisi MT, Bermo MS, Alessio AM, Sharp SE, Gelfand MJ, Shulkin
50.
Bar-Sever Z. Scintigraphic evaluation of gastroesophageal
BL. Optimization of pediatric PET/CT. Semin Nucl Med.
reflux and pulmonary aspiration in children. Semin Nucl Med.
2017;47:258–74.
2017;47:275–85.
Parisi MT, Eslamy H, Mankoff D. Management of differentiated thy-
51. Vali R, Daneman A, McQuattie S, Shammas A. The value of repeat
roid cancer in children: focus on the American Thyroid Association
scintigraphy in patients with a high clinical suspicion for Meckel
pediatric guidelines. Semin Nucl Med. 2016;46:147–64.
diverticulum after a negative or equivocal first Meckel scan. Pediatr
Sharp SE, Trout AT, Weiss BD, Gelfand MJ. MIBG in neuroblastoma
Radiol. 2015;45:1506–14.
diagnostic imaging and therapy. Radiographics. 2016;36:258–78.
52. Biassoni L, Easty M, Sinha C. Technetium-99m-pertechnetate scin-
Uslu L, Donig J, Link M, Rosenberg J, Quon A, Daldrup-Link
tigraphy in children with symptomatic Meckel’s diverticulum. Nucl
HE. Value of 18F-FDG PET and PET/CT for evaluation of pediatric
Med Commun. 2015;36:406–7.
malignancies. J Nucl Med. 2015;56:274–86.
Hybrid Imaging of the Peripheral
Lymphatic System 37
Paola Anna Erba, Roberto Boni, Martina Sollini,
Andrea Marciano, Rossella Di Stefano,
Contents
37.1 Anatomy and Physiology of Peripheral Lymphatic Circulation 952
37.2 Lymphedema 954
37.2.1 Lymphangitis 956
37.2.2 Lymphatic Malignancies 957
37.2.3 Intracavitary Lymphedema 957
37.3 Workup of Patients with Lymphedema... 958
37.4 Role of Imaging in the Management of Lymphedema 958
37.5 Treatment and Management of Patients with Lymphedema 959
37.6 Peripheral Lymphoscintigraphy 960
37.6.1 Radiopharmaceuticals 960
37.6.2 Injected Volume and Activity 961
37.6.3 Factors Affecting Kinetics of Radiocolloids Injected Interstitially 961
37.6.4 Injection Sites and Method 961
37.6.5 Imaging 963
37.6.6 Visual Interpretation 964
37.6.7 Lymphoscintigraphy with Stress Test........... 964
37.6.8 Quantitative Lymphoscintigraphy 965
37.6.9 Absorbed Dose 968
37.7 Lymphoscintigraphy in the Management of Lymphedema 971
References 973
P. A. Erba (*) · A. Marciano · G. Mariani

Learning Objectives
Research and Advanced Technologies in Medicine, University of • Acquire basic notions of the anatomy and physiology
Pisa, Pisa, Italy of the peripheral lymphatic circulation.
e-mail: paola.erba@unipi.it • Define the main diseases affecting the peripheral lym-
R. Boni phatic circulation: lymphedema, lymphangitis, lym-
Nuclear Medicine Service, “Papa Giovanni XXIII” Hospital, phatic malignancies, and intracavitary lymphedema.
Bergamo, Italy
• Acquire basic notions on lymphedema including etiol-
M. Sollini ogy, pathophysiology, clinical signs, classification,
Department of Biomedical Sciences, Humanitas University,
Milan, Italy
and staging system.
• Acquire basic knowledge of lymphangitis including
R. Di Stefano
Cardiovascular Disease Unit, Department of Surgical, Medical,
causes, pathophysiology, and clinical signs.
Molecular Pathology and Medical Emegencies, University of Pisa, • Acquire basic knowledge of the main lymphatic
Pisa, Italy malignancies.

• Acquire basic knowledge of different types of intra- tions along the drainage route where fluid and cell exchange
cavitary lymphedema: chylous ascites, chylothorax, between blood and lymph occur), and other lymphoid tis-
and chylopericardium, including etiology and sue, particularly the spleen and bone marrow. The lymphatic
pathophysiology. system, initially recognized solely as a network of pathways
• Understand how to implement the workup of patients for the removal of tissue products, has a significant role in
with lymphedema, including treatment and management. maintaining the balance of tissue fluid volume and an appro-
• Understand the role of imaging in the management of priate cellular chemical environment (the capillary plasma
lymphedema patients. filtration process). An important role of the lymphatic circu-
• Learn the main clinical indications for peripheral lym- lation is the transport of the antigen and immune cells from
phoscintigraphy and the role of lymphoscintigraphy in peripheral tissues to lymph nodes. The interaction between
the management of lymphedema. lymphatic endothelial cells and leukocytes controls the
• Learn how to perform peripheral lymphoscintigraphy immune cell migration and immune responses by regulating
including knowledge of available radiopharmaceuti- lymphatic flow and various secreted molecules such as che-
cals, injection procedures and acquisition protocols, mokines and cytokines [1, 2] in tissue.
dosimetry, image analysis and interpretation. The lymphatic system can be divided into peripheral
(from the interstitial space to and within the nearest lymph
node) and central (efferent lymphatics, cisterna chyli, and
37.1 Anatomy and Physiology thoracic duct, all lymphatic organs) [1]. The lymphatic
of Peripheral Lymphatic Circulation organs (spleen, thymus, pharyngeal lymphoid ring, and lym-
phoid tissue in the gut) are composed of connective tissue
The lymphatic circulation is part of the peripheral circula- and lymphocytes. The lymphatic vessels are a complex net-
tion, linking blood circulation at its origins (the interstitial work, similar to arteries and veins (Fig. 37.1). Along the
space) and at its final drainage point (the thoracic duct). To course of these thin-walled vessels, there are lymph nodes,
a large extent, the anatomy of lymphatic channels parallels lymphoid tissue formations that receive lymph from afferent
that of the veins, and the two systems show many similari- vessels and let it flow centripetally through the efferent lym-
ties in structure and function. The lymphatic system includes phatic vessels toward higher-echelon lymph nodes or the
fluid (lymph), the lymphatic vessels, the lymph nodes (sta- major collecting lymph trunks.
Fig. 37.1 Diagrammatic endothelium smooth muscle adventitia

representation of the structure
of different components of the
vascular system with variable
complexity from the blood lymphatic
capillary to different levels of capillary vessel valves
the arterial and the venous
sides of the general
circulation, as well as of an
initial lymphatic vessel. The
most important difference arteriole venule
between blood vessels and
lymphatic vessels is that
lymphatic vessels have
virtually no smooth muscle,
and their adventitia is very small small vein
thin. Relative proportions of artery
smooth muscle and adventitia
change considerably at smooth muscle
in adventitia
different levels of blood
vessels. Lumen of arterial
vessels is more regularly
shaped than lumen of venous
vessels (modified from http:// medium/large
vein valves
slideplayer.com/ artery
slide/5227321/)
37 Hybrid Imaging of the Peripheral Lymphatic System 953
Lymph nodes may be single, chained, or grouped and may • Lymph capillaries: the lymphatic circulation begins with
have different sizes from few millimeters to some centimeters. blind endings in the peripheral tissues; they have many
Lymph nodes are the “stations” of the lymphatic system anastomoses which form a very fine network.
which act as filters and operate in the immune defense. The • Pre-collecting vessels: short and thin, they connect the
lymphatic vascular network consists of small blind-ended lymph capillaries to the collecting lymphatics; they have
capillaries and larger collecting vessels. The lymphatic capil- rare valves that are not sufficient to avoid lymph flux
laries are composed of a single layer of overlapping endothe- inversion.
lial cells and lack a continuous basement membrane and • Collecting lymphatics: larger contractile lymphatics which
pericytes (i.e., the smooth muscle-like contractile cells that have valves as well as smooth muscle walls that may be
wrap around the outer surface of blood vessels). Therefore, divided into deep (with a course generally satellite of
the lymphatic capillaries are highly permeable to interstitial blood vessels) and superficial (with a course often inde-
fluid and macromolecules, such that, when the surrounding pendent from blood vessels). Along their course there are
interstitial pressure changes, these lymphatics either expand the lymph nodes. Collecting lymphatics are a wide and
and fill with lymph or contract and push lymph—thanks to a complex outflow system located at pre- and post-nodal
system of valves similar to valves in veins. These lymphatic level, anastomosing with other collecting lymphatics.
capillaries first drain into pre-collecting lymphatic vessels, These vessels form the afferent or pre-lymph nodal collec-
which merge into larger secondary collecting lymphatic ves- tors. The efferent or post-lymph nodal collector, a new
sels covered by smooth muscle cells that provide contractile vessel with identical morphological characteristics to the
activity to assist lymph flow and possess a continuous base- afferent vessel, leaves the lymph node. Afferent collectors
ment membrane. Tissue fluid collected in the larger collect- carry unfiltered lymph into the node, while efferent collec-
ing lymphatic vessels drains into the thoracic duct and is then tors carry filtered lymph out of the node.
returned to the blood circulation through lymphatic–vascula- • Lymphatic trunks: the fluid is then transported to pro-
ture connections at the junction of the jugular and subclavian gressively larger lymphatic vessels formed by the conflu-
veins [3]. Proceeding from the periphery, the lymphatic ves- ence of the efferent collectors of the most important
sels can be identified as (Fig. 37.2): lymph node groups that culminate in the right lymphatic
Tissue fluid tissue cells

a Lymphatic Pulmonary b
capillaries capillaries lymphatic capillaries
Lymph Pulmonary
node circuit
lymphatic
Valve duct
Lymphatic
vessel
blood capillary
Systemic venule
circuit arteriole
Lymph
node
Lymphatic Systemic
capillaries Tissue fluid capillaries
Fig. 37.2 (a) Schematic representation of the relationship between right subclavian vein) and through the major thoracic duct (which
blood circulation (arterial system in red, venous system in blue) and drains the remainder of the body into the left subclavian vein). (b)
lymphatic circulations (in green). Lymphatic capillaries drain tissue Magnified schematic view of the tight interconnections between vascu-
fluid, macromolecules, and cell debris from tissues, merging into major lar and lymphatic circulation in peripheral tissues ((a) from http://ency-
lymphatic vessels that direct lymphatic fluid to lymph nodes. Lymph clopedia.lubopitko-bg.com/Lymphatic_Circulation.html; (b) from
returns to the blood circulation through the right lymphatic duct (which http://www.med-ars.it/galleries/lymphnodes_4.htm)
drains the right arm, right portion of the thorax, neck, and head into the
duct (for lymph from the right upper body) and the tho-
racic duct (for the rest of the body); both ducts drain into • The lymphatic vessels can be classified as lymph
the circulatory system at the right and left subclavian capillaries, pre-collecting vessels, collecting lym-
veins. phatics, and lymphatic trunks.
• Lymph nodes may be single, chained, or grouped
The lymphatic vasculature is essential for the adsorption and may have different sizes from few millimeters
of lipids from the intestine. The major products of lipid to some centimeters. Lymph nodes are the “sta-
digestion (fatty acids and monoglycerides) enter the entero- tions” of the lymphatic system which act as filters
cyte by simple diffusion or via a specific fatty acid trans- and operate in the immune defense.
porter protein in the membrane. Once inside the enterocyte, • The lymphatic system has a significant role in
fatty acids and monoglyceride are transported into the endo- maintaining the balance of tissue fluid volume. The
plasmic reticulum, where they are used to synthesize tri- lymphatic system has a significant role also in
glycerides. Beginning in the endoplasmic reticulum and maintaining an appropriate cellular chemical envi-
continuing in the Golgi apparatus, triglycerides are pack- ronment, controls the immune cell migration and
aged with cholesterol, lipoproteins, and other lipids into par- immune responses by regulating lymphatic flow
ticles called chylomicrons. Instead of being adsorbed and various secreted molecules.
directly into capillary blood as in the case of sugars and
amino acids, chylomicrons are transported first into the lym-
phatic vessels that penetrate into each intestinal villus.
Further, chylomicron-rich lymph drains into the lymphatic 37.2 Lymphedema
system, which rapidly flows into the blood. Blood-borne
chylomicrons are rapidly disassembled and their constituent Under normal circumstances, water acts predominantly as a
lipids utilized throughout the body. Intestinal lymphatics solvent or vehicle for the colloids, cells, and other materials
offer a specialized absorption pathway through which lipids, that can be drained only via the lymph system. This role of
fat-soluble vitamins, lipophilic xenobiotics, and lipophilic the lymphatic circulation as a “safety valve” against fluid
drugs can gain access to the systemic circulation. Compounds overload involves the lymphatic system in every form of
absorbed by the intestinal lymphatics drain via the cisterna edema. In most cases, edema occurs in the extracellular liq-
chili and the thoracic duct, thus entering systemic circula- uid compartment but may involve the intracellular compart-
tion at the junction of the left internal jugular vein with the ment as a result of depression of the metabolic tissue system
left subclavian vein. or lack of proper cell nutrition. On the other hand, the extra-
Lymph is a transparent, straw yellow, or milky fluid, con- cellular edema is the result of an alteration of the physiologi-
taining sugars, proteins, salts, lipids, amino acids, hormones, cal equilibrium existing between capillary filtration and
and vitamins. It has a composition comparable to that of lymphatic drainage.
blood plasma, but it differs slightly for the protein content; In general, lymphedema may be defined as a pathologi-
since the lymph is derived from the interstitial fluid, its com- cal condition in which the increase in the size of a body area
position continually changes as the blood and the surround- is the direct consequence of a decrease in lymphatic drain-
ing cells continually exchange substances with the interstitial age occurring without any increase in capillary filtration.
fluid. This condition recognizes several causes and can occur as a
result of different mechanisms including congenital dyspla-
sia due to genetic mutation or for the presence of hereditary
syndromes or anatomical obliteration of lymphatic vessels
Key Learning Points (see Table 37.1). Primary lymphedemas can be congenital
• The lymphatic system includes fluid (lymph), the (either in a sporadic form or in a familial form), in turn dis-
lymphatic vessels, lymph nodes (stations along the tinguished as praecox of tardal according to the age of
drainage route where exchange of fluid and cells onset. Secondary lymphedemas can be caused by surgery,
between blood and lymph occurs), and other lym- radiation therapy, trauma, lymphangitis, or infection.
phoid tissue (spleen and bone marrow). Primary lymphedema is a rare condition (<0.5% of the
• The lymphatic system can be divided into periph- general population) which may occur at birth (congenital),
eral (from the interstitial space to and within the during puberty (praecox), or, rarely, in adulthood (tardal) as
nearest lymph node) and central (efferent lymphat- a consequence of the imperfect development of the lym-
ics, cisterna chyli, and thoracic duct, all lymphatic phatic vascular system in utero. In the familial form (as with
organs). Milroy’s disease and Meige’s syndrome), or genetic, such as
those associated with Turner’s and Noonan’s syndromes or
Table 37.1 Mechanisms and causes of lymph drainage failure retroperitoneal), radiotherapy, and direct tumor cells inva-
Mechanism Causes sion. In the industrialized world, secondary lymphedema is
Reduction of lymphatic Aplasia-hypoplasia of the whole mostly iatrogenic, while in developing countries it is mainly
pathways vessel due to infectious diseases such as filariasis, cellulitis, acute
Acquired obliteration of the lymphangitis, tuberculosis, venereal disease, and leprosy. In
lymphatic lumen
Poorly functioning Failure of pump contractility
such cases, an alteration in the flow of the lymph is mani-
lymphatic pathways fested through the progressive obstruction (i.e., sclerosis) of
Obstructed lymphatic Scar from lymphadenectomy, lymphatic collectors or through lymphangiothrombosis.
pathways radiation therapy, or infection Lymphedema frequently coexists with lipedema, a bilat-
Incompetent lymphatic Megalymphatics eral, symmetrical swelling of the lower extremities extend-
pathways with reflux Lymphatic hyperplasia ing from the pelvic brim to the ankles affecting only women.
Reproduced with modifications from: Di Stefano R, Erba PA, D’Errico Histologically, its hallmark is a gross increase in the subcu-
G. The pathophysiology of lymphatic circulation in different disease
conditions. In: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés
taneous fat layer. The patient may be normal in weight, or
Olmos RA, Eds. Atlas of Lymphoscintigraphy and Sentinel Node even thin in the upper half of the body, but grossly obese
Mapping – A Pictorial Case-Base Approach. Milan: Springer-Verlag; from the pelvic brim down. Lymph vessels in lipedema are
2013: 7–16 coiled, the prelymphatic canals and initial lymph vessels are
abnormal, lymph transport velocity is reduced, and lymph-
the congenital vascular Klippel-Trenaunay syndrome, mal- edema involves both lower extremities.
formed lymphatics coexist with an aberrant venous system. A special case of lymphedema occurs in patients with
Sporadic cases of primary lymphedema are more frequent cancer (where the tumor obstructs lymphatic vessels) or as
than the familial or genetic associated forms, being distal the result of anticancer treatment(s). The most common
hypoplasia of the legs’ lymphatics the most common cause example is upper arm lymphedema occurring after axillary
of primary lymphedema of lower limbs. lymph node dissection for breast cancer surgery and/or radi-
In hereditary human lymphedemas, mutations involve ation therapy. The incidence of breast cancer-related lymph-
genes that are crucial during lymphovascular development. edema ranges from 6% to 70%; however, it may be a common
For example, the FOXC2 gene is the hallmark mutation in underreported morbidity, considering the close connections
the lymphedema-distichiasis syndrome (LD), an autosomal between lymphatic drainage of the upper extremity and of
dominant disorder with variable expression that represents the breast in the caudal part of the axilla [7]. Lymphedema is
one of the most common forms of primary lymphedema. The generally localized in the arm on the side of breast surgery,
FOXC2 gene encodes for FOXC2, a transcription factor and patients with a high peripheral blood vascular filtration
involved in the development of the lymphatic and vascular rate seem to be predisposed to this complication [8]. Although
system [3]. In this syndrome, lymphedema of the limbs and a clear relationship between the number of lymph nodes
distichiasis (double row of eyelashes) occur during late removed and the risk of lymphedema has not been definitely
childhood or pubertal age. Venous insufficiency is also pres- established, some correlation was found between the number
ent in half of the patients. Other associations have been of removed axillary lymph nodes and the risk of lymph-
reported, including congenital heart disease, ptosis, cleft lip/ edema. Arm swelling is also more likely to occur in over-
palate, and spinal extradural cysts. Another example of weight women, or in patients receiving taxane-based
genetic primary lymphedema is the Milroy’s disease, a con- chemotherapy [9]. Currently, clinical trials are focusing on
dition characterized by early-onset congenital lymphedema. the reduction of axillary lymph node dissections even in the
In this syndrome, heterozygous missense mutations in the presence of a positive sentinel lymph node [10].
FLT4 (VEGFR3) gene inactivate the kinase activity of Primary and secondary lymphedemas share the dysfunc-
VEGFR-3. Other mutations of genes involved in lymphatic tion of lymphatic endothelial cells that make up the vessels
development described in primary lymphedema include and are responsible for the absorption and transport of the
GJC2, CCBE1, PTPN14, GATA2, and SOX18 [4]. The lymph. The alteration of lymphatic transport modifies the
Connell algorithm, a classification framework based on clin- interstitial fluid and the clearance of macromolecules, affect-
ical phenotype originally published in 2010 [5] has more ing the osmotic and hydrostatic gradients maintained by blood
recently been updated to include newly identified genes [6]. vessels. Stagnation of lymph in the interstitium increases the
Secondary or acquired lymphedema is more common colloid osmotic pressure, further promoting the accumulation
than primary lymphedema, and it may be caused by different of fluid. All these processes cause regional ischemia and result
conditions including injury (e.g., burns, wide and peripheral in the inflammatory reactions that characterize chronic lym-
wounds of the extremity) or functional deficiency (e.g., phatic failure. In the extremities, the persistence of edema
paralysis), pregnancy, contact dermatitis, rheumatoid arthri- results in hypercellularity of the skin, progressive fibrosis, and
tis, radical lymph node dissection (e.g., axillary, inguinal, pathological increase in the deposition of subcutaneous and
Table 37.2 Classification of limb lymphedema according to the

International Society of Lymphology • Lymphedema frequently coexists with lipedema.
Stage Characteristics Lymph vessels in lipedema are coiled, the prelym-
Stage 1 (a) Absence of edema in the presence of alterations of phatic canals and initial lymph vessels are abnor-
lymphatic vessels mal, lymph transport velocity is reduced, and
(b) Slight edema reversible with declining position and
lymphedema involves both lower extremities.
nighttime rest
Stage 2 Persistent edema only partially reversible with declining • Primary and secondary lymphedemas share the
position and nighttime rest dysfunction of lymphatic endothelial cells.
Stage 3 Persistent (not spontaneously reversible with declining • Chronic lymphatic failure is characterized by
position) and progressive (acute erysipeloid inflammatory reactions resulting from regional
lymphangitis) edema
ischemia due to alteration of lymphatic transport,
Stage 4 Fibrotic lymphedema (with early lymphostatic
verrucosis) and a column-shaped limb abnormal clearance of macromolecules, stagnation
Stage 5 Elephantiasis with severe limb deformity; sclero- of lymph in the interstitium, and increased colloid
indurative pachydermitis and lymphostatic verrucosis osmotic pressure.
(modified with permission from: Gasbarro V, Michelini S, Antignani PL, • In the extremities, the persistence of edema results
Tsolaki E, Ricci M, Allegra C. The CEAP-L classification for lymphede- in hypercellularity of the skin, progressive fibrosis,
mas of the limbs: the Italian experience. Int Angiol. 2009;28:315–24) and pathological increase in the deposition of sub-
cutaneous and subfascial adipose tissue.
subfascial adipose tissue. In addition, due to the variable dys- • The International Society of Lymphology has estab-
function of the affected site, lymphedema has a negative lished a staging system for defining the severity of
impact on the quality of life and self-perception of patients, this disease, for monitoring the progression of the
thus possibly resulting in depression, anxiety, and social adap- condition, and for evaluating treatment response.
tation problems.
The classification of limb lymphedema proposed in 2009
is similar to that used for chronic venous insufficiency, which
considers the clinical, etiologic, anatomic, and pathophysio- 37.2.1 Lymphangitis
logic (CEAP) factors. It adopts the acronym CEAP and adds
the letter “L” (lymphatic) to underline the aspect “lymph- Lymphatic vessels, together with lymph nodes, constitute a
edema” [11]. The International Society of Lymphology (ISL) host defense system preventing or limiting noxious agents
has established a staging system for defining severity of this to enter the systemic circulation. When the inflammatory
disease (Table 37.2). It is thus possible to identify the progres- response is profound, an overt lymphangitis and/or lymph-
sion of the condition and the potential for successful treatment adenitis can arise. Lymphangitis is an inflammation of the
and improvement. This staging system, which applies only to lymphatic collectors and is clinically evident as a red streak
the limbs (arms and legs), is based on the degree of swelling up the limb corresponding to the inflamed vessels. Edema
and on the condition of the skin and tissues. A subsequent ver- is often an accompanying feature. Infection is generally
sion of the ISL lymphedema staging system reduces stages limited to the lymph nodes, and lymphadenitis may give
from five to four levels (graded from 0 to 3) but introduces a rise to painful swelling in the groin or axilla (depending on
more complex description of the clinical picture [12]. the site of infection). Lymphangitis may occur without any
demonstrable inflammation or can be recurrent. When lym-
phatic insufficiency exists and the local system fails in its
Key Learning Points host defense duty, recurrent infection can occur, presenting
• Lymphedema is a pathological condition in which clinically as recurrent cellulitis/erysipelas. Erysipelas is a
the increase in size of a body area is the direct con- skin infection that is usually caused by β-hemolytic group
sequence of a decrease in lymphatic drainage occur- A streptococci. After having had erysipelas in an extremity,
ring without any increase in capillary filtration. a significant percentage of patients develop persistent
• Lymphedema recognizes several causes and can swelling or suffer from recurrent erysipelas caused by sec-
occur as a result of different mechanisms including ondary lymphedema. Patients presenting with a first epi-
congenital dysplasia due to genetic mutation or for sode of erysipelas often have signs at lymphoscintigraphy
the presence of hereditary syndromes or anatomical of pre-existing lymphatic impairment in the other, clini-
obliteration of lymphatic vessels. cally non-affected limb [13], suggesting that subclinical
• Secondary lymphedemas can be caused by surgery, lymphatic dysfunction of both limbs may be an important
radiation therapy, trauma, lymphangitis, or infection. predisposing factor. Therefore, treatment of erysipelas
should focus not only on the infection but also on the lym-
phatic aspects, including treatment for lymphedema to pre- chylothorax defined by these criteria represents a hetero-
vent recurrence of erysipelas and aggravation of the geneous group of clinical entities. In the case of an acute
pre-existing lymphatic impairment. or chronic chylothorax due to possible injury of the tho-
racic duct, accurate evaluation is crucial, as surgical liga-
tion of the thoracic duct is often entertained. In contrast, a
37.2.2 Lymphatic Malignancies cholesterol-rich effusion is typically the result of long-
standing pleurisy with elevated cholesterol levels in the
Solid tumors can originate in the lymphatic tissues, e.g., pleural space; most cases of cholesterol pleural effusions
lymphangiosarcoma, which causes primary or secondary are attributed to tuberculous or to rheumatoid pleurisy.
lymphedema. Lymphangioleiomyomatosis is characterized Distinguishing between a chylothorax and a cholesterol
by the infiltration of abnormal smooth-like cells through the effusion is critical. A chylothorax develops after injury or
pulmonary interstitium, perivascular spaces, and lymphatics obstruction of the thoracic duct, leading to leakage of
of young women leading to lymphatic disruption and final chyle into the pleural space, and is characterized by
respiratory failure. Kaposi’s sarcoma is an angiogenic tumor increased triglyceride concentration and by the presence of
of lymphatic endothelial cells. chylomicrons. In contrast, a cholesterol effusion is a long-
The lymphatic vessels also provide a route for tumor cells standing effusion associated with an elevated cholesterol
to metastasize, and the lymph node microenvironment may concentration (usually greater than 250 mg/dL) and with a
select tumor cells for increased metastatic potential. thick pleural rind; this condition represents a form of lung
entrapment.
The accumulation of chyle in the pericardial space, or
37.2.3 Intracavitary Lymphedema chylopericardium, occurs most frequently after trauma, car-
diac and thoracic surgery, or in association with tumors,
Chyle is a “milky lymph” which flows from the lacteals of tuberculosis, or lymphangiomatosis. When its precise cause
the gut through the cisterna chili and then through the tho- cannot be identified, it is called primary or idiopathic chylo-
racic duct. Chylous ascites is the accumulation of milky pericardium, a quite rare clinical condition. Chylopericardium
chyle in the peritoneal cavity. Chylous ascites has been can be caused by various clinical conditions, particularly
reported after surgeries such as abdominal aortic aneurysm trauma (thoracic duct lesions), neoplasms (primary, such as
repair, radical gastrectomy, duodenectomy, nephrectomy, lymphangioma, or through invasion of the lymphatic system
and Wilms tumor resection. The neoplastic infiltration of by other neoplasms), and filaria infection. Primary forms are
abdominal lymphatic and lymph nodal structures, traumas the result of malformation of the intestinal lymph circulation
(abdominal trauma, often iatrogenic, as in lymphadenec- and its relationship with the systemic circulation, resulting in
tomy in the surgical treatment of renal carcinoma), lym- megalymphatics that develop fistulas following even mini-
phatic dysplasia, intraperitoneal lymphatic fistula (in a mal trauma and that can be located at atypical sites in the
scenario of congenital defects of the lymphatic system), body.
rupture of a lymphatic cyst, as well as an infectious and
inflammatory process in lymph nodes account for other
causes of chylous ascites. Abdominal distention is the most Key Learning Points
common symptom. Ascitic fluid with triglyceride levels • Lymphangitis, an inflammation of the lymphatic
greater than 110 mg/dL is diagnostic of chylous ascites. collectors, becomes clinically obvious as a red
Gross appearance of the ascitic fluid correlates poorly with streak up the limb corresponding to the inflamed
absolute triglyceride levels, because turbidity also reflects vessels.
size of the chylomicrons. A high total leukocyte count with • Erysipela is a skin infection that is usually caused
marked lymphocytic predominance is also present. Total by β-hemolytic group A streptococci.
protein content varies usually from 1.4 to 6.4 g/dL, with a • Solid tumors can originate in the lymphatic tissues,
mean of 3.7 g/dL. and the lymphatic vessels also provide a route for
Chylothorax represents a form of lipid effusion charac- tumor cells to metastasize.
terized by the presence of chyle in the pleural space, which • Intracavitary lymphedema includes chylous ascites,
can be the result of obstruction or disruption of the tho- chylothorax, and chylopericardium resulting from
racic duct or one of its major tributaries. A triglyceride the accumulation of chyle in the peritoneal cavity,
concentration >110 mg/dL is virtually diagnostic, and the pleural, or pericardial spaces, respectively.
presence of chylomicrons confirms the diagnosis. However,
37.3 Workup of Patients as these techniques measure only the overall volume of the
with Lymphedema limbs, volume differences from left-right dominance, muscle
atrophy, fibrous tissue deposition, or weight gain may be
The diagnosis of lymphedema is generally clinical: edema inaccurately attributed to fluid accumulation [20].
without fovea, orange peel, skin fibrosis, and positive Stemmer
sign. Lymphedema should be considered with any edematous
extremity without pain or inflammation. Chronic venous Key Learning Points
insufficiency can be difficult to differentiate from early • The diagnosis of lymphedema is generally clinical:
lymphedema because both have pitting edema, while skin edema without fovea, orange peel, skin fibrosis, and
changes typical of late-stage lymphedema are not yet present. positive Stemmer sign.
Additionally, chronic venous insufficiency is often bilateral, • Medical history, physical examination, and diag-
rather than unilateral as in lymphedema. Lymphoscintigraphy nostic tests can help to rule out alternative causes of
may be necessary to distinguish the two conditions, although lower extremity edema.
the distinction cannot always be achieved since chronic • Exclusion of general medical causes of lower
venous insufficiency per-se can lead to secondary lymph- extremity swelling such as renal failure, protein-
edema. Similarly, deep vein thrombosis can cause a postphle- losing nephropathy, hypoalbuminemia, congestive
bitic syndrome, which can result in lipodermatosclerosis and heart failure, pulmonary hypertension, drug-
chronic swelling of the limb [14, 15]. In nonfilarial regions of induced edema, obesity, pregnancy, chronic venous
tropical Africa, Central America, and the Indian subcontinent, insufficiency, postphlebitic syndrome, filariasis,
there is a condition that has a presentation similar to filariasis, nonfilarial elephantiasis, lipedema, “armchair legs,”
called podoconiosis, or nonfilarial elephantiasis. and postoperative swelling should be considered as
Exclusion of general medical causes of lower extremity part of the differential diagnosis of lymphedema.
swelling should be a priority. These causes include, but are
not limited to, renal failure, protein-losing nephropathy,
hypoalbuminemia, congestive heart failure, pulmonary
hypertension, drug-induced edema, obesity, and pregnancy
[16]. Other considerations include lipedema (also known as 37.4 R
ole of Imaging in the Management
lipomatosis of the leg), “armchair legs” (descriptive term of Lymphedema
that results from sitting in a chair all day and night with one’s
legs in a dependent position), and postoperative swelling, a Imaging is generally used to confirm the presence and degree
dysfunction presumably due to secondary damage from of lymphatic failure, also for prognostic purposes and in the
overload or recurrent cellulitis [17]. In addition to medical perspective of choosing the best therapeutic option.
history and physical examination, diagnostic tests that help A venous Doppler ultrasound is often needed to detect
to rule out alternative causes of lower extremity edema deep venous thrombosis or venous disease, which can be
include a complete metabolic profile, serum albumin, and concomitant, as 20–30% of patients with advanced chronic
urinalysis to screen for renal failure, hypoalbuminemia, and venous disease have an associated lymphatic disease.
protein-losing enteropathy. If there is suspicion of filariasis, Bioimpedance spectroscopy (BIS), a noninvasive proce-
one can perform a blood smear (collected at night) looking dure, attempts to directly measure lymph fluid volume [21]
for the presence of microfilariae, although antigen testing by by measuring impedance to passage of an electrical current
immunochromatographic card test (Binax) or enzyme-linked in a body segment. Fat and bone act as insulators, while elec-
immunosorbent assay (TropBio) is more sensitive [18]. trolytic fluids conduct electricity, therefore allowing for the
Objective measurement of limb swelling can be problem- determination of properties unique to lymphatic fluid; in par-
atic. Documenting differences in extremity size and deter- ticular, low-frequency currents selectively pass through
mining quantitative discrepancies between patients’ extracellular fluid compartments, whereas high-frequency
unaffected and affected limbs is critical, particularly in the currents pass through both intra- and extracellular fluids
early evaluation of lymphedema. Estimating differences in [22]. Reduced impedance values indicate the presence of
limb volume has been used as an indirect measure of changes lymphedema. Other techniques described to measure limb
in lymph fluid volume over time or with treatment and is typi- edema have not been sufficiently validated or are too com-
cally done through circumferential measurements or immer- plex or expensive for routine use. Furthermore, extravascular
sion techniques to measure volume displacement [19]. fluid volume undergoes cyclic changes over days or weeks,
Disadvantages of these methods include the significant and limb volume also has a pronounced circadian variation.
amount of time necessary to perform the measurements and High-resolution cutaneous ultrasonography (20 MHz
the high potential for errors in the measurements. Additionally, probe) identifies lymphedema based on increased dermal
thickness and decreased echogenicity as compared with • Evaluation of patients with recurrent erysipelas episodes
lipedema and venous insufficiency [23]. In lymphedema (e.g., cellulitis, acute lymphangitis) without an obvious
both skin and dermal hypoechogenicity is observed, versus clinical cause, to define the presence of unknown lym-
superficial dermal hypoechogenicity which is present in case phatic pathology.
of venous insufficiency [24]. These findings reflect the accu- • Selecting patients for prophylaxis.
mulation of protein-rich exudative interstitial fluid in the • Monitoring prophylaxis.
skin and subcutaneous tissue in the case of lymphedema, • Evaluation of the effect of antibiotic treatment on altera-
whereas in venous insufficiency there is a mobile transudate tions of lymph flow.
edema that accumulates in the superficial dermis. • Evaluation of response to treatment.
Direct lymphangiography using an iodine oil agent visu- • Pre-surgical evaluation before lymphadenectomy, to iden-
alizing the major lymphatic vessels [25] is no longer rou- tify patients at risk to develop lymphedema.
tinely performed because it can lead to life-threatening • Evaluation of lymphatic rearrangement post-
complications and is difficult to perform [26]. lymphadenectomy, to determine the real need for physical
X-ray CT or MR imaging of the lower extremity can also and/or medical therapy.
detect a “honeycomb” pattern of the subcutaneous tissue
characteristic of lymphedema, although they are mostly used
to identify the morphologic changes due to the subcutaneous Key Learning Points
lipomatous hypertrophy [27–29]. • Imaging is used to confirm the presence and degree
MR lymphangiography (MRL) with intracutaneous injec- of lymphatic failure, for prognostic purposes and in
tion of an extracellular, paramagnetic contrast agent has the perspective of choosing the best therapeutic
recently been introduced for high-resolution imaging of the option.
abnormal lymphatic pathways [30, 31]. However, extravas- • Imaging that may be useful in diagnosing a lym-
cular intracutaneous injection is an off-label administration phatic disease includes venous Doppler ultrasound,
modality for MR contrast agents. bioimpedance spectroscopy, high-resolution cuta-
Whenever the clinical diagnosis of lymphedema is con- neous ultrasonography, direct lymphangiography,
troversial, it can be either confirmed or ruled out with lym- X-ray CT or MR imaging, MR lymphangiography,
phoscintigraphy, which is considered the method of choice and lymphoscintigraphy.
to evaluate the lymphatic pathways and their drainage pat- • The main clinical indications for peripheral lympho-
tern. Lymphoscintigraphy provides images of both the lym- scintigraphy are initial evaluation of any lymphatic
phatic vessels and lymph nodes, as well as quantitative data dysfunction, identification of patients with asymp-
on tracer (lymph) transport [32]. Peripheral lymphoscintig- tomatic filariasis, evaluation of patients with recur-
raphy enables to distinguish normal or altered radiocolloid rent erysipelas episodes without an obvious clinical
transport along the deep and superficial lymphatic vessels cause, selection or monitoring of patients for pro-
and tributary lymph nodes, as well as the existence of short- phylaxis, evaluation of the effect of antibiotic treat-
circuit, stenosis, bypass, or dermal backflow. Progression ment on alterations of lymph flow, evaluation of
of the radiopharmaceutical along lymphatic collectors pro- response to treatment, identification of patients at
vides a map of the active vessels draining the site of injec- risk to develop lymphedema, and evaluation of lym-
tion. Although lymphoscintigraphy does not provide an phatic rearrangement post-lymphadenectomy.
etiologic diagnosis, it can detect a latent lymphatic disease
whose clinical manifestation can be triggered by other fac-
tors (e.g., traumas, infections). In the acute phase of infec-
tion, an increased flow of lymph drainage and weak lymph
node visualization are the most common scintigraphic find- 37.5 Treatment and Management
ings. After repeated infectious episodes, the status of the of Patients with Lymphedema
lymphatic vessels is variable. Generally, patients with a
previous history of cellulite and lymphangitis present with As a chronic and potentially disabling condition, lymph-
low or no lymphatic flow as a result of the obstruction of edema is associated with significant morbidity in terms of
lymphatic collectors and lymphatic thrombosis. the functional, cosmetic, and emotional consequences.
The main clinical indications for peripheral lymphoscin- Treatment efforts are focused on minimizing the associated
tigraphy are listed here below: swelling, restoring cosmesis and functionality of the limb,
and preventing potential complications associated with
• Initial evaluation of any lymphatic dysfunction. lymphedema (e.g., cellulitis, lymphangitis). Treatments are
• Identification of patients with asymptomatic filariasis time consuming and expensive and involve a multidisci-
before onset of symptoms. plinary approach including rehabilitative lymphedema ther-
apy (elevation, exercise, compression devices, manual lymph the other physicochemical properties of particles undergoing
drainage), skin care, and surgery [33]. phagocytosis; a certain fraction of the radiocolloid remains
Drug therapy has so far been disappointing in the man- inside the lymph node, where phagocytosis occurs, while the
agement of lymphedema. While diuretic drugs are not help- remaining portion proceeds through the efferent collector
ful in lymphedema, coumarin has been found to have a toward the next lymph node station. Furthermore, also physi-
favorable effect, most likely due to reduction of vascular per- cal exercise affects the outcome of a lymphoscintigraphic
meability and thus capillary filtration. Furthermore, couma- examination.
rin is thought to activate macrophages, thus leading to Particles smaller than a few nanometers usually leak into
increased protein degradation and reduction of tissue fibro- blood capillaries, whereas larger particles (up to about
sis. However, no long-term data on treatment of lymphedema 100 nm) can enter the lymphatic capillaries and are
with coumarin are available, and potential hepatotoxicity transported to lymph nodes, where phagocytosis occurs [35,
causes some concern [34]. 36]. Particles <30 nm in size migrate rapidly; a small propor-
tion of such radiocolloid particles remains in the first lymph
node encountered, therefore resulting in the visualization of
Key Learning Points additional nodes along the same lymphatic pathway. Particles
• Lymphedema is associated with significant >100 nm in size are trapped in the interstitial compartment
morbidity. for a relatively long period [35], and the fraction of tracer
• Treatment for lymphedema aims at minimizing the that is phagocytized locally or in lymph nodes increases with
associated swelling, at restoring cosmesis and func- increasing size [37]. The smaller the molecule, the less con-
tionality of the limb, and at preventing potential vection influences its transport. Convection of dextrans of
lymphedema-related complications. similar shape and charge is significantly faster at a molecular
• A multidisciplinary approach is crucial to properly weight of 71 kDa than lighter (3 and 40 kDa) and heavier
manage and treat lymphedema patients. (2 MDa) dextrans [38]. On the other hand, albumin is trans-
ported three times slower than dextran of similar weight
(71 kDa), suggesting that both the shape and uneven charge
distribution on albumin surface influence its ability to move
37.6 Peripheral Lymphoscintigraphy through the extracellular matrix. Furthermore, negatively
charged dextrans are transported faster than neutrally charged
Peripheral lymphoscintigraphy consists of imaging the lym- dextran of similar size and shape; faster convection of nega-
phatic drainage of radiocolloids injected in the interstitium tively charged molecules through the extracellular matrix
(the hand or foot for the evaluation of the arm or leg, respec- can probably be linked to the fixed negative charge of gly-
tively) toward regional lymph nodes. Protocols for lymphos- cosaminoglycans in the extracellular matrix. In particular,
cintigraphy are not standardized, and remarkable differences repelling forces between the negatively charged molecules
among centers persist. The main differences include the type and negatively charged extracellular matrix components
of tracer administered, the site of injection, the use of might reduce mechanical interaction and thus reduce the
dynamic and/or static acquisitions, and the acquisition times. resistance against convection. Differences in surface charac-
teristics of the colloids may also account for differences in
lymph node uptake [39]. Specific surface properties, such as
37.6.1 Radiopharmaceuticals charge, hydrophobicity, and the presence of targeting ligands,
can influence both the rate of particle drainage from a subcu-
Different radiopharmaceuticals can be used to evaluate the taneous injection site and the distribution within the lym-
lymphatic system (see Chap. 2). phatic system. In rats, for instance, small, negatively charged
The transport of (macro)molecules from the interstitium liposomes localize in lymph nodes more effectively than
to the lymphatic vessels is influenced by many factors. positively charged liposomes [40, 41].
Composition of the extracellular matrix and properties of the The above considerations are the main determinants for
solute significantly affect their ability to move through the selecting suitable molecules for either peripheral lymphos-
interstitium into the lymphatics. Colloids enter and exit the cintigraphy or sentinel lymph node mapping. In fact, when
lymphatic circulation at different speeds, depending on their quantification in peripheral lymphoscintigraphy is based on
sizes. Other properties that affect the ability of colloids to assessment of tracer retention in local lymph nodes, the
move through the interstitium into the lymphatics are weight, radiopharmaceutical should present high retention in the
shape, and charge. The speed of migration through the lym- lymph nodes, i.e., have a molecular size, which promotes
phatic system is inversely related to colloid particle size, phagocytosis. Conversely, if washout techniques are used
while retention of radiocolloids in lymph nodes depends on during peripheral lymphoscintigraphy, smaller-sized tracers
that mimic in vivo transport of plasma proteins in the lym- trally through collecting ducts and then to lymph nodes [49].
phatics with faster interstitial and lymphatic transport and Therefore, investigation of the functional integrity of the
less local retention are needed to result in faster and more lymphatic system begins with the demonstration of a normal
reliable clearance data [42]. process of extracellular fluid transport into lymphatic vessels
and continues with the evaluation of the pattern of lymph
flow along the collectors into the initial lymph node(s).
37.6.2 Injected Volume and Activity The site of injection affects the pattern of lymphatic drain-
age visualized at lymphoscintigraphy. In this regard, both
The effects of varying concentrations of the particles and the subcutaneous and intradermal injections are routinely
influence of injected volume and activity parameters on the utilized for evaluating superficial lymphatics of the extremi-
outcome of lymphoscintigraphy are still unclear. Bourgeois ties. There are however discrepant views about which injec-
has investigated the effect of variable concentration (0.02 mg tion technique is best. As recommended by many investigators
versus 0.2 mg) and volume (0.2 mL versus 1.0 mL) of 99mTc- [50], subcutaneous radiocolloid injection has the advantage
nanocolloidal albumin injected subcutaneously in the foot on of negligible drainage through the blood vessels. According
lymph node uptake at 1 h postinjection. Activity in inguinal to Mostbeck and Partsch, subcutaneous injection of 99mTc-
lymph nodes was greatest using the highest quantity in the nanocolloidal albumin allowed accurate differentiation,
lowest volume [43]. using quantitative parameters, between patients with lymph-
The sentinel lymph node identification rate increased from edema and healthy volunteers [51]. On the other hand, intra-
83% to 94% with a 50% increase in injected activity [44]. dermal radiocolloid injection is preferred by other
Furthermore, because of the nonphysiologic increase in inter- investigators [52].
stitial pressure, the administration of a large volume of injec- In this regard, the optimal route of injection may vary
tate may lead to drainage in both homoregional non-sentinel depending upon the tracer employed, the subcutaneous route
lymph nodes and in additional drainage regions [45]. being optimal for the colloidal agents [51], while intradermal
injection of non-colloidal agents (e.g., 99mTc-human serum
albumin) is associated with very fast lymphatic transport,
37.6.3 Factors Affecting Kinetics which facilitates rapid evaluation and better quantification of
of Radiocolloids Injected Interstitially lymphatic flow. On the other hand, better tracer kinetics is
observed after intradermal injection of 99mTc-HSA than after
Mechanical massage over the radiocolloid injection site subcutaneous injection of the same radiopharmaceutical,
enhances the uptake and reduces the inverse correlation which results in slow or no transport [53].
between particle size and speed of lymphatic drainage. Intramuscular (subfascial) radiotracer injection is utilized
Besides the influence of particle surface properties on radio- for investigations of the deep lymphatic system of the
colloid uptake [39], an increase in venous pressure decreases extremities. “Two-compartment” lymphoscintigraphy is per-
concentration of macromolecules and leukocytes in the formed by injecting sequentially both in the epifascial (i.e.,
lymph [46]. Radiocolloid uptake is also temperature- and superficial) space and in the subfascial space. This approach
pH-dependent. is preferable for distinguishing the possible mechanisms of
Exercise increases lymph flow [47]. Type and intensity of edema in the limbs [54]. In fact, evaluating both the deep and
the exercise influence lymphatic drainage, therefore also the the superficial circulation increases the diagnostic accuracy,
outcome of lymphoscintigraphy. since it is possible to identify lymphatic abnormalities pres-
ent at either level of lymphatic circulation [55]. For both
phases of two-compartment lymphoscintigraphy, the radio-
37.6.4 Injection Sites and Method colloid is injected using a 25 G, 15-mm-long needle, admin-
istering a small volume of the radiocolloid suspension
Lymphoscintigraphy is based on the interstitial injection of a (0.2–0.3 mL) containing an activity of about 11–18 MBq.
suitable radiopharmaceutical that enters the initial lymphat- The procedure adopted in our center is described as fol-
ics both through interendothelial openings and by vesicular lows. For the deep lymphatic circulation of the lower extrem-
transport through the endothelial cells [48]. Interstitial radio- ities, we inject two aliquots of radiocolloid (7 MBq each, in
pharmaceutical administration is based on the lymphatic cir- a volume of 0.1 mL) in the first and second inter-metatarsal
culation anatomy, considering that lymphatic vessels space (Fig. 37.3) on each side. After identifying correct posi-
originate in the connective interstitium nearby the blood ves- tioning by palpating the distal heads of the metatarsal bones,
sels. The initial lymphatics are closely interconnected in a the needle is inserted by about 12–13 mm to reach the inter-
hexagonal pattern, through a set of precollectors, with deeper metatarsal muscles below the deep fascia plantaris (Fig. 37.4).
lymphatics in the dermis, where lymph is transported cen- For the deep lymphatic circulation of the upper extremities,
Superficial and Intermetatarsal the radiocolloid (similar volume and activity as for the lower
deep fascia space extremities) is injected in the second and in the third inter-
dorsalis
metacarpal space (identified by palpating the palms of both
hands the fossa in the inter-metacarpal space immediately
Metatarsal I proximal to the distal heads of the metacarpal bones; see
Fig. 37.5) on each side, inserting the needle by about
Metatarsal V Fascia 10–12 mm to reach the inter-metacarpal muscles below the
plantaris
deep fascia palmaris (Fig. 37.6).
For superficial lymphoscintigraphy, we prepare syringes
in a similar manner as for deep lymphoscintigraphy but with
slightly higher radioactivity content (about 15–18 MBq). We
inject two aliquots on the dorsum of either each foot or each
Aponeurosis
Deep fascia plantaris hand (for the lower or the upper extremities, respectively),
plantaris inserting the needle subdermally in sites corresponding
approximately to the prior palmar injections, about 1–2 cm
Fig. 37.3 Schematic representation of the modality of radiocolloid
proximally to the interdigital web (Fig. 37.7). We strongly
injection for evaluating the deep lymphatic circulation of lower limbs.
Each of two aliquots is injected in the first and in the second inter- recommend against injecting the radiocolloid directly into
metatarsal space, respectively. Correct localization is achieved by palpat- the interdigital web (as generally indicated by other authors),
ing the soles of both feet, immediately proximal to the distal heads of the since this procedure may result in the visualization of either
metatarsal bones (reproduced with permission from: Erba PA, Sollini M,
the superficial and/or the deep lymphatic systems.
D’Errico G, Mariani G. Methodological aspects of lymphoscintigraphy:
bicompartmental versus monocompartmental radiocolloid administra- Due to the faster and more complex pattern of the super-
tion. In: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, ficial lymphatic circulation, we prefer to perform full assess-
Eds. Atlas of Lymphoscintigraphy and Sentinel Node Mapping – A ment of the deep lymphatic system first, followed by
Pictorial Case-Based Approach. Milan: Springer; 2013:27–38)
a b
Fig. 37.4 Radiocolloid injection in the first (a) and second (b) inter- mental versus monocompartmental radiocolloid administration. In:
metatarsal space. The needle is inserted by about 12–13 mm, so to reach Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, Eds.
the inter-metatarsal muscles below the deep fascia plantaris (repro- Atlas of Lymphoscintigraphy and Sentinel Node Mapping – A Pictorial
duced modified with permission from: Erba PA, Sollini M, D’Errico G, Case-Based Approach. Milan: Springer; 2013:27–38)
Mariani G. Methodological aspects of lymphoscintigraphy: bicompart-
superficial lymphoscintigraphy as the final step of the com- 37.6.5 Imaging

bined procedure. The injection sites are prepared by swab-
bing the area with either an iodine solution (especially in Images should be recorded with a dual-detector gamma cam-
patients with frank lymphedema) or alcohol. Both limbs are era, using high-resolution parallel-hole collimators (setting a
always injected, using one side as a control for patients with 20% window on the 140 keV energy peak of 99mTc), both in
unilateral lymphedema. the spot view mode and in the whole-body mode. Spot views
can be acquired from the feet to pelvis (for lower limbs) or
Deep fascia from the hands to axilla (including the chest, for upper limbs)
palmaris
Fascia Aponeurosis for about 3–5 min, starting from the most distal to the most
palmaris palmaris proximal portion of the limbs. Final spot views of the abdo-
manus
men should also be acquired, to confirm passage of the radio-
colloid to the systemic blood circulation within a
physiological time window (as demonstrated by visualiza-
tion of the liver and spleen), repeating the acquisitions up
Metacarpal I Metacarpal V until 4–6 h post-administration in case of delayed radiocol-
loid drainage. After acquiring the spot images, a whole-body
scan can be useful, from the distal feet until the abdomen for
lower limbs and from the hands to the chest and upper abdo-
Superficial and men for upper limbs, using a scan speed of 12 cm/min.
deep fascia Intermetacarpal Dynamic imaging is necessary if quantitation of lym-
dorsalis manus spaces
phatic flow is planned (see below). The improvement of
Fig. 37.5 Schematic representation of the modality of radiocolloid
anatomo-topographic mapping of sentinel lymph nodes for
injection for evaluating the deep lymphatic circulation of upper extrem- radioguided surgery (see Chap. 16) has raised interest in the
ities. Each of two aliquots is injected in the first and in the second inter- application of hybrid SPECT/CT imaging also for peripheral
metacarpal space, respectively. Correct localization is achieved by lymphoscintigraphy. In this regard, SPECT/CT is generally
palpating the palms of hands, immediately proximal to the distal heads
of the metacarpal bones that are identified by palpating the palms of
not acquired, although a recent report emphasizes the clini-
both hands, immediately proximal to the distal heads of the metacarpal cal impact of this imaging approach for assessing the effi-
bones (reproduced with permission from: Erba PA, Sollini M, D’Errico cacy of lymphovenous anastomosis patients with
G, Mariani G. Methodological aspects of lymphoscintigraphy: bicom- lymphedema of the lower limbs [56].
partmental versus monocompartmental radiocolloid administration.
In: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA,
For intracavitary lymph effusions, two-phase lymphos-
Eds. Atlas of Lymphoscintigraphy and Sentinel Node Mapping – A cintigraphy is generally performed. Subcutaneous injection
Pictorial Case-Based Approach. Milan: Springer; 2013:27–38) in the interdigital space of both feet is preferred since the
a b
Fig. 37.6 Radiocolloid injection in the second (a) and third (b) inter- versus monocompartmental radiocolloid administration. In: Mariani
metacarpal space. The needle is inserted by about 10–12 mm, so to G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of
reach the inter-metacarpal muscles below the deep fascia palmaris Lymphoscintigraphy and Sentinel Node Mapping – A Pictorial Case-
(modified from: Erba PA, Sollini M, D’Errico G, Mariani Based Approach. Milan: Springer; 2013:27–38)
G. Methodological aspects of lymphoscintigraphy: bicompartmental
a b
Fig. 37.7 For superficial lymphoscintigraphy, we prepare syringes in a sion from: Erba PA, Sollini M, D’Errico G, Mariani G. Methodological
similar manner as for deep lymphoscintigraphy but with slightly greater aspects of lymphoscintigraphy: bicompartmental versus monocompart-
activity (about 15–18 MBq). The two aliquots are injected either on the mental radiocolloid administration. In: Mariani G, Manca G, Orsini F,
dorsum of each foot (a) or hand (b), inserting the needle subdermally in Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of Lymphoscintigraphy and
sites corresponding approximately to the previous deep injections, Sentinel Node Mapping – A Pictorial Case-Based Approach. Milan:
about 1–2 cm proximal to the interdigital web (reproduced with permis- Springer; 2013:27–38)
superficial circulation of the lower limbs accounts for the 37.6.6 Visual Interpretation
majority of lymph transport. In preparation for the scan
before radiocolloid injection, if a surgical drainage is pres- Qualitative lymphoscintigraphy is in many cases sufficient to
ent it should be closed. The gamma camera is generally establish a reliable diagnosis [66]. However, there is still lack
positioned over the site of the effusion, and dynamic images of consensus on the criteria to visually interpret a normal
are acquired from the time of radiocolloid injection until scan, and expertise plays a critical role in this setting [67]. A
the evidence of radioactivity accumulation. Usually images typical example of normal two-compartment lymphoscinti-
are acquired for up to 8 h with timings of about 30 min. graphic pattern is shown in Figs. 37.8 and 37.9. Abnormal
During the second phase, delayed imaging is acquired con- patterns include asymmetrical visualization of lymphatic
sisting in the acquisition of the region of interest in the channels and collateral lymphatic channels, interrupted lym-
same conditions of previous phase but with the surgical phatic vessels and lymph collection, asymmetrical or absent
drainage open. Although a certain fraction of the radiocol- visualization of regional lymph nodes, and the presence of
loid may remain trapped in the inguinal lymph nodes, the “dermal flow” and/or “dermal backflow” [54]. Figs. 37.10,
fraction progressing through lymph nodes is sufficient to 37.11, 37.12, and 37.13 show different lymphoscintigraphic
reach into the thoracic duct and thus demonstrates a possi- patterns in patients with lymphedema.
ble lymphatic leak. For intracavitary lymph effusion (Fig. 37.14), the evalu-
Static images followed by SPECT, or better SPECT/CT, ation of differences in intensity of the activity present in
acquisitions may complete the set of images, depending on all the images at the site of interest defines (1) a normal
the site of the intracavitary effusion. SPECT/CT imaging is pattern, when no significant differences in the distribution
particularly useful in patients with intracavitary chylous of radiocolloids on all images until the 24th hour are
effusions, as suggested by occasional case reports in case of detected, and (2) an abnormal test, when focal radiocolloid
chylothorax or chylopericardium [57–62] and confirmed by uptake increases in all the acquisitions until the end of the
more systematic studies that have confirmed the clinical first acquisition phase but disappears in the last acquisition
impact of this imaging approach in the identification of the at 24 h.
site of leakage causing either chylothorax or chylopericar-
dium [63–65].
Scintigraphic acquisitions should be displayed with the 37.6.7 Lymphoscintigraphy with Stress Test
intensity minimized, to saturate the display of residual activ-
ity at the injection site and depict the small fraction of radio- Lymphoscintigraphy can be performed following an inter-
colloid that migrates from the injection site to the more vention intended to augment lymphatic flow—such as
proximal lymphatic stations. changes in temperature, physical exertion, or administration
Fig. 37.8 Normal
lymphoscintigraphic pattern a b
of lower limbs following
two-compartment radiocolloid
injection (anterior views). (a)
Deep injection visualizes
symmetrical migration of the
radiocolloid along the vessels
of the deep lymphatic
circulation, clearly identifying
both the popliteal and the
inguinal lymph nodes. (b) In
addition to the deep lymphatic
vessels (still visualized by
prior deep injection),
subsequent subdermal
radiocolloid injection
visualizes the more variable
superficial lymphatic
circulation. This image
represents therefore the sum
of the two lymphatic systems,
deep and superficial. Deep
and superficial lymphatic
vessels merge at the groins
into the inguinal lymph nodes
and continue in the main
pelvic and abdominal
lymphatic vessels (reproduced
with permission from: Erba
PA, Sollini M, D’Errico G,
Mariani G. Methodological
aspects of
lymphoscintigraphy:
bicompartmental versus
monocompartmental
radiocolloid administration.
In: Mariani G, Manca G,
Orsini F, Vidal-Sicart S,
Valdés Olmos RA, Eds. Atlas
of Lymphoscintigraphy and
Sentinel Node Mapping – A
Pictorial Case-Based
Approach. Milan: Springer;
2013:27–38)
of a pharmacologic agent. Although stress lymphoscintigra- min for 20 min), 30 fist clenchings, 100 submaximal contrac-
phy is recommended for enhanced sensitivity and for quanti- tions in 10 min, bouts of arm cranking for 5 min at 0.6 W/kg
tation of lymphatic flow [68], this approach is not universally or 75 contractions in 2.5 min at 50% maximum voluntary
employed [69]. In the lower extremities, stress tests include contraction, and 12 repeated sets of arm cranking for 2.5 min
walking [70], standing [71], limb massage [52], standardized at 0.6 W/kg or 12 repeated sets of arm cranking for 2.5 min
treadmill exercise [72], and bicycle ergometer exercise [73]. at 0.3 W/kg.
For the upper extremities, repetitive squeezing of a rubber
ball, use of a handgrip exercise device [73], or massage [52]
have been proposed. Massage, exercise, and standing each 37.6.8 Quantitative Lymphoscintigraphy
enhance radiotracer drainage from the injection site in the
lower extremities [68]. More standardized stress tests pro- Quantitation of lymphatic flow with lymphoscintigraphy has
posed by different investigators include, among others, pas- been proposed by many authors to enhance sensitivity in the
sive electric foot ergometer at 30 cycles/min for 2 h, climbing identification of lymphatic drainage abnormalities. Different
150 steps, horizontal treadmill at 3.2 km/h, ergometric bicy- procedures may be applied, and different semiquantitative
cle at 25–75 W, flexed and straightened feet (20 movements/ and/or quantitative indices may be derived, as follows:
(b) Transit time (TT): it is the time over which the radiocol-
loids reach the inguinal nodes. Variable time is necessary
to obtain adequate quality images, since large-sized par-
ticles undergo slow drainage from the injection site to
lymph nodes. Use of filtered 99mTc-sulfur colloids or of
99m
Tc-nanocolloidal albumin overcomes this problem, as
it shortens the acquisition time and improves image
interpretation [74]. TT identifies lymphatic dysfunction
and correlates with its severity; however, a certain vari-
ability in TT estimation has been reported, suggesting
that quantitative parameters should be interpreted in
association with visual interpretation of images for
improved diagnostic accuracy [75].
(c) Tracer appearance time (TAT): it is calculated as the
time for drainage of the radiotracer to locoregional
lymph nodes (normal value <10 min) [71].
(d) Transport capacity (TC): it calculates the clearance rate
from the injection site as the percent amount of activity
transported from the depots to the groin lymph nodes in the
first 2 h after injection. The normal reference limit is 15%.
This parameter identifies early impairment of lymphatic
drainage, even before the appearance of clinical or qualita-
tive changes in the lymphoscintigraphic pattern [13].
(e) Removal rate constant (RRC): it represents local lymph flow
per unit distribution volume of the flow tracer and is reduced
(by about 25%) in the presence of lymphatic dysfunction [76].
(f) Depot activity transported to inguinal lymph nodes after
attenuation correction is a modified method to determine
the depot clearance rate which takes into account the
individual variation in tissue depth in order to improve
Fig. 37.9 Normal lymphoscintigraphic pattern of upper limbs follow-
ing two-compartment radiocolloid injection (the two hands are joined
quantification of lymph nodal activity and to make this
together in the lower part of the imaging field). Upper panel: deep injec- quantitative parameter more reliable [77].
tion visualizes symmetrical migration of the radiocolloid along the ves- (g) Uptake index of the left inguinal (UIL) and of the right
sels of the deep lymphatic circulation, clearly identifying axillary inguinal lymph nodes (UIR): it provides the activity ratio
lymph nodes (arrows). Lower panel: in addition to the deep lymphatic
vessels (still visualized by prior deep injection), subsequent subdermal
for inguinal nodes calculated on total body images at 2 h
radiocolloid injection visualizes the more variable superficial lymphatic after radiocolloid injection. Radioactivity at the injection
circulation. Image in the lower panel represents therefore the sum of the site is measured immediately after administration to cal-
two lymphatic systems, deep and superficial. Epitrochlear lymph nodes culate the percentage of radiocolloid injected into the left/
are clearly visualized following superficial injection of the radiocolloid,
associated with enhanced activity accumulation in the axillary lymph
right foot (WpL/WpR) that migrates to inguinal lymph
nodes (reproduced with permission from: Erba PA, Sollini M, D’Errico nodes. Therefore, UIL/UIR is calculated as percentage of
G, Mariani G. Methodological aspects of lymphoscintigraphy: bicom- count rate in the lymph nodes of left/right inguinal region
partmental versus monocompartmental radiocolloid administration. divided by the count rate over the total body × WpL/WpR
In: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés Olmos RA,
Eds. Atlas of Lymphoscintigraphy and Sentinel Node Mapping – A
[75]. This parameter identifies lymphatic dysfunction;
Pictorial Case-Based Approach. Milan: Springer; 2013:27–38) however, because of some variability in UIL/UIR deter-
mination, quantitative parameters should be interpreted in
(a) Transport index (TI) [73]: it consists of a numerical association with visual interpretation of the images in
index of transport kinetics ranging from 0 (normal) to 45 order to improve diagnostic accuracy.
(pathological), obtained by combining visual assessment (h) The washout rate constant (k) and depot half-life

of five criteria: spatial and temporal distribution of the (T1/2 = ln2/k) provide a quantitative measure of lymphatic
radionuclide, appearance time of lymph nodes, and function calculated by linear regression analysis of the
graded visualization of lymph nodes and vessels. TI in a activity/time curve derived from images obtained over
healthy extremity should be <10. Changes in these 20–25 min starting 40 min after the injection of 10 MBq
parameters after treatment are significantly correlated of 99mTc-HSA (followed by 30 min of rest + 10 min of
with volume changes of the extremities in response to ergometer bicycle at 75 W) [51]. However, these param-
treatment. eters are heavily influenced by the injection technique,
Fig. 37.10 Lymphoscintigraphy in a woman with left lower limb (spot views in center panels) results in visualization of one single left
edema; Doppler ultrasound was normal, and there was no family his- inguinal lymph node but still with no clear visualization of lymphatic
tory of lymphedema. Spot images visualize clear impairment of lymph vessels on this side (reproduced with permission from: Erba PA, Sollini
flow of the left lower limb, while the right panel better displays the M, Boni R. Lymphoscintigraphy for the differential diagnosis of periph-
overall pattern as depicted by the whole-body scan. Minimal dermal eral edema and intracavitary lymph effusion. In: Mariani G, Manca G,
flow is evident at the distal medial portion of the left leg after deep Orsini F, Vidal-Sicart S, Valdés Olmos RA, Eds. Atlas of
radiocolloid injection (left panels), and no popliteal or inguinal lymph Lymphoscintigraphy and Sentinel Node Mapping – A Pictorial Case-
nodes are visualized on the left side. Superficial radiocolloid injection Based Approach. Milan: Springer; 2013:39–86)
thus yielding quantitative results that are difficult to stan- Visual interpretation of clearance from the injection site
dardize for assessing lymphatic dysfunction [78]. may not allow discrimination between normal subjects and
(i) Liver to nodal ratio (L/N ratio): it is calculated based on patients with lymphedema [54]. Quantitation of disappear-
regions of interest (ROI) defined on the 45- and 150-min ance rates from the injection site is preferred when using
images (anterior view) around the right liver lobe and on labelled human serum albumin. When comparing quantitative
ilioinguinal groups of nodes bilaterally corrected for back- and qualitative lymphoscintigraphy in 308 extremities with
ground (ROIs on the same horizontal levels as the liver and different grades of lymphedema, qualitative interpretation
the ilioinguinal nodes lateral to the right edge, respectively). confirmed the diagnosis of lymphedema in 70% of the
Normal values are of 1.8 pixel−1 × 10−6 (range 1.3–5.5) at extremities, while quantitative analysis detected abnormal
45 min and 2.5 pixel−1 × 10−6 (range 1.5–5) at 150 min. The lymphatic function in all 308 examined limbs. All cases
150 min L/N ratio is most frequently found to be abnormal missed by qualitative interpretation were mild, grade I lymph-
in patients with lymphatic dysfunction, correlated to sever- edema [77].
ity of lymphoscintigraphic abnormalities [79].
a b c
Fig. 37.11 (a) Post-traumatic stage V lymphedema of the lower left leg and up to the mid-thigh), although with preservation of the main
limb with cutaneous retraction of the proximal and medial portion leg, lymphatic vessels. (c) The whole-body scans acquired after deep radio-
hyperkeratosis, lymphatic vesicles, eczema, and ulcerations in a colloid injection (left panel) and after superficial injection (right panel)
36-year-old man. After a crush injury the patient had multiple surgical better depict the overall lymphoscintigraphic pattern (reproduced with
procedures; grade 3 disability according to the Ricci scale was present. permission from: Erba PA, Sollini M, Boni R. Lymphoscintigraphy for
(b) Spot images acquired after deep radiocolloid injection (left panels) the differential diagnosis of peripheral edema and intracavitary lymph
show near-normal deep lymphatic circulation in both limbs, with physi- effusion. In: Mariani G, Manca G, Orsini F, Vidal-Sicart S, Valdés
ologic visualization of popliteal and inguinal lymph nodes on both Olmos RA, Eds. Atlas of Lymphoscintigraphy and Sentinel Node
sides. Whereas, superficial radiocolloid injection (spot images in right Mapping – A Pictorial Case-Based Approach. Milan: Springer;
panels) visualizes severe dermal backflow on the left side (involving the 2013:39–86)
37.6.9 Absorbed Dose

over the radiocolloid injection site, exercise, venous
After the interstitial administration of about 37 MBq Tc- 99m pressure, temperature, and pH.
HAS, the absorbed dose is about 4.44 mGy at the site of • The site of injection affects the pattern of lymphatic
injection, 0.27 mGy in the lymph nodes, and <0.01 mGy in drainage visualized by lymphoscintigraphy.
the liver, spleen, bone marrow, and ovaries. • Subcutaneous and intradermal injections are rou-
tinely utilized for evaluating superficial lymphatics
of the extremities. Intramuscular (subfascial) radio-
Key Learning Points tracer injection is utilized for investigations of the
• Peripheral lymphoscintigraphy consists of imaging deep lymphatic system of the extremities.
the lymphatic drainage toward regional lymph • “Two-compartment” lymphoscintigraphy is per-
nodes of radiocolloids injected in the interstitium. formed by injecting sequentially both in the epifas-
• Different radiopharmaceuticals can be used to eval- cial space and in the subfascial space. Due to the
uate the lymphatic system. faster and more complex pattern of the superficial
• The rate of particle drainage from a site of intersti- lymphatic circulation, it is preferred to perform full
tial injection and distribution within the lymphatic assessment of the deep lymphatic system first, fol-
system can be influenced by several factors that lowed by superficial lymphoscintigraphy as the
should be considered when selecting suitable mol- final step of the combined procedure.
ecules for either peripheral lymphoscintigraphy or • Both limbs are always injected, using one side as a
for sentinel lymph node mapping. control for patients with unilateral lymphedema.
• The effects of varying concentrations of the parti- • Lymphoscintigraphy should include spot images
cles and the influence of injected volume and activ- and whole-body scan. Dynamic imaging is neces-
ity parameters on the outcome of lymphoscintigraphy sary if quantitation of lymphatic flow is planned.
are still unclear. • SPECT/CT is generally not acquired, although a
• Factors affecting the kinetics of radiocolloids recent report emphasizes the clinical impact of this
injected interstitially include mechanical massage imaging approach for assessing the efficacy of lym-
Fig. 37.12
Lymphoscintigraphy in a
33-year-old man with right
pelvic (adjacent to the urinary
bladder) and right inguinal
swelling. CT had revealed
multiple lesions suspected for
cystic lymphangiomatosis
localized in the splenic lodge,
in bones (ribs, vertebral
bodies, pelvic bones), and in
the pelvis. [18F]FDG PET/CT
was negative for lesions with
increased glucose
metabolism. Composite
whole-body images following
deep radiocolloid injection in
the lower extremities (left
panel) display mild delay of
lymph drainage on the left
side. Abnormal radiocolloid
accumulation is seen in the
right pelvis, adjacent to the
urinary bladder,
corres

Nuclear Medicine Textbook: Editors

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nuclear Medicine Textbook: Editors

Uploaded by

Copyright:

Available Formats

Duccio Volterrani

Paola Anna Erba

Nuclear Medicine Textbook

Ignasi Carrió H. William Strauss

ISBN 978-3-319-95563-6 ISBN 978-3-319-95564-3 (eBook)

© Springer Nature Switzerland AG 2019

Pisa, Italy Duccio Volterrani

Part I Basic Science

1 Fundamentals of Natural and Artificial Radioactivity

15 Essentials of MR Image Interpretation������������������������������������������������������������������� 317

Part II Organ/Apparatus-Specific Applications

17 Radionuclide Imaging for Non-tumor Diseases of the Brain������������������������������� 391

Part III Practice and Procedures (with Supplemental Online Files)

39 Radiopharmacy/Radiochemistry for Conventional Single-Photon

47 Teaching Cases in Nuclear Medicine: Non-oncological Applications������������������� 1139

Gayane Aghakhanyan Regional Center of Nuclear Medicine, Department of Translational

Davide Caramella Diagnostic and Interventional Radiology, Department of Translational

Montserrat Estorch Nuclear Medicine Department, Sant Pau Hospital, Autonomous

Valerio Lanni Nuclear Medicine Unit, Fondazione Policlinico Universitario A. Gemelli

Roberto Pani Department of Medical and Surgical Sciences and Biotechnologies, Sapienza

Flavia Ticconi Nuclear Medicine, Department of Health Sciences, University of Genoa,

Alberto Del Guerra and Daniele Panetta

Learning Objectives • To acquire the knowledge of the various types of decays

© Springer Nature Switzerland AG 2019 3

1.1  rief History of Radiation

X-rays were discovered by Wilhelm Röentgen on Christmas

1.3 Radioactive Decays  = 0,1, ¼, n - 1  (1.8)

Frequency (Hz) 104 108 1012 1015 1016 1018 1020

In the previous equation, N0 represents the number of A ( t ) = -dN / dt = - N 0 ( -l ) exp ( -l t ) = l N ( t ) = N ( t ) / t

0.3 Key Learning Points

232Th 228Th 234Th 230Th 231Th 227Th

220Rn 222Rn 219Rn

218At 219At 215At

216Po 212Po 218Po 214Po 210Po 215Po 211Po

212Pb 208Pb 214Pb 210Pb 206Pb 211Pb 207Pb

Thorium family Uranium family Actinium family

Fig. 1.5 Basic scheme of a a

the reaction 98Mo(n,γ)99Mo [8]. Due to the relatively small

Key Learning Points

1.6 I nteraction of Charged Particles

When the particles of a radiation field hit a target material

µ 2 , for b < 0.96 (1.29a) 13

where Zp is the atomic number of the primary particle, Z is

Protons / ions Electrons / positrons

Rayleigh scattering Photoelectric absorption

Compton scattering Pair production

a dx neutrons is particularly high for Boron-10. For instance,

x 1. ICRU. Report 85. Fundamental quantities and units for ionizing

Key Learning Points

F. Orsini (*) · E. Puta

© Springer Nature Switzerland AG 2019 21

2.9 Other Radiopharmaceuticals 54

Learning Objectives Radiopharmaceuticals consisting of only the radionuclide

active transport requires energy (in the form of ATP or an 99m

Fig. 2.3 Axial SPECT

resin, which constitutes the stationary phase of the system usu-

geometric conformation, the transition metal remaining in 2.3.1 99mTc-Sodium Pertechnetate

Key Learning Points

Key Learning Points HO P O P OH

anion pumps (Na+/I− symporter). OH OH

uptake, as observed in the case of reflex sympathetic dystro-

mulate at sites of bone mineral rearrangement by 99m

Introduced in clinical practice in 1970 [4], 99mTc-DTPA is

Pisa, Italy Duccio Volterrani

Part I Basic Science

1 Fundamentals of Natural and Artificial Radioactivity

15 Essentials of MR Image Interpretation�� 317

Part II Organ/Apparatus-Specific Applications

17 Radionuclide Imaging for Non-tumor Diseases of the Brain�� 391

Part III Practice and Procedures (with Supplemental Online Files)

39 Radiopharmacy/Radiochemistry for Conventional Single-Photon

47 Teaching Cases in Nuclear Medicine: Non-oncological Applications�� 1139

1.1 rief History of Radiation

1.3 Radioactive Decays  = 0,1, ¼, n - 1 (1.8)

1.6 I nteraction of Charged Particles

2.9 Other Radiopharmaceuticals 54

geometric conformation, the transition metal remaining in 2.3.1 99mTc-Sodium Pertechnetate

2.3.7 99mTc-Mannosyl-DTPA-Dextran 2.3.8 99mTc-Macroaggregated Albumin

2.3.11 99mTc-Hexakis-2-Methoxy-2-Isobutyl- Localization of 99mTc-sestamibi in the parathyroid tissue

2.6 201Tl-Chloride quality of gamma camera imaging with 201Tl is suboptimal

remainder remains airborne and is exhaled. The pattern of 2.8.3 Xenon-133

3.1 ection I: Positron-Emitting

fore can be applied without interference or competition by 3.1.7 11C-Labeled Radiopharmaceuticals

3.1.9.2 Radiopharmaceuticals Labeled with Very

3.2.4 Substrates for Phospholipid Synthesis a H3C 11CH