Textbook of Therapeutic Cortical Stimulation

TEXTBOOK OF THERAPEUTIC CORTICAL STIMULATION
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.
TEXTBOOK OF THERAPEUTIC CORTICAL STIMULATION
SERGIO CANAVERO
EDITOR
Nova Biomedical Books

New York
Copyright 2009 by Nova Science Publishers, Inc.

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Textbook of therapeutic cortical stimulation / [edited by] Sergio Canavero. p. ; cm. Includes bibliographical references and index. ISBN 978-1-61122-495-5 (Ebook)
Published by Nova Science Publishers, Inc. New York
To Marco and Serena Who dares wins!
CONTENTS
Preface Basic Aspects Chapter 1 Anatomical Methods of Cortical Localization Michael C. Park, Marc A. Goldman, Gerhard M. Friehs and Sergio Canavero From Localization to Surgical Implantation Youichi Saitoh and Koichi Hosomi Principles of Surgical Implantation and Complication Avoidance Benoit JM Pirotte, Philippe Voordecker, Marc Levivier and Danielle Baleriaux Transcranial Magnetic and Direct Current Stimulation: A Primer Filippo Cogiamanian, Sara Marceglia, Lorenzo Rossi, Ernesto Della Torre and Alberto Priori Mechanisms of Action Vincenzo Di Lazzaro, Paolo Profice, Fabio Pilato, Michele Dileone, Federico Ranieri, Riccardo Di Iorio and Pietro A. Tonali Optimization of Parameter Selection and Electrical Targeting Amorn Wongsarnpigoon and Warren M. Grill A Neuroengineering Model of Extradural Cortical Stimulation Ljubomir Manola and Jan Holsheimer
xi 1 3
Chapter 2 Chapter 3
17 33
Chapter 4
45
Chapter 5
57
Chapter 6 Chapter 7
69 91
viii Clinical Conditions Chronic Pain Chapter 8 Chapter 9 Chapter 10
Contents 113 115 117 139
Noninvasive Stimulation for Chronic Pain Sergio Canavero, Massa Micon Barbara and GianLuca Zollino Extradural Cortical Stimulation for Central Pain J. E. Arle, J. L. Shils and S. Canavero Extradural Cortical Stimulation for Peripheral (Including Trigeminal) Neuropathic Pain D. Rasche and V.M. Tronnier Neuroimaging and Neurophysiological Studies of ECS for Chronic Pain Sergio Canavero, J. E. Arle and J. L. Shils Movement Disorders
161
Chapter 11
169 181 183
Chapter 12 Chapter 13
Noninvasive Stimulation for Treatment of Movement Disorders Eman M. Khedr A. Invasive Cortical Stimulation for Parkinsons Disease and Movement Disorders B. Cioni, A. R. Bentivoglio, C. De Simone, A. Fasano, C. Piano, D. Policicchio, V. Perotti, M. Meglio B. Invasive Cortical Stimulation for Parkinsons Disease (PD):Why, Where and How Sergio Canavero Stroke Rehabilitation
201
217 229 231
Chapter 14
Noninvasive Stimulation for Stroke Rehabilitation Paolo Profice, Michele Dileone, Fabio Pilato, Federico Ranieri, Fioravante Capone, Lucia Florio, Pietro A. Tonali and Vincenzo Di Lazzaro Extradural Cortical Stimulation for Stroke Recovery A: Review of Clinical Studies Jeffrey A. Brown and Hyoung-Ihl Kim B: Surgical Technique Hyoung-Ihl Kim, Yong-Il Shin and Sung-Keun Moon Coma Rehabilitation
Chapter 15
249 257 273 275
Chapter 16
Extradural Bifocal Cortical Stimulation for the Post-Traumatic Permanent Vegetative State Sergio Canavero, Barbara Massa-Micon, Franco Cauda and Federico DAgata
Contents Psychiatric Disorders Chapter 17 Invasive and Non-Invasive stimulation for Psychiatric Conditions P. Eichhammer, B. Langguth, M. Landgrebe, S. Canavero and G. Hajak Epilepsy Chapter 18 Chapter 19 Cortical Stimulation for Medically Refractory Epilepsy David J. Mogul, Sergio Canavero and Ananda S. Fine Closed-loop Stimulation for Control of Focal Epilepsy J. Smith, K. Fountas, M. Murro, Y. Park, P. Jenkins, M. Morrell, E. Esteller and S. Canavero Tinnitus Chapter 20 Chapter 21 TMS for Refractory Tinnitus B. Langguth, D. de Ridder, P. Eichhammer and G. Hajak Auditory Cortex Stimulation for Intractable Tinnitus Dirk De Ridder, Berthold Langguth, Mark Plazier, Elsa van der Loo,Tim Vancamp, Jan Ost, Olivier Meeus, Stefan Sunaert, Silvia Kovacs, Tomas Menovsky, Paul Van de Heyning Prosthetics Chapter 22 Cortical Visual Neuroprosthesis E. Fernndez, F. Pelayo, S. Romero, J.M. Ferrandez, C. Botella, J. Albisua and R. A. Normann
ix 287 289
307 309 319
335 337 353
363 365
Conclusion Chapter 23 Cortical Stimulation Versus Deep Brain Stimulation in Neurological and Psychiatric Disorders: Current State and Future Prospects Damianos E. Sakas and Ioannis G. Panourias
389
391 423
Index
Upon a slight conjecture I have ventured on a dangerous journey, And I already behold the foothills of new landsThose who have the courage to continue the search Will set foot upon them. Immanuel Kant
PREFACE
Over the past 20 years a silent revolution in functional neurosurgery has been unfolding. This revolution -which is still in the making- is the development of cortical electrical neuromodulation (see box for select historical milestones). Unlike deep brain stimulation which began in the 1950s following the advent of stereotactic technology and bloomed over the past 20 years, CS elbowed its way forward all too slowly. The reason is simple. The cortex, with its vast expanse and deceptively simple structure, baffled scientists for the past 150 years or so. Only in the past 25 years, the introduction of in vivo functional neuroimaging triggered an explosion of knowledge on cortical anatomy and function. As a result, cortical targets for therapeutic modulation emerged slowly. Certainly, the fact that each persons brain is wired differently, so that parameters and/or target areas of stimulation vary from person to person, is compounded by the extent of the cortex, as compared to small deep nuclei. No book covering the field of therapeutic cortical stimulation exists and only scarce attention has been given to this new field of neuromodulation in all textbooks and related books. No wonder then that so many neurologists and neurosurgeons the majority, I dare say- are unaware of this technology. Yet, even those who ventured into this field often ignore the full range of current and potential applications, or other colleagues contributions, sacrificing study time to menial surgery. Others jumped on the bandwagon, without proper training, and, given the free-for-all attitude governing the field, pursue approaches which can be dangerous, such as subdural implantation of electrodes, a practice that should be restricted to very few patients. In a word, the field is like the Far West of yore. The aim of this book is thus twofold: cross-fertilization among neurologists, other health professionals and functional neurosurgeons, and defining indications and limits of cortical stimulation, invasive and noninvasive (or implantable and nonimplantable, as suggested by Damianos Sakas). The absence of large-scale commercial backing has made research in this area relatively slow, but the hope for drug-free treatments has sustained the field. Much opposition came (and still comes) from the many functional neurosurgeons that pursue deep brain stimulation and, given some undisputed successes, have grown blind to other alternatives. The fact that cortical targets are still not completely defined for any indication calling for burdensome individualization- diminishes the appeal of the technique. However, the recent interest from the device industry might change the status quo.
xii
Sergio Canavero
Box. Milestones of Cortical Stimulation

(with David J Mogul and Ananda S. Fine)
---Antiquity: Scribonius Largus in his De Compositionibus Medicamentorum mentions the use of electrical currents to treat headaches and pain through the application of electric torpedo fish to the affected region or through placement of painful extremities into a pool of water containing torpedo fish. ---1831: Faradays law of electromagnetism. ---1870: Gustav Fritsch and Hitzig published their results on the first case of direct electrical stimulation of the brain, disproving the widely held view that the cortex was not an excitable substance. Electrical stimulation could lead to anything from small arm movements to generalized seizures, depending on the current applied. ---1874: Robert Bartholow in the USA described a 30-year-old woman who developed a skull ulcer as a result of irritation from a whale bone in her wig, which led to exposure of approximately 2 inches of both parietal lobes. Portions of both parietal lobes were stimulated at various currents via needle electrodes, eliciting muscle twitches, crying episodes (which ceased when stimulation ceased), loss of consciousness and ultimately a seizure involving the left side predominantly. ---1880S: Sir Victor Horsley, a British neurosurgeon, published reports of the use of cortical stimulation to better localize epileptogenic areas in three patients. With the assistance of Ferrier and J.H. Jackson, an epileptologist, he used stimulation to induce a typical seizure in each of the patients presumed epileptic zone (Horsley 1886). Also, David Ferrier and colleagues showed that direct electrical brain stimulation could change behavior and that activation of specific regions correlated with certain behavioral changes. ---1896: DArsonval produced phosphenes when a volunteers head was placed inside a coil driven at 42 Hz producing a time-varying magnetic field. ---early 1900s: in Vienna, psychiatrists working down the street from Freud, Adrian Pollacsek and Berthold Beer, filed a patent to treat depression and neuroses with an electromagnetic device that looked surprisingly like a modern TMS apparatus. --- 1930-1950s: Penfield and Jasper, using direct stimulation of the brain and simultaneous recording, were able to map out motor, sensory and language function in awake patients. In their acute experiments, they observed that in some cases focal electrical stimulation of the exposed cortex resulted in a flattening of the local electrocorticography (both normal rhythms and spontaneous epileptiform discharges). ---1972: Alberts using Delgado cortical strips modulated parkinsonian symptoms in humans. ---1984: Medical physicist Anthony Barker et al. in Sheffield succeeded in developing a magnetic stimulator that delivered field pulses short enough to elicit motor responses and allow neurophysiologic recording, first in the spinal cord, then the brain. ---1991: first published report of extradural motor cortex stimulation for central pain by Tsubokawa et al ---1995: first published report of extradural parietal cortex for central pain by Canavero ---1998: first report on tDCS effects by Priori ---2000: first published report of extradural motor cortex stimulation for Parkinson disease by Canavero ---2002: first patients receiving extradural cortical implants for stroke rehabilitation by Canavero and Brown ---early 2000s: first closed-loop cortical stimulator for epilepsy by NeuroPace ---2004: first published report of extradural auditory cortex stimulation for tinnitus by de Ridder ---2008: completion of first study assessing prefrontal extradural cortical stimulation for depression Selected references Bartholow R. Experimental investigations into the functions of the human brain. The American Journal of the Medical Sciences 1874; 134, 305-313 Ferrier D. (1873). Experimental Researches in Cerebral Physiology and Pathology. J Anat Physiol 8, 152-155. Fritsch G, Hitzig, E., ed. On the electrical excitability of the cerebrum. Charles C Thomas, Springfield, IL, 1960 Hitzig E. (1900). Hughlings Jackson and the cortical motor centres in the light of physiological research. Brain 23, 545581. Horsley S. Brain surgery. British Medical Journal 1886; 2, 670-675. Jackson J. (1890). On convulsive seizures. Lancet 1, 685-688, 735-738, 785-788. Penfield W, Jasper, H. Epilepsy and the Functional Anatomy of the Human Brain. Little and Brown, Boston, MA, 1954
Preface
xiii
An ideal therapy should be tailored to the individual and based on detailed knowledge of the pathophysiology of the specific patients condition, underlying disease and degree of disability. Such therapy should selectively target the involved neural networks, have minimal or no adverse effects, be effective and be financially and clinically feasible. CS, guided by neuroimaging and neurophysiologic measures of a patients derangement, can theoretically meet all these requirements. Presently, implantable cortical stimulation has proved to be superior to noninvasive approaches in such fields as chronic pain and movement disorders. Given its minimal invasiveness, high accuracy and transdural contact with the brain, it is likely that the surgical approach will continue to have the lead over nonsurgical approaches in the immediate future. However, the continuing progress in the field of noninvasive stimulation may forebode a time when a surgeon will not be needed any more. Besides the disorders covered in this textbook, other conditions may become the focus of surgical stimulation in the near future. Some of these include: 1- drug and smoke (by stimulating the insula) addiction; 2- multiple sclerosis (central fatigue: by motor and sensory cortex stimulation; immune derangement: by prefrontal cortex stimulation) and other immune-mediated disorders, by modulation of inflammation; 3- memory deficits due to dementia (by stimulating e.g. the dorsolateral prefrontal cortex, but also the motor cortex through its indirect effects on the hippocampus); 4- age-related cognitive impairment (by boosting GABA levels that typically decline with aging); 5- subarachnoid hemorrhage- associated vasospasm (via neuroprotective and vascular vasodilatatory- effects of on-site stimulation); 6- endocrinological disorders (e.g. by boosting the release of TSH). Outside clinical medicine, preliminary evidence suggests that cortical stimulation can improve cognitive performance under severe exhaustion, boost memory and learning and speed up geometric puzzle solving, temporarily energize sleep-deprived individuals, and even unleash savant skills. In other words, cortical stimulation may enhance human neural function (and fall into the scope of military applications). The serendipitous induction of an alien arm and an extracorporeal experience in patients by modulation of the tripartite right parieto-temporal-premotor circuit, which contributes to sensorimotor transformations, opens the way to recreational virtual reality cortical stimulation. In a world where boredom is shunned as a disease, the creation of virtual realities by cortical stimulation may become tomorrows new revolutionary market, even open the doors to spiritual experiences, as happens after ingestion of psychedelic drugs. Only, it would be in a controlled fashion. Also, the possibility of inducing orgasmic activity as witnessed on occasion during deep brain stimulation- may well become a topic of future research. Finally, since most thought arises from the coordinated firing of many brain regions, one may even think of simultaneously stimulating multiple brain areas invasively or noninvasively- in a coordinated way, thereby achieving control of behavior (mind control).
xiv
Sergio Canavero Cortical stimulation enters the stage! Sergio Canavero Turin, January 2009
BASIC ASPECTS
In: Textbook of Therapeutic Cortical Stimulation Editor: Sergio Canavero
ISBN 978-1-60692-537-9 2009 Nova Science Publishers, Inc.
Chapter 1
ANATOMICAL METHODS OF CORTICAL LOCALIZATION

Michael C. Park1,, Marc A. Goldman1, Gerhard M. Friehs1 and Sergio Canavero2
1
Department of Clinical Neurosciences Program in Neurosurgery, The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI, USA; 2 Turin Advanced Neuromodulation Group (TANG), Turin, Italy.
INTRODUCTION
Methods of localizing cortical landmarks e.g., the central sulcus (CS), a.k.a. Rolandos fissure, which separates the primary motor cortex (MI, precentral gyrus) anteriorly from the primary sensory cortex posteriorly, and the lateral sulcus (LS)- based on the relationship of cortical fissures and convolutions as they project to the cranial bony structures have been developed and utilized since the 19th century. In fact, cortical landmarks that identify function result from consistent sulcal and gyral folding that occur during development and are not merely coincidental occurrences. Despite the development and the availability of neuronavigation and image guided systems over the past 20 years, these craniocerebral topographic techniques are still utilized in surgical planning for most cranial surgery. A common reference is an imaginary line running almost straight up from the external auditory meatus near the midline, which identifies the superior aspect of MI, with the premotor cortex lying 2-3 cm anteriorly. As such, basic craniocerebral localization techniques also play an important role in therapeutic electrical cortical stimulation. One critical component of MI extradural cortical stimulation (ECS) is the localization and placement of stimulating electrode strips parallel or perpendicular to the CS. Success of the treatment highly depends on the correct placement of the electrode.
Correspondence concerning this article should be addressed to: Dr. Michael C. Park, M.D.: email: Michael_Park@brown.edu.
Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Numerous cortical localization methods exist in the literature today. In this chapter, we present the more well-known and widely utilized techniques for the localization of the CS and the LS, as well as the more recent techniques of identifying the precentral knob, or the hand representation, in MI for the purpose of therapeutic cortical stimulation (CS).
METHODS OF LOCALIZATION OF THE CENTRAL AND LATERAL SULCI

A- Craniometric Methods 1) Broca-Championnire In 1871, Pierre Paul Broca attended to a 38 year-old male who sustained an injury to his left frontoparietal scalp from a horse. While under his care, the patient worsened due to an intracranial abscess located in the language area localized to the temporal lobe and the patient underwent a craniotomy based on his localization method (Broca 1876). Broca described his experience: Je dterminai dabord, daprs les donnes topographiques indiques cidessus (p. 52), le lieu o devait tre applique la couronne du trpan. Sur une ligne horizontale mene par la base de lapophyse orbitare externe, je pris un longueur de 5 centimtres; puis, sur lextrmit postrieure de cette ligne, jlevai une perpendiculaire longue de 2 centimtres et jarrivai ainsi sur le point trpaner (Broca 1876). He followed a horizontal line from the base of the external orbital apophysis for 5 cm and then a perpendicular line 2 cm superiorly from that point to perform the craniotomy and drained the abscess. This localization of the proposed language area was later developed with the assistance of Lucas-Championnire into the procedure for localizing the Rolandic (central) and Sylvian (lateral) fissures, which was described by Auguste Broca and Pierre Maubrac in 1896: En 1871, P. Broca mit en pratique un procd opratoire grce auquel, sur le vivant, it ouvrit un abcs qui comprimait le pied de la 3e frontale. Ce procd est le suivant : en arrire du bord postrieur, toujours facile sentir, de lapophyse orbitaire externe, on trace une ligne horizontale longue de 6 centimtres, au bout de laquelle on lve une perpendiculaire haut de 2 centimtres; et lon est ainsi au niveau du centre du langage articul. Cela tant, il tait facile, en se portant 1 centimtre plus haut et 2 centimtres plus en arrire de marquer lorigine de la scissure de Rolando. De l le procd rgularis par Lucas-Championnire pour trouver lextrmit infrieure de la scissure de Rolando : a partir du bord postrieur de lapophyse orbitaire externe tirer une horizontale longue de 7 centimtres et au bout de celle-ci lever une perpendiculaire haute de 3 centimtres; chez la femme, la ligne horizontale naura que 6cm,1/2. Pour marquer le point suprieur, on se souviendra quil est en moyenne 47 millimtres en chiffres ronds 5 centimtres en arrire du bregma ; reste donc marquer le bregma (Broca and Maubrac 1896).
Anatomical Methods of Cortical Localization
In the Broca-Championnire method, lextrmit infrieure or the inferior Rolandic point (IE) is localized by determining a 7-cm line which starts at the external orbital apophysis of the frontal bone, parallel to the zygomatic arch, and le point suprieur or the superior Rolandic point (SP) is located on average 4.7 cm behind the bregma (Broca and Maubrac 1896). The CS is located on the line connecting the two points (Figure 1).
Figure 1. Top and lateral views of the skull illustrating Broca-Championnire method of central sulcus or Rolandic fissure localization. B: bregma (black dot), EOA: external orbital apophysis, IE: inferior end (inferior Rolandic point), RL: Rolandic line (solid black line), SP: superior point (superior Rolandic point), SS: sagittal suture, Z: zygomatic arch line (white dashed line).
2) Reid In 1884, Robert W. Reid published his guides by which, on examining a persons head, one might readily be able to localize the chief sulci, and thereby the position of the principal convolutions. In Observations on the relation of the principal fissures and convolutions of the cerebrum to the outer surface of the scalp he wrote: What I propose to do is to show that by taking large and easily felt landmarks on the head, and drawing from these certain lines, those lines will indicate accurately enough for all practical purposes the position of the principal sulci, and that by removing in any of these lines a piece of the scalp and skull an inch square, or by applying the one inch trephine to the skull, with the centre pin on the line, we can expose the fissure in any part of its course (Reid 1884) .
Based on skull landmarks which can be easily identified or palpated, Reid formulated reference lines which indicate the positions of the sulci. First and foremost, the base line, known as Reids base line, is defined as that which runs through the lowest part of the infraorbital margin and the middle of the external auditory meatus. The Sylvian fissure is localized by drawing a line from a point one inch and a quarter behind the external angular process of the frontal bone to a point three-quarters of an inch below the most prominent part of the parietal eminence. The Rolandic fissure is localized by drawing a pair of perpendicular lines to Reids base line. The first line is drawn from the depression in front of the external and auditory meatus and the second line is drawn from the posterior border of the mastoid process at its root. This creates a four-sided figure, bounded above and below by the lines for the longitudinal fissure and horizontal limb of the fissure of Sylvius respectively, and in front and behind by the two perpendicular lines. The diagonal line of this four-sided figure represents the fissure of Rolando (Figure 2).
Figure 2. Top and lateral views of the skull illustrating Reids method of central sulcus or Rolandic fissure localization. EAP: external angular process of the frontal bone (black dot), LF: longitudinal fissure, MPL: mastoid process line, PE: parietal eminence (black square), PEL: pre-external auditory meatus line, RBL: Reids base line, RF: Rolandic fissure (solid black line), SF: Sylvian fissure.
3) Poirier In 1891, Paul-Julien Poirier published La topographie cranio-ncephalique where he utilized craniocerebral topography for localizing the CS and LS, a practical system applicable to brain surgery (Poirier 1891, Charpy and Poirier 1899-1901). The superior Rolandic point is 2 cm behind the midpoint between the nasion and inion on the midline. The inferior Rolandic point is 7 cm above the zygomatic arch immediately anterior to the tragus, perpendicular to the zygomatic arch. The LS lies on the nasion-lambda line which starts at the temporal canthus to a point 1 cm anterior to the lambda on the midline (Figure 3).
Figure 3. Top and lateral views of the skull illustrating Poiriers method of central sulcus or Rolandic fissure localization. IRP: inferior Rolandic point (black star), ML: midline, NLL: nasion-lambda line, RL: Rolandic line (solid black line), SRP: superior Rolandic point (black open square), 50%: halfway point between nasion and inion.
4) Taylor-Haughton In 1900, Edward H. Taylor and William S. Haughton published a method for ascertaining the position of the CS and LS with reference to the surface of the head in Some recent researches on the topography of the convolutions and fissures of the brain. The introduction of the Taylor-Haughton lines, based on a cadaveric study of the relationship between the cranium and cerebrum utilizing x-rays, a relatively new technology at the time, summarized a
method based on easily identified landmarks, which was applicable to intact heads of various sizes without the need for delineating deeper structures through dissection.
Figure 4. Top and lateral views of the skull illustrating Taylor-Haughton method of central sulcus or Rolandic fissure localization. Central sulcus (RL) is the line connecting the SRP and IRP. IRP: inferior Rolandic point (black star), NI: nasion-inion line, OTA: orbito-temporal angle (solid black square), PAL: pre-auricular line, PAP: pre-auricular point (solid black dot), RL: Rolandic line (black line), SL: Sylvian line, SRP: superior Rolandic point (open black square), Z: zygomatic arch (upper margin) line, 50%: halfway point on the nasion-inion line (open black dot), 75%: three-quarters point on the nasioninion line.
Taylor and Haughton emphasized the localization of three fundamental fissures, the Sylvian fissure, the Rolandic fissure, and the external parieto-occipital fissure, which are necessary in identifying and separating the four main lobes of the brain from each other. Of these, the Rolandic fissure (CS or RL; Figure 4) is localized in the following manner by Taylor and Haughton: The fissure of Rolando is found as follows: - In the adult take a point (superior Rolandic) three-quarters of an inch behind the centre of the naso-inion line. This will usually be found to correspond to 53 to 55 per cent of the naso-inion distance, measuring from before backwards. The lower Rolandic point is found by erecting a perpendicular to the upper margin of the zygoma, starting from the pre-auricular point to meet the Sylvian line.
Should, however, the upper margin of the zygoma be difficult to determine, the perpendicular (pre-auricular line) may be erected on a line connecting the center of the infra-orbital margin with the center of the external auditory meatus or auricular point. The Rolandic line may now be drawn by connecting these two points. In 1980, the Taylor-Haughton line was re-evaluated using CT imaging for localization of the CS and LS (Taylor et al 1980). The study concluded that the Taylor-Haughton line provides a good approximation of the CS and the motor and sensory cortex can be accurately located for planning surgical procedures for anatomic studies.
Figure 5. Top and lateral views of the skull illustrating the method of CS and LS localization as described by Dr. Albert L. Rhoton, Jr. CS: central sulcus (solid black line), FZL/SF: frontozygomatic line / Sylvian fissure, FZS: frontozygomatic suture (solid black square), IRP: inferior Rolandic point (black star), NI: nasion-inion line, SRP: superior Rolandic point (open black square), 50%: halfway point on the nasion-inion line, 75%: three-quarters point on the nasion-inion line.
5) Rhoton In 1995, Albert L. Rhoton, Jr. published his procedure for localizing CS and LS based on a large number of cadaver studies (Rhoton 2002, Reis et al 2008). The LS is located on the frontozygomatic line which connects the frontozygomatic suture to the 75% point of the naso-inion line. The SRP is located 2 cm behind the 50% point of the naso-inion line.
10
Another line is drawn from the midportion of the zygomatic arch to the SRP where IRP is the point of intersection with the LS or frontozygomatic line. The localization procedure is summarized in Figure 5.
B-Image-Guided Methods Localization of neuroanatomical structures using neuroimaging of cadaveric and living patients brains has been described. 1) Kido Kido et al (1980), using CT axial views of cadaveric brain specimens with specially marked superior frontal, precentral and central sulci, as well as that of live patients (86 hemispheres, 50 patients), concluded that the approximate location of CS - even when it cannot be visualized on the scan- is a point 4.7 cm behind the coronal suture and 2.1 cm from the midline, and then drawing a line through this point at an angle of 67.9 from the midline (Figure 6). However, they did not extend their localization to a specific landmark, such as the precentral knob (see ahead).
Figure 6. Top view of the skull illustrating the predicted position of the central sulcus based on data from Kido et al (1980). The lines and measurements are superimposed onto the skull to estimate the location of the central sulcus.
2) Iwasaki Using CT and MRI axial views, Iwasaki et al (1991) described a method of identifying the precentral and postcentral gyri on the basis of the medullary branches of the cerebral white matter. The superior frontal gyrus, which runs parallel to the sagittal plane, will form an angle with the precentral gyrus; the CS will be localized posterior to it. 3) Naidich Using magnetic resonance (MR) sagittal images, Naidich et al (1995) described the cingulate sulcus as a landmark for recognition of the CS. The cingulate sulcus can be
11
followed in a midline sagittal image posteriorly, where its ascending segment is known as the marginal ramus or sulcus. The notch formed immediately anterior to it (at the vertex) is formed by the CS. The MR lateral sagittal image is also used to localize the central sulcus. The anterior horizontal and anterior ascending rami of the inferior frontal gyrus are identified. They have a Y-shaped appearance. The major descending sulcus immediately posterior to the "Y" is the precentral sulcus, followed posteriorly by the CS. The precentral sulcus is formed by two discontinuous grooves called the inferior precentral sulcus (Ebeling et al 1989) and the superior precentral sulcus. The CS courses approximately parallel to the zigzag precentral sulcus and almost never intersects the Sylvian fissure. Rather, the bottom end is closed by the fusion of the precentral and postcentral gyri. The CS runs uninterrupted in 92% of the cases, while the precentral sulcus is regularly interrupted in 100% of the cases (Yasargil 1994). 4) Yousry Yousry et al (1997) proposed a new landmark corresponding to the neural elements involved in motor hand function. Using functional MRI, the segment of the precentral gvrus that most often contained motor hand function was a knob-like, broad based, posterolaterally directed structure. It usually (90% of cases) has an inverted omega shape and sometimes a horizontal epsilon shape in the axial plane with a mean diameter of 1.4 cm. On average it is located about 23 mm from the midline, just posterior to the junction of the superior frontal sulcus with the precentral gyrus and 19 mm from the lateral surface. In the sagittal plane, this knob has the form of a posteriorly directed hook (92% of the cases), with a mean depth and height of 17 and 19 mm, respectively. It is located in the sagittal plane on the same section on which the insula can be identified, perpendicular to its posterior end. On the cortical surface of cadaver specimens, this precentral knob corresponded precisely to the characteristic middle knee' of the CS that has been described by various anatomists in the XIX century (see above). This precentral knob is a reliable landmark for identifying the precentral gyrus directly, in particular the neural elements involved in motor hand function (Boling et al 1999) . It faces and forms the middle knee of the CS, is located just at the cross point between the precentral sulcus and the CS, and is therefore also visible on the cortical surface. Based on measurements from the MRI, this hand area can thus be localized to an area 30 mm lateral to the midline and posteriorly following the superior frontal sulcus until it intersects with the precentral sulcus. Above this knob, the cortical portion subserving the chest and inferior limb can be found, with extension to the medial edge of the hemisphere or even in the interhemispheric part of the strip. 5) Park The omega shaped precentral knob for the hand representation in the primary motor cortex can also be localized on a CT. Based on CT scans of patients, the precentral knob on MI was consistently located 45.1 5.2 mm posterior with respect to the coronal suture line and 33.9 3.4 mm lateral to the midline on the right hemisphere, and 44.6 5.7 mm posterior and 33.2 2.5 mm lateral on the left hemisphere (Park et al 2007). This can also be transposed onto the skull to provide the approximate location of the precentral knob on MI with respect to the external skull markings (Figure 7). The approximate location of the
12
precentral knob based on CT is in agreement with the previously described methods. Based on their findings, Park et al concluded that current CT imaging can be used to consistently predict and localize structures such as the precentral knob, much like MRI. This indirect CTbased localization method can convey critical anatomic information to the neurosurgeon in the event that MRI or frameless guidance are not available or are contraindicated (e.g. in the presence of an implanted cardiac or neural pacemaker, metal debris or metal fragments in the body). The advantage of the CT scan in comparison to MRI would be the relative ease and availability of the imaging.
Figure 7. Top view of the skull illustrating the position of the precentral knob of the primary motor cortex with respect to skull landmarks, based on the measurements from CT of the brain (Park et al 2007). PCK: precentral knob.
Anatomical localization of MI can also be accomplished using image-guided neuronavigation systems, based on standard MRI, functional MRI (fMRI), positron emission tomography (PET) or magnetoencephalography (MEG) (Sobel et al 1993, Mogilner and Rezai 2001). 3D image-guided navigation systems help to determine the exact location of anatomical reference points, CS, LS, interhemispheric sulcus, superior and inferior frontal sulci, which can be clearly visualized on the reconstructed images. The superior frontal sulcus separates the superior frontal gyrus from the middle frontal gyrus, coursing anterior to posterior and perpendicular to the precentral sulcus. The next sulcus posteriorly may be identified as the CS. Referring to the cortical representation of body parts (cortical homunculus), the face area lies below the inferior frontal sulcus, the neck at the level of the inferior frontal sulcus, the arm and hand between the inferior and superior frontal sulci and the upper chest at the level of the superior frontal sulcus (Velasco et al 2002). These structures can be visualized by oblique MRI sections parallel to a plane from the sylvian to the interhemispheric fissure in the axial views (the Rolandic fissure is the deepest mesiolateral sulcus extending from the sagittal to the Sylvian fissure), a feat not possible with conventional MRI views (Velasco et al 2002). By applying vitamin E capsules glued onto the skin overlying vertex, inion, nasion and pterion and superimposed on the oblique MRI sections, Velasco et al (2002) found that the Rolandic fissure begins mesially at a point 3.903.96 cm behind the vertex and with a postero-anterior trajectory over the convexity toward a
13
point 0.3-0-9 cm above the angle formed by the zygomatic arch and the external orbital rim. The Rolandic fissure follows an angled trajectory of about 45 with respect to the sagittal plane (i.e. it is oblique to the cortical surface), its course being rather tortuous with two distinct and constant curves, one with an anterior concavity at the level of the superior frontal sulcus and another with an anterior convexity between the superior and the inferior frontal sulci. The oblique course may be relevant when the place of an extradural MI stimulator is decided on the basis of N20-P20 reversal (Chapter 2). The vertex-pterion line invariably lies anterior to the upper two thirds of the motor strip, which may be critical when implanting stimulating paddles parallel to the Rolandic fissure in order to target arm, pelvis, leg and thorax territories. On the other hand, cerebral localization using the 10-20 EEG system may not be not adequate (Homan et al 1987), although some surgeons use it for localization purposes. From a stereotactic perspective, the strikingly constant association of the CS with the midcallosal plane must be stressed. Lehman and colleagues (1992) demonstrated the reliability of a midcallosal line to intersect the inferior CS. Talairach et al (1993) recognized that most of the CS lies between two vertical planes extending from the AC and PC. The most superior aspect of the CS arises just behind the PC and the inferior sulcal limit just in front of the AC. The y coordinate (which determines the anteroposterior location) has the smallest range of values for face and tongue sensory activations consistent with the fixed position of the inferior CS in relation to midline commissural fiber bundles (Boling et al 2002). A special reference should be made to the portion of the primary sensory cortex responsible for facial and buccal sensation in reference to cortical stimulation for facial pains. In open neurosurgical approaches, the tongue sensory area is easily recognized as a triangular gyral configuration in the most inferior aspect of the postcentral gyrus, just above the Sylvian fissure. The distance between the Sylvian fissure and the thumb-hand area is no more than 3 cm (Boling et al 2002). Anterior tongue sensory function can be localized to the superior aspect of the triangle. Lower face sensation is subserved by the narrow portion of the postcentral gyrus, which is situated immediately above the tongue area. The apex of the tongue sensory area thus narrows immediately into a thin segment of the postcentral gyrus, which is the substratum of low face sensation (Boling et al 2002). The tongue sensory area is the only area where the postcentral sensory gyrus is not thinner than the precentral motor cortex. When placing an extradural stimulating strip straddling the motor and sensory cortex, it is important to bear in mind this particular triangular anatomic configuration.
CONCLUSIONS
Individual variability causes the motor strip to lie anywhere from 4 to 5.4 cm behind the coronal suture, and landmark- based linear measurements may vary by up to 2 cm. This is mainly due to the natural individual variability and must be kept in mind when linear measurements based on anatomical landmarks from CT images are to be used to localize the precentral knob area.
14
One might think that the advent of frameless guidance system or neuronavigation has largely replaced localization relative to bony landmarks. Nonetheless, Reis et al (2008) concluded that the above-described classic craniocerebral topographic procedures are reliable methods for localizing CS and LS as well as associated landmarks and most cogently- that modern imaging and neuronavigation have improved upon these methods by only 1 -3 mm! Velasco et al (2008) too found that Rolandic fissure localization by external landmarks is satisfactory (mean error: 1.6 mm, range: 0-7 mm). Box: Craniometric Points BOX: CRANIOMETRIC POINTS Pterion: region of approximation of the frontal, parietal, temporal and sphenoid (greater wing) bones; it lies approximately 2 finger-breadths above the zygomatic arch, and a thumb's breadth behind the frontal process of the zygomatic bone. Asterion: junction of the lambdoid, occipitomastoid and parietomastoid sutures. It overlies the junction of transverse and sigmoid sinus. Lambda: junction of the lambdoid and sagittal sutures. Stephanion: junction of the coronal suture and the superior temporal line. Glabella: the most forward projecting point of the forehead at the level of the supraorbital ridge in the midline. Opisthion: the posterior margin of the foramen magnum in the midline.
REFERENCES
Boling W, Olivier A, Bittar RG, Reutens D. Localization of hand motor activation in Broca's pli de passage moyen. J Neurosurg 1999; 91: 903-910 Boling W, Reutens DC, Olivier A. Functional topography of the low postcentral area. J Neurosurg. 2002; 97:388-95 Broca A, Maubrac P. Trait de chirurgie crbrale. Paris: Masson et Cie, 1896. Broca P. Sur la topographie cranio-crbrale ou sur les rapports anatomiques du crane et du cerveau. Rev dAnthrop 1876; 5: 193-248. Charpy A, Poirier PJ. Trait danatomie humaine: systme nerveux [in French]. Paris: Masson, 1899-1901. Ebeling U, Steinmetz H, Huang YX, Kahn T. Topography and identification of the inferior precentral suIcus in MR imaging. AJNR Am J Neuroradiol 1989; 10: 937-942 Homan RW, Herman j, Purdy P. Cerebral location of international 10-20 system electrode placement. Electroencephalogr Clin Neurophysiol 1987; 66: 376-382 Iwasaki S, Nakagawa H, Fukusumi A, et al. Identification of pre- and postcentral gyri on CT and MRI images on the basis of the medullary pattern of cerebral white matter. Radiology 1991;179:207-213, Kido DK, LeMay M, Levinson AW, Benson WE. Computed tomographic localization of the precentral gyrus. Radiology 1980; 135: 373-377.
15
Lehman RM, Olivier A, Moreau JJ, Tampieri D, Henri C. Use of the callosal grid system for the preoperative identification of the central sulcus. Stereotact Funct Neurosurg 1992; 58:179-188 Mogilner AY, Rezai AR. Epidural motor cortex stimulation with functional imaging guidance. Neurosurg Focus 2001; 11: issue 3, E4 Naidich TP, Valavanis AG, Kubik S. Anatomic relationships along the low-middle convexity: Part I-Normal specimens and magnetic resonance imaging. Neurosurgery 1995; 36:517-532 Park MC, Goldman MA, Park MJ, Friehs GM. Neuroanatomical localization of the precentral knob with computed tomography imaging. Stereotact Funct Neurosurg 2007; 85: 158-161. Poirier PJ. La topographie cranio-encphalique. Paris: Lescrosnier et Bab, 1891. Reid RW. Observations on the relation of the principal fissures and convolutions of the cerebrum to the outer surface of the scalp. Lancet 1884; 2: 539-540. Reis CV, Sankar T, Crusius M, et al. Comparative study of cranial topographic procedures: Brocas legacy toward practical brain surgery. Neurosurgery 2008; 62: 294-310. Rhoton AL Jr. The cerebrum. Neurosurgery 2002; 51(4 suppl.): S1-S51. Sobel DF, Gallen CC, Schwartz BJ, et al: Locating the central sulcus: comparison of MR anatomic and magnetoencephaloraphic functional methods. AJNR Am J Neuroradiol 1993; 14: 915-925 Talairach J, Tournoux P, Missir O. Referentially Oriented Cerebral MRI Anatomy: An Atlas of Stereotaxic Anatomical Correlations for Gray and White Matter. pp 51-75, New York: Thieme Medical, 1993, 51-75 Taylor AJ, Haughton VM, Syvertsen A, Ho KC. Taylor-Haughton line revisited. AJNR Am J Neuroradiol 1980; 1: 55-56. Taylor EH, Haughton WS. Some recent researches on the topography of the convolutions and fissures of the brain. Trans R Acad Med Ireland 1900; 18: 511-522. Velasco M, Velasco F, Brito F, et al. Motor cortex stimulation in the treatment of deafferentation pain. I. Localization of the motor cortex. Stereotact Funct Neurosurg 2002; 79: 146-167 Velasco F, Arguelles C, Carrillo-Ruiz JD, et al. Efficacy of motor cortex stimulation in the treatment of neuropathic pain: a randomized double-blind trial. J Neurosurg 2008; 108: 698-706 Yasargil MG. Microneurosurgery, Vol. IVA. CNS tumors. Stuttgart-New York: Thieme, 1994:1-114. Yousry TA, Schmid UD, Alkadhi H, et al. Localization of the motor hand area to a knob on the precentral gyrus. A new landmark. Brain 1997; 120: 141-157.
Chapter 2
FROM LOCALIZATION TO SURGICAL IMPLANTATION

Youichi Saitoh and Koichi Hosomi
Department of Neurosurgery, Osaka University Graduate School of Medicine, Japan.
INTRODUCTION
In its landmark paper introducing MI ECS for the treatment of central pain, Tsubokawa et al (1991) included a brief description of the localization of the motor cortex: Location of the motor cortex was estimated by bony landmarks with conventional methods. Paramedian incision was made 1-4 cm lateral to the midline contralateral to the painful area. The trephination was then placed over the estimated area of the motor cortex. The locations of the sensory and motor cortices were confirmed from phase reversal of the N20 wave of somatosensory evoked potential recorded from the electrode. When the electrode was moved from the sensory cortex to the motor cortex, the N20 wave turned positive. The location of the motor cortex was again confirmed by motor evoked potential recoded in response to stimulation with the electrode. The motor cortex was mapped as carefully as possible and the electrode was placed in the region where muscle twitch of painful area could be observed at the lowest threshold. Identification of the precise location of the central sulcus (CS) remains one of the key steps in this kind of surgery, and several techniques have been used for this purpose, although they have not been systematically compared. Most neurosurgeons use classical anatomical landmarks to determine the exact position for a craniotomy (chapter 1). Somatosensory evoked potentials (SSEP) and intraoperative motor evoked potential (MEP) measurements are also employed. In recent years, navigation systems have become increasingly important.
Correspondence concerning this article should be addressed to: Dr. Youichi Saitoh, M.D., Ph.D. e-mail: neurosaitoh@mbk.nifty.com
18

There is no current consensus on the ideal surgical approach. Most neurosurgeons implant an electrode in the epidural space, but some implant it subdurally or in the interhemispheric or within the central sulci. The direction of the implanted paddle also remains controversial. This chapter summarizes and compares these various procedures.
ANATOMICAL LOCALIZATION
As detailed in chapter 1, there are several landmarks for detecting the location of CS on the scalp and cortical surface. The CS can be expected to lie approximately 4 - 5.4 cm posterior to the coronal suture on the scalp midline, and can be localized with the aid of Taylor-Haughton lines (Figure 1). The CS is easily located with preoperative magnetic resonance imaging (MRI): it is characterized by the lack of sulcal branches, and lies just
Figure 1. Taylor-Haughton line indicates the position of the central sulcus from the scalp. Taylor-Haughton (T-H) lines can be constructed on an angiogram, CT scout film, or skull x-ray, and can then be reconstructed on the patient based on visible external landmarks. The Frankfurt plane (a.k.a. baseline) is the line from the inferior margin of the orbit through the upper margin of the external auditory meatus (EAM) (as distinguished from Reid's base line, running from the inferior orbital margin through the center of the EAM). The distance from the nasion to the inion is measured across the top of the calvaria and is divided into quarters (this can be done simply with a measuring tape). The posterior ear line runs perpendicular to the baseline through the mastoid process (intersecting the skull sagittal midline about 1 cm behind the vertex and 3-4 cm behind the coronal suture). The condylar line runs perpendicular to the baseline through the mandibular condyle (intersecting the line representing the sylvian fissure). The Sylvian fissure (a.k.a. lateral fissure) is approximated by a line connecting the lateral canthus to the point 3/4 of the way posterior along the arc running over convexity from nasion to inion (or by a line drawn 45 to Reids line starting at the pterion). A point 5 cm straight up from the external auditory meatus intercepts MI. The angular gyrus (which includes Wernickes area) is located just above the pinna, with significant individual variability in its location.
From Localization to Surgical Implantation
19
Figure 2. Reconstructed oblique MRI view of the central sulcus (arrow: central sulcus).
anterior to the pars marginalis of the cingulate sulcus on the interhemispheric surface (Naidich et al 1995, 2001). Oblique MRI views easily show the CS (Figure 2). Penfields Homunclus is commonly used for identification of the corresponding body parts on the precentral or postcentral gyrus. As discussed in chapter 1, the precentral knob sign corresponding to the hand is easily identified on surface MRI anatomic scans. From these findings, the position for the craniotomy can be decided. A simple way to locate the motor strip is to use a Callosal Grid system (Lehman and Kim 1995). Establishing a horizontal plane (HP) through the inferior border of the genu and the splenium of the corpus callosum creates such a proportional grid system. Three vertical planes perpendicular to HP are constructed: the anterior callosal plane (AC), the posterior callosal plane (PC), and the midcallosal plane (MC) in the midpoint between the AC and MC. If the grid is overlapped over the cortical surface, the junctional point between the HP and the MC corresponds with the inferior point of the CS, where a central artery enters the CS. The superior CS lies 4mm anterior to PC. Imaginary lines connect the inferior and superior points that constitute the CS.
ELECTROPHYSIOLOGICAL LOCALIZATION
SSEPs can be measured by stimulating the contralateral median nerve at the wrist; stimuli consist of single shocks (0.5 ms, 4.7 Hz, 20 mA) to produce a small, but consistent contraction of the thumb. SSEPs are recorded from each cortical electrode referenced to the ipsilateral ear lobe. Individual SSEP signals are differentially amplified and filtered: 200 are averaged through a digital signal analyzer with sample interval of 40 s. The phase reversal
20
of the N20 (sensory cortex) /P20 (motor cortex) waves is used to confirm the location of the CS (Wood et al 1988, Velasco et al 2002), using a 20/32-contact grid and the central scalp EEG leads or directly using the definitive 4-contact strip overlying the dura (Figure 3). Polarity inversion of potentials across the sulcus is less reliable and technically more difficult for trigeminal SEPs (McCarthy et al 1993) and only occasionally a phase reversal has been described for tibial nerve SEPs (Maegaki et al 2000). Other later components (P25-N25; P30N30) also present phase reversals, but may not be observed. Phase reversal seems to be a rather constant feature of SSEPs, provided that the orientation of the electrode is perpendicular to the Rolandic fissure, otherwise it may not be present. The Rolandic fissure being tortuous and oblique along the convexity of the brain, it is difficult to believe that phase reversal may be obtained in all cases without the assistance of visual inspection of the sulci, as in the case of epidural recordings (Velasco et al 2002). Although N20 is generally recordable, the P20 component may be missing, even in awake or mildly sedated patients, in part due to the dipole generator of the early component at each side of the fissure which, having an oblique depth posterior trajectory with respect to the cortical surface, orients the N20 component toward the surface and the P20 component toward the depth. Moreover, the N20/P20 reversal is only useful for hand representation targeting. The rolandic tortuosity that in some parts becomes parallel to the midsagittal line complicates the placement of the 4contact strip on MI by determining reversal only in a single point. Even worse, maximal N20 amplitude can be reached on either motor or sensory cortex in several patients (Wood et al 1988, Velasco et al 2002). Both components can be severely attenuated by nervous system lesions: patients with phantom-limb pain, brachial plexus avulsion, severe stroke or other similar conditions may thus show no SSEPs. Most importantly, inferring the position of the motor hand area from the position of the maximum amplitude median nerve SEP is impossible. Indeed, Woolsey et al (1979) demonstrated that the face-arm boundary is situated more laterally on the postcentral gyrus than on the precentral gyrus, by 12 cm. It is therefore necessary to map the motor cortex. Penfield and Boldrey (1937) first systematically stimulated the sensory-motor cortex and described the sensory and motor homunculus. They utilized a bipolar direct stimulation of the cortex, applying 5060 Hz stimuli up to 20 mA for 14 s, and looked for movements or sensations in the awake patient. This technique required awake surgery, often induced complex movements involving more than one muscle and provoked epileptic seizures in a high percentage of cases (2025%). As regards the leg representation, Woolsey et al (1979) found that in only one third of the cases the lower extremity is on the medial surface of the hemisphere, in two thirds of the cases it extends on the lateral surface and, in 27% of the cases, the whole lower extremity is on the lateral surface. Recently, an enlarged and displaced motor map for the hand area was described in Parkinson's disease patients. Map shifts were found in the majority of the patients (12/15), both in untreated early cases and treated cases of long duration, with a correlation between the inter-side difference in the severity of PD symptoms (UPDRS) and interhemispheric map displacement (Thickbroom et al 2006). In sum, guidance of epidural electrode placement is often inadequate or impossible by SSEPs. According to Velasco et al (2002), recording corticocortical evoked responses (CCER) is simple and reliable and superior to SSEPs. MI stimulation elicits negative CCER over the
21
frontal scalp, whereas SI stimulation elicits positive responses over parietal and occipital scalp regions.
Figure 3. SSEPs show phase reversal of the N20 wave, when the median nerve is stimulated. The black wide line indicates the estimated location of the central sulcus.
Most neurosurgeons attempt intraoperative test stimulation by using the quadripolar or the grid electrodes. Test bipolar stimulation (210-1000 s generally 400-500-s, 1-5 Hz up to 100Hz, at increasing voltage or intensity up to 50 mA, anodally, but also cathodally) is applied by means of the contacts situated over the motor or sensory cortex. In general, the amplitude needed to produce motor responses is higher using epidural rather than subdural stimulation. Motor contraction can be elicited at relatively lower amplitudes when general anesthesia is not employed. 1Hz stimulation is preferred to higher frequencies, since the former does not habituate and has less potential to trigger seizures. Muscle responses are recorded from muscle bellies of the contralateral hemibody, with EMG needle electrodes or visually. The supposed advantage of a grid electrode is nixed by the observation that over
22
75% of the cortical surface is not covered by its contacts, not to mention the imperfect superposition with the definitive strip electrode. Under local anesthesia, the patients may describe pain reduction, paresthesias in the painful body part, muscle twitching or contraction or nothing at all (Canavero and Bonicalzi 2002, Velasco et al 2002). Interestingly, stimulation of both MI and SI can elicit similar motor or sensory responses and both motor and sensory responses can be obtained from the same contacts (Schmid et al 1980, Wood et al 1988, Canavero and Bonicalzi 2002, 2007), making it difficult to rely on motor and sensory responses to differentiate MI from SI. In some institutions, motor evoked potentials (MEPs) are measured under general anesthesia. Holsheimer et al (2007) stressed the importance of intra-operative MEP measurement obtained by monopolar and bipolar stimulation for determining the location of the electrodes bringing most pain relief during chronic ECS. Monopolar stimulation appeared superior at determining the optimal point for chronic motor cortex stimulation. They concluded that the anode yielding the largest intra-operative MEP should be selected as the cathode for chronic stimulation. Intraoperative D-wave recording of corticospinal MEPs has been utilized to optimize electrode placement; the D-wave has been recorded with a flexible wire electrode placed epidurally in the cervical (C3-4) spinal cord during high intensity anodal monopolar stimulation of each plate electrode under general anesthesia. The aim was to evoke the D-wave of highest amplitude. Pain reductions significantly correlated with the recorded amplitude of the D wave employing the same stimulation electrode. The result was analgesia with lower voltages than generally required (Yamamoto et al 2007).
LOCALIZATION BY FUNCTIONAL MRI (FMRI)

fMRI has been explored extensively in terms of functional localization. In particular, sequential tapping of fingers in a predetermined fixed order or repetitive opposition of the thumb and each of the remaining fingers activates MI contralaterally (but also the ipsilateral MI, supplementary motor and premotor areas and the primary somatosensory areas bilaterally). This method appears to be superior to the competing methods described above. Pirotte et al (2005) utilized fMRI to identify the hand and tongue motor area. fMRI data were coregistered on 3D T1-weighted MRI anatomical scans and matched with data from intraoperative neurophysiology. The operation was performed under TIVA (see ahead); they utilized median-nerve SEP phase reversal to identify the CS and mapped the motor cortex using the 60 Hz Penfield's technique. In 61% of cases, they found a good correlation between fMRI and intraoperative neurophysiological data, with a mean distance of 3.8 1.3 mm between the two hot spots, which is sufficiently accurate, considering that the activation area of an electrode measures 5 mm; in 33% of the cases, intraoperative neurophysiology provided ambiguous results because of electrical artifacts, influence of anesthesia, SEP attenuation, diffuse motor responses, or sensorimotor disconnection. Some of these problems were due to an inadequate mapping technique; in 6% of the cases they reported poor localization with both fMRI and intraoperative neurophysiology. Blood oxygenation level depending (BOLD) fMRI data of motor functions of the tongue, arm or leg are obtained by using standardized paradigms, such as repetitive contraction of the
23
lips, cyclic finger tapping of the contralateral hand, or flexion-and-extension of the toes of the contralateral foot at a rate of 1 Hz after a training session is performed. Post-stroke motor deficits may hamper examination, but fMRI is particularly useful for amputees or plexus avulsion patients, since virtual movements of the phantom or paralytic limb easily induce contralateral precentral and postcentral gyri activations. Blocks of 30 seconds of alternating activation and rest are repeated a few times (Canavero and Bonicalzi 2007). Generally, a focal cortical activation area (diameter 5-10 mm) after hand motor tasks is localized to the contralateral precentral gyrus, but differences between the two sides in the surface and minor displacements of the precentral activation area are frequently observed (also due to cortical remapping). It remains a matter of debate which area is more suitable as a stimulation target, the mirror image of the activation site on the healthy side or the displaced activation site on the affected side. Stippich et al (2004) developed a fully automated super-fast fMRI method for SI localization. Diffusion tensor imaging (DTI) tractography has been suggested as a means to identify the motor cortex (Kamada et al 2005), but it should be borne in mind that DTI is a mathematical probability function, not an anatomical image. CS veins have been used as landmarks during subdural approaches, but, being sometimes located deep in the sulcus, cannot be identified by examining only the cortical surface (Saitoh and Yoshimine 2007).
NEURONAVIGATION
Several kinds of navigation systems for neurosurgical assistance can be used to estimate the position of the central sulcus or other structures, both on the dura mater and scalp (Tirakotai et al 2007). Neuronavigation combined with fMRI data help to decide the best position for craniotomy and for placement of the stimulating paddle (Rasche et al 2006). While in most cases fMRI data can be satisfactory matched with navigation data, in some cases, motion artifacts and low signal levels interfere with fMRI data analysis. A drawback of neuronavigation is the requirement that the patients head be fixed in a 3-point pin holder or vacuum headrest (Tirakotai et al 2007), which several patients may not tolerate under local anesthesia. For this reason, other surgeons prefer not to fix the patients head and operate without navigation.
TECHNIQUES OF IMPLANTATION
Implanted electrodes used in reported studies have generally been Resume (Medtronic) or Lamitrode (ANS) stimulating quadripolar strips; eight-contact paddles have been used in one failed trial (NCT00122915). Most neurosurgeons (including Tsubokawa) implant the strip in the epidural space, but a few insert the strip subdurally (Saitoh et al 2000, 2007, Kleiner-Fisman et al 2003, Strafella et al 2007). In patients with extensive painful areas, two strips are positioned.
24
Anethesia is induced with a loading dose of Remifentanil in continuous infusion followed after 58 min by Propofol as induction dose (Total intravenous anesthesia, TIVA). Endotracheal intubation is facilitated by vecuronium bromide; no further doses of muscle relaxants are administered throughout surgery. The lungs are mechanically ventilated with a 50% O2 in air mixture, in order to maintain end tidal concentrations of CO2 at 3035 mmHg. Anesthesia is maintained with Remifentanil and Propofol. At the end of the surgical procedure, all patients are awakened within 1530 min from cessation of TIVA. A small craniotomy or burr-hole is made around the CS. The four-contact electrode array (each contact 5 mm in diameter; inter-contact distance center-to-center 1 cm) is usually placed in the epidural space. The best location and orientation of the electrode array are generally determined in such a way that bipolar stimulation with an appropriate pair of electrodes can be attained. Some surgeons place the electrode perpendicular to the CS above the precentral (cathode) and postcentral (anode) gyri for the supposed improved selectivity (e.g. Nguyen et al 1999), others in a parallel fashion, i.e. with all contacts on MI or SI (e.g. Canavero and Bonicalzi 2002, Rasche et al 2006), but there appears to be no difference between the results of these two approaches (Tsubokawa et al 1991, 1993). Moreover, no polarity-related difference in pain relief is seen for most patients with epidural electrodes (Katayama et al 1998). Saitoh et al (2000; Hosomi et al 2008) implanted the paddle subdurally on the cerebral or interhemispheric surface) or within the CS. The latter makes it possible to stimulate MI more directly (Takahashi et al 2002, White et al 1997). Nuti et al (2005) implanted the strip electrode on the interhemispheric surface to treat leg pain. In some patients with brain atrophy, the cortical surface and the dura mater are wide apart, in which case patients may fail to respond to extradural stimulation: a subdural approach may be considered in a few, highly select cases. After the implantation of the electrodes, a test period of a few days to 2-4 weeks follows. If the stimulation proves effective, under general anesthesia, a pulse generator is then implanted subcutaneously below the clavicle and connected to the paddle via a subcutaneous extension. For movement disorders patients, a single surgical procedure may be used (Canavero and Bonicalzi 2007b). Stimulating paddles have been implanted through several approaches:
1-Epidural Single Burr-Hole (Figure 4) Both Tsubokawa et al (1991) and Meyerson et al (1993) performed MI ECS using a single burr-hole made on the central sulcus under local anesthesia. For leg pain, a paddle is placed on the medial edge of the hemisphere, but involves some risk of developing an epidural hematoma. This technique may require relocations before an optimal position is found, and thus increase the risk of epidural bleeding due to dural detachment. However, this was not a problem in recent navigated series (Rasche et al 2006).
25
Figure 4. Single burr-hole surgery with the assistance of neuronavigation system and fMRI. Insertion of the stimulating paddle (which is shown reversed for clarity) is performed via a single burr hole (with permission from IASP).
Figure 5. Two burr-hole surgery is shown. The locations of burr holes are marked on the scalp depending on the anatomical landmarks (courtesy of Prof. Canavero).
26
2-Two Epidural Burr-Holes (Figure 5) Canavero (Canavero and Bonicalzi 2002) makes an oblique linear skin incision (6-10cm) parallel to and 1 cm ahead of or behind the projection of the CS and then drills two burr holes at a distance of 2-4 cm (plus a bony groove parallel to the paddle to accommodate the connector between the looping lead and the extension). A stimulating paddle is inserted from the edge of one burr hole into the epidural space overlying the precentral gyrus or post central gyrus contralateral to the painful area or most disabled side for movement disorders. The bony bridge between the two holes will then hold the plate in place and simultaneously reduce the durocortical gap. For facial or leg targets, the paddle can be gently advanced caudally or rostrally by up to 2 cm. This technique entails no risk of epidural hematoma, and accidental displacement of the electrode has never been observed (S Canavero, personal communication).
3-Epidural Bone flap Because of greater availability of the epidural area for electrophysiological exploration and mapping, the procedure has been proposed to result in improved outcome (Nguyen et al 1999), but this has not been confirmed. A small craniotomy (4-5 cm) is made on the central sulcus. The center of the craniotomy should correspond to the target as determined by imaging. This technique allows SSEP recordings from electrodes placed on the dura mater. The paddle is fixed to the dura mater with two stitches (making accidental displacement impossible), which may theoretically catch on a vessel and cause intracerebral bleeding. The risk of inadvertent opening of the dura during bone detachment must be borne in mind.
4-Subdural Method In patients with advanced cortical atrophy, epidural stimulation may fail due to the durocortical separation. The cortical surface and interhemispheric surfaces subdurally may be elected as targets for stimulation. However, large bridging veins sometimes interfere with implantation on the interhemispheric surface and adhesion may occur due to subarachnoid hemorrhage. Moreover, dissection of the CS involves the risk of developing new neurological deficits due to brain damage or vein obstruction. For upper limb and/or face pain, the arachnoid membrane of the CS must be carefully dissected and the vessels within the central sulcus must be freed with a microsurgical procedure to expose the hidden lateral walls of the precentral and postcentral gyri (Hosomi et al 2008). Since the paddle is too stiff to be placed within the CS, it must be trimmed off (Figure 6). Saitoh et al (Hosomi et al 2008) limited most of the implantations within the CS to patients with severe motor weakness or lack of function. In their series, test stimulation of MI within the CS was more effective in most cases than subdural stimulation on the cerebral surface, but long-term clinical results were not significantly superior to ECS (Saitoh and Yoshimine 2007) and most of the patients who received the implantation within the CS gained only temporary pain reduction (maximum: 6
27
months) (Hosomi et al 2008). At the end of surgery, the lead extension is fixed to the dura or the border of the burr hole with a silk suture to prevent dislocation. However, migration of the electrodes seems to be more of a problem with a subdural than an extradural approach. A meticulous, watertight dural closure is mandatory to minimize the risk of cerebrospinal fluid leakage.
Figure 6. After trimming to reduce stiffness, a Resume paddle is implanted within the central sulcus (with permission from IASP).
COMPLICATIONS
Of all published cases of invasive cortical stimulation, 11.4% were associated with one or more adverse effects. Speech disorders (aphasia/dysarthria), although rare and generally temporary, have been observed (Ebel et al 1996, Canavero and Bonicalzi 2002). Some cases of headache reportedly associated with stimulation of the face area may actually be due to contraction of the temporalis muscle (Canavero and Bonicalzi 2002). However, local pain may be relieved by incision and resuturing of the dura around the electrode or, more simply, by bipolar coagulation. Other reported side effects include fatigue, paresthesia and dysesthesia (2.2% of cases), but also, exceptionally, impairment in motor imagery tasks (Tomasino et al 2005) and supernumerary phantom arms (Canavero and Bonicalzi 2002, 2007). Montes et al (2002) analyzed event-related potentials (ERPs) and behavioral performance during an auditory target-detection task in 11 consecutive patients obtained during MI ECS and 10 minutes after switching off stimulation. While sensory responses remained unaffected by MCS, there was a significant delay of brain potentials reflecting target detection in the
28
older patients (N2 and P3), rapidly reversible after MI ECS discontinuation. No effect was observed in patients younger than 50 years. Individually, the effect was highly variable from no effect to a delay of tens of milliseconds. Cognitive effects of MCS appeared as mild and non-specific, directly related to the stimulation period (i.e. with no post-effect), in a manner reminding of cognitive effects reported during MI rTMS. Thus, MCS may interfere with relatively simple cognitive processes such as those underlying target detection, notably in the elderly and in the presence of preexistent cerebral lesions. Occurrence of epileptic seizures, probably due to differences in testing conditions, has been reported during test stimulation in a minority of patients. The low rate of epileptic seizures during chronic stimulation (0.2%) means that stimulation of MI within an appropriate range of parameters is reasonably safe. The most serious reported complications are epidural or subdural hematomas. These are definitely exceptional with an extradural approach, and some surgeons never observed one, making the risk of peri-operative hemorrhage much lower compared to deep brain stimulation. However, in our series using a subdural approach, two patients developed cerebral hemorrhage: one died and the other remained in a vegetative state (Hosomi et al 2008). This is especially true for patients with post-stroke pain, who are likely to develop a new stroke in the years following the first one. Some wound infections have been reported by most surgeons. If the infection occurs, all devices including the paddle, extension leads, and pulse generators must be removed temporally. Patients with post-stroke pain frequently have diabetes mellitus and thus are at greater risk. The implanted pulse generator (IPG) can accidentally turn off due to electromagnetic interference from household devices in close (<10 cm) proximity, such as electric appliances of any kind, but also anti-theft devices and metal detectors or magnets in loudspeakers. At impedances >2000 , a connection problem, such as a broken cable or a lead fracture, must be suspected. The operator should thus measure impedance in a unipolar configuration in order to assign a value to the single contact. The so-called radio test may be useful: IPGs emit a signal at 500-550 kHz which can be received as a continuous hum on a small battery operated AM radio receiver.
SAFETY OF STIMULATORS
The output of commercial stimulators are either of the controlled current (CC) or controlled voltage (CV) type. CC output circuits are somewhat more complex and less power-efficient than the CV type, but may provide a more stable level of stimulation, especially where the electrode impedance may fluctuate as a result of e.g. changing contact with tissue, formation of scar tissue, polarization potentials on the metal/electrolyte interface. Most of the stimulus voltage of macro-electrodes is dissipated in the impedance of the tissue. If the impedance changes in a CC circuit, the output voltage automatically rises to keep the current flow constant. If the impedance changes in a CV circuit, the voltage stays the same and the current changes. Stimulators use pulsatile rather than sinusoidal current waveforms. The stimulus waveform may be monophasic or biphasic. In biphasic stimulation, the negative voltage
29
applied is balanced by an equal amount of positive voltage. This is generally considered to be far safer than monophasic stimulation as it allows for a balance of ionic exchange at the electrode-cortex interface. If the charge is not balanced, it is possible that metal ion deposition will occur at the interface, which may cause deleterious effects for both the tissue and the electrode (Polikov et al 2005). A fast-rising rectangular pulse of negative current is the most efficient stimulating waveform. Each pulse delivers a charge (Q) of current per phase (CPP): (1) Q = A x PW Charge density (CD) of the different cathodal pulses is given by: (2) QD= A x PW / cathodal area. These formulae indicate that: 1) maximal safety (i.e. minimal tissue damage) is obtained by applying short pulse durations (Crago et al 1974), i.e. slightly greater than chronaxie, which are also ideal to evoke neuronal responses (Tehovnik 1996). 2) QD for a given current is limited by the size of the electrode: more current is required to stimulate the cortex using large electrode than using small electrodes, but decreasing the surface area of the electrode may increase the extent of histological damage. Consequently, paddle or strip electrodes containing multiple contacts are ideal to recruit more neurons in the stimulation field. The QD threshold is lower when using surface electrodes (ECS) versus depth electrodes (DBS). Luckily, functional alteration can be achieved by charge densities much lower than those required for histological damage. The least damaging pulse waveform is that with no net direct current (DC), which can lead to tissue damage even at very low intensity. The current density and charge per phase (CPP)- but not frequency, waveform or periods between pulses- are likely the most important factors in determining safety of a particular stimulation protocol (McCreery et al 1990). Ideally, the CPP and current density ought to be minimized. According to Pudenz et al (1975, 1977), CPP must not exceed 0.3 C(oulombs) in each half of the stimulating pulse. In a human study, subdural stimulation with 0.3 ms square wave pulses at 50 Hz and 12.5-15 mA delivered for 24 hours achieved a maximum CPP of 4-4.4 C, with a maximum charge density of 52-57 C/cm2, which suggests a greater ability of the human brain to accommodate higher currents and current densities (Gordon et al 1990). Choice of stimulation features must aim to prevent electrode dissolution and generation of electrochemical toxic products at the electrode interface (by electrode polarization and hydrolysis: Bartlett et al 1977). Electrodes should be made of corrosion-resistant noble metals or alloys, such as platinum-iridium (actually, there is a small corrosion rate, which may in principle lead to toxic cumulative effects over many years). A charge-balanced symmetrical biphasic waveform with two closely spaced pulses of equal charge, cathodal followed by anodal, is the stimulus waveform recommended for avoiding tissue damage and electrode corrosion. In fact, the electrochemical reactions occurring during the first pulse (phase) are
30
reversed by the following pulse of opposite polarity. However, biphasic symmetrical stimulus waveforms result in less selectivity than monophasic waveforms. Prolonged stimulation at an intensity well below threshold for histologically detectable neural damage may nonetheless induce pronounced and prolonged elevation of the electrical threshold of stimulated neurons.
REFERENCES
Bartlett JR, Doty RW, Lee BB, Negrao N, Overman WH, Jr. Deleterious effects of prolonged electrical excitation of striate cortex in macaques. Brain Behav Evol 1977;14:46-66. Canavero S, Bonicalzi V: Therapeutic extradural cortical stimulation for central and neuropathic pain: A review. Clin J Pain 2002; 18:48-55 Canavero S, Bonicalzi V: Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007, pp 27-36 Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, pp 223-232 Crago PE, Peckham PH, Mortimer JT, Van der Meulen JP. The choice of pulse duration for chronic electrical stimulation via surface, nerve, and intramuscular electrodes. Ann Biomed Eng 1974;2:252-264. Ebel H, Rust D, Tronnier V, Spies EH, Boker D, Kunze S: Chronic precentral stimulation in trigeminal neuropathic pain. Acta Neurochir 1996; 138:1300-1306. Gordon B, Lesser RP, Rance NE, et al. Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation. EEG Clin Neurophysiol 1990; 75: 371-377 Holsheimer J, Lefaucheur JP, Buitenweg JR, Goujon C, Nineb A, Nguyen JP. The role of intra-operative motor evoked potentials in the optimization of chronic cortical stimulation for the treatment of neuropathic pain. Clin Neurophysiol 2007; 118: 22872296 Hosomi K, Saitoh Y, Kishima H, Oshino S, Hirata M, Tani N, Shimokawa T, Yoshimine T: Electrical stimulation of primary motor cortex within the central sulcus for intractable neuropathic pain. Clin Neurophysiol 2008; 119: 993-1001 Kamada K., Sawamura Y., Takeuchi F., et al. Functional identification of the primary motor area by corticospinal tractography, Neurosurgery Suppl.1 2005; 56: 98-109 Katayama Y, Fukaya C, Yamamoto T. Poststroke pain control by chronic motor cortex stimulation: neurological characteristics predicting a favorable response. J Neurosurg 1998;89:585-91. Kleiner-Fisman G, Fisman DN, Kahn FI, Sime E, Lozano A, Lang AE. Motor cortical stimulation for Parkinsonism in Multiple Systemic Atrophy Arch Neurol 2003; 60:15541558 Lehman RM, Kim HI. Partial seizures with onset in central area: Use of the callosal grid system for localization. Acta Neurochir 1995;64:79-82.
31
McCarthy G., Allison T., Spencer D.D. Localization of the face area of human sensorimotor cortex by intracranial recording of somatosensory evoked potentials, J Neurosurg 1993; 6:874-884 McCreery DB, Agnew WF, Yuen TG, Bullara L. Charge density and charge per phase as cofactors in neural injury induced by electrical stimulation. IEEE Trans Biomed Eng 1990; 37, 996-1001 Maegaki Y., Najm I., Terada K., et al. Somatosensory evoked high-frequency oscillations recorded directly from the human cerebral cortex. Clin Neurophysiol 2000; 111:19161926 Meyerson BA, Lindblom U, Linderoth B, Lind G, Herregodts P. Motor cortex stimulation as treatment of trigeminal neuropathic pain. Acta Neurochir 1993; 58:150-153. Montes C, Mertens P, Convers P, et al. Cognitive effects of precentral cortical stimulation for pain control: an ERP study. Clinical Neurophysiology 2002;32: 313-325 Naidich TP, Valvanis AG, Kubik S: Anatomical relationships along the low-middle convexity: part 1: normal specimens and magnetic resonance imaging. Neurosurgery 1995; 36: 517-532 Naidich TP, Blum JT, Firestone MI. The parasagittal line: an anatomic landmark for axial imaging. AJNR 2001; 22:885-895 Nguyen JP, Lefaucheur JP, Decq P, et al. Chronic motor cortex stimulation in the treatment of central and neuropathic pain. Correlations between clinical, electrophysiological and anatomical data. Pain 1999;82:245-51. Nuti C, Peyron R, Garcia-Larrea L, Brunon J, Laurent B, Sindou M, Mertens P: Motor cortex stimulation for refractory neuropathic pain: Four year outcome and predictors of efficacy. Pain 2005; 118: 43-52. Penfield W , Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation Brain 1937; 60:389-443 Pirotte B, Neugroschl C, Metens T, et al. Comparison of Functional MRI-Guidance to Electrical Cortical Mapping for Targeting Selective Motor Cortex Areas in Neuropathic Pain: A Study Based on Intraoperative Stereotactic Navigation. AJNR Am. J. Neuroradiol. 2005; 26:2256-2266. Polikov VS, Tresco PA, Reichert WM. Response of brain tissue to chronically implanted neural electrodes. J Neurosci Methods 2005; 148, 1-18. Pudenz RH, Bullara LA, Jacques S, Hambrecht FT. Electrical stimulation of the brain. III. The neural damage model. Surg Neurol 1975; 4: 389-400 Pudenz RH, Agnew WF, Bullara LA. Effects of electrical stimulation of brain. Brain Behav Evol 1977; 14: 103-125 Rasche D, Ruppolt M, Strippich C, Unterberg A, Tronnier VM. Motor cortex stimulation for long-term relief of chronic neuropathic pain: A 10 year experience. Pain 2006; 121: 4352 Saitoh Y, Shibata M, Hirano S, Hirata M, Mashimo T, Yoshimine T: Motor cortex stimulation for the central and the peripheral deafferentation pain. J Neurosurg 2000; 92:150-155
32
Saitoh Y, Yoshimine T. Stimulation of primary motor cortex for intractable deafferentation pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2 Wien: Springer-Verlag, 2007, pp. 51-56. Schmid D, Ebeling U, reulen HJ. Electrophysiological localization of the human sensorimotor cortex. J Neurosurg 1980; 70; 817-818 Stippich C, Romanovsky A, Nennig E, Kress B, Haehnel S, Sartor K. Fully automated localization of the human primary somatosensory cortex in one minute by functional magnetic resonance imaging. Neurosci Lett 2004; 364: 90-93 Strafella AP, Lozano AM, Lang AE, Ko JH, Poon Y-Y, Moro E. Subdural motor cortex stimulation in Parkinsons disease does not modify movement-related rCBF pattern. Mov Disorders 2007; 22: 2113-2116. Takahashi N, Kawamura M, Araki S: Isolation hand palsy due to cortical infarction: localization of the motor hand area. Neurology 2002; 58:1412-1414. Tehovnik EJ. Electrical stimulation of neural tissue to evoke behavioral responses. J Neurosci Methods 1996;65:1-17. Thickbroom GW, Byrnes ML, Walters S, Stell R, Mastaglia FL. Motor cortex reorganisation in Parkinson's disease. J Clin Neurosci 2006; 13: 639-42 Tirakotai W, Hellwig D, Bertalanffy H, Riegel T: Localization of precentral gyrus in imageguided surgery for motor cortex stimulation. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2 Wien: Springer-Verlag, 2007, pp. 75-79 Tomasino B, Budai R, Mondani M, Skrap M, Rumiati RI. Mental rotation in a patient with an implanted electrode grid in the motor cortex. Neuroreport 2005; 16: 1795-1800 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation of the treatment of central pain. Acta Neurochir Suppl (Wien) 1991;52:137139. Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation in patients with thalamic pain. J Neurosurg 1993;78:393-401. Velasco M, Velasco F, BritoF, Velasco AL, Nguyen JP, Marquez I, Boleaga B, Keravel Y: Motor cortex stimulation in the treatment of deafferentation pain. 1. Localization of the motor cortex. Stereotact Funct Neurosurg 2002;79:146-167. Yamamoto T, Katayama Y, Obuchi T, et al. Recording of corticospinal evoked potentials for optimum placement of motor cortex stimulation electrodes in the treatment of post-stroke pain. Neurol Med Chir (Tokyo) 2007;47:409-414. White LE, Andrewa TJ, Hulette C, Richards A, Groelle M, Paydarfar J, et al: Structure of the human sensorimotor system: I. Morphology and cytoarchitecture of the central sulcus. Cereb Cortex 1997;7:18-30 Wood CC, Spencer DD, Allison T, McCarthy G, Williamson PD, Goff WR: Localization of human sensorimotor cortex during surgery by cortical surface recording of somatosensory evoked potentials. J Neurosurg 1988; 68: 99-111. Woolsey CN, Erickson T, Gilson WE. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J Neurosurg 1979; 51:476-506.
Chapter 3
PRINCIPLES OF SURGICAL IMPLANTATION AND COMPLICATION AVOIDANCE

Benoit JM Pirotte1,, Philippe Voordecker1, Marc Levivier1 and Danielle Baleriaux2
1
Departments of Neurosurgery and 2Neuroradiology, Hpital Erasme, Universit Libre de Bruxelles, Bruxelles, Belgium.
INTRODUCTION
ECS remains a technique under evaluation and a definitive surgical protocol has not been established yet, not least because of a basic lack of understanding of how it works. Since the first report in 1991, the surgical procedure has known successive adaptations to improve the accuracy of the technique and to increase its success rate.
PRINCIPLES OF IMPLANTATION
ECS is generally accomplished in two steps. The procedure starts with a first surgical procedure to implant the ECS stimulating array, which is connected subcutaneously in the retromastoid region (and not inferiorly in the soft tissues of the neck, as the connector may break or erode through the skin with time) to a temporary cable that exits through the skin behind the ear (by way of a small incision) and is then plugged into an external stimulator. A 2-4 week test-period follows, during which the external stimulator, programmed by the attending clinician, provides electrical pulsed waves to the implanted electrodes in order to test different stimulation conditions and combinations and to search for the highest and more reproducible therapeutic effect. Whatever the results, the temporary cable exiting through the
Correspondence concerning this article should be addressed to: Dr. Benoit Pirotte, M.D., Ph.D., e-mail : bpirotte@ulb.ac.be.
34
Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.
skin is cut in order to allow the skin hole to heal properly. Then, in good and excellent responders, a second surgical procedure is performed under general anesthesia to implant the pulse generator (IPG) for chronic stimulation, according to the same electrical combinations and conditions that induced benefit in the test-period. Clinical practice has showed that ECS efficacy strongly depends on electrode positioning, i.e. the stimulating strip must overlie the somatotopic projection of the appropriate segment on the cortical motor strip. Inaccurate positioning of the electrode is the first issue in non-responders. This step is paramount and must be confirmed prior to fixing the electrode to the dura mater or otherwise (chapter 2). However, although functional areas corresponding to the face and the hand have a large cortical representation, functional reorganization in the primary cortical areas might occur both in chronic neurogenic pain and movement disorders, so that anatomical landmarks may have to be integrated with functional targeting methods (see below and chapter 2). As highlighted in chapter 2, the procedure can be performed under local anesthesia with a one- two burr holes or with flap craniotomy under general anesthesia. In this second approach, the patients head is secured to a head-holder clamp for neuronavigation and a large skin incision is performed after hair shaving; a 4-hour-long surgical procedure is usually necessary for a complete functional mapping procedure, although other authors do it in much less. In our experience, severely disabled patients are much more concerned by the efficacy of ECS than by the cosmetic issues related to skin incision, hair shaving or size of the craniotomy.
Figure 1. Stereotactic planning in different views for targeting the fMRI activation areas on the left motor strip after motor tasks of the right finger and upper lip.
Principles of Surgical Implantation and Complication Avoidance A-Surgical procedure (Figures 1-3)
35
STEP 1. Operative Position and Craniotomy Our preferred position is the lateral decubitus (Park bench) allowing horizontal orientation of the operative field and easy access to the contralateral upper limb under the table (Pirotte et al 2005, 2008). We recommend a 4x4 cm square craniotomy rather than a burr hole. However, several authors prefer this latter approach (Canavero and Bonicalzi 2002, 2007a,b). The craniotomy is centered on the cortical projection area of the painful somatic segment (parasagittal for lower limb pain; temporofrontal for facial pain) or, in case of movement disorders, contralateral to the worst affected side.
Figure 2. Per-operative 3-D reconstructed CT view of the right hemisphere in a patient suffering from left hemi-facial pain from trigeminal neuropathy. Data from the pre-operative fMRI study are combined with those from the intra-operative cortical brain mapping (iBM). The sensory-motor electrical dipoles corresponding to the central sulcus (white line) and to the left hand and tongue recorded by SSEPs (black crosses and lines) are compared spatially with the targets corresponding to the highest significance (white areas and crosses) of the activation area on fMRI (discontinued black lines) after motor paradigms of the left hand and tongue.
STEP 2. Intra-operative Cortical Brain Mapping (iBM) IBM of the primary sensorimotor cortex [including intraoperative recording of somatosensory evoked potentials (iSEP) and bipolar stimulo-detection (iBS)] is used as the most accurate method to localize the CS and the functional target to stimulate on the motor cortex (Pirotte et al 2005, 2008). Such iBM-guided MCS procedures combine iBM recordings and epidural fixation of a stimulation electrode in a single procedure. Some authors (see chapter 2) implant a grid for cortical brain mapping, record SEP for one week and fix the stimulation electrode in a separate surgical procedure. For iBM, a grid (or the
36
electrode used for ECS) is placed at different locations on the dural surface over the central sulcus (CS) region. The coordinates of every iSEP recording contact covering the CS region are registered in the navigation workstation. Caveats are in order. In the context of marked denervation, iBM may show wave attenuation, diffuse motor responses, increased sensitivity to electrical artifacts or lack of reproducibility due to lesion of the somatosensory tracts. In a personal series of 18 patients, iBM was highly accurate in localizing the functional target in 50% of the cases, provided an approximative target in 3/18 and a non-reproducible target in 6/18 cases (Pirotte et al 2005, 2008). Moreover, even in physiological conditions, the facial area, although largely represented on the pre-central gyrus, is not accurately identified by iBM. Finally, in amputees or in patients with brachial plexus lesions, iBM may be simply unfeasible because the limb is missing or has degenerated. The CS is defined by means of the N20/P30 wave phase reversal (confirmed on 3 repeated recordings), as described in chapter 2. The motor target of the hand is thereafter determined by the wave showing the highest amplitude after stimulation at the median nerve. iSEP after facial stimulation are used for facial pain. Although peripheral stimulation of median, tibial or trigeminal nerves are efficient for studying each segment, a systematic median nerve stimulation is recommended in all pain distribution patterns because the highly reproducible hand target obtained can be helpful as a reference target when recordings after lower limb or facial stimulation give ambiguous data (Pirotte et al 2005, 2008). The location of the motor target can be confirmed by iBS through the stimulation electrode (5 mm space tips bipolar stimulator probe; isolated square-wave pulses with a duration of 1 ms; 60 Hz; from 5 to 20 mA)(Pirotte et al 2005, 2008). One limitation of the ECS technique is the cortical projection of the lower limb on the interhemispheric portion of the motor strip. This anatomical arrangement increases the distance between the electrode fixed on the convexity and may reduce the efficacy of ECS. On the other hand, subdural implantation in the interhemispheric fissure is not safe on a routine basis (chapter 2). In practice, the functional target corresponding to the lower limb in iBM (confirmed by fMRI) is projected on the para-sagittal convexity in more than one third of the patients (Canavero et al 2002). As regards possible mistargeting due to plastic rearrangements, as seen in phantom pain, no guidelines can be offered, due to lack of correlative data. However, it should probably not represent a major source of mistargeting, at least for central pain (Roux et al 2001; see also discussion in Canavero and Bonicalzi 2007c, who downplay neuroplastic changes in central pain). STEP 3. Epidural Electrode Fixation and Closure Once the target identified on the motor strip, the stimulation electrode is fixed epidurally, parallel to CS (Canavero and Bonicalzi 2007,a,b) or perpendicular to the CS with 3 poles anterior to the CS. Very careful hemostasis is highly recommended for avoiding epidural infectious or hemorrhagic complications. Superficial denervation of the dura also allows to reduce the local (cranial) pain described by many patients when switching the stimulation on. Initially, we opened the dura with scissors and resutured it with 3-0 silk. Later on, we denervated the dura by bipolar coagulation on its surface. The electrode must be fixed to the dura with very tight sutures to minimize the space between the electrode and the dura, in which scar tissue can develop and thereby increase the impedance of the stimulation and require higher voltage (Canavero and Bonicalzi 2007a,b). We usually fix the electrode by
Principles of Surgical Implantation and Complication Avoidance
37
means of a single 3-0 silk suture every 5 mm, with 2 sutures bridging the electrode in order to maintain it against the dura. We recommend to drill a groove at the posterior margin of the bone flap, to avoid crushing the wires after bone flat fixation to the skull. A drain can- or cannot- be inserted in the epidural space or in the subcutaneous space to avoid hematomas, which may delay wound healing. The skin is closed with separated 2-0 Dermalon sutures. We avoid the use of staples because they can damage the wires and be a source of electrical failure, but others use them regularly
Figure 3. Summary of the 3 different situations encountered when combining per-operative data from iBM and fMRI for the hand. In about 60% of the cases (figure 3A), the iBM-defined motor target of the hand is located within the motor fMRI activation area for standard values (p<0.001), but also for more restrictive values (p<0.00001). In about 30% of the cases (figure 3B), the iBM-defined target is not strictly reproducible (repeated recordings provide different targets among which one corresponds to the fMRI activation area). In about 5% of the cases (figure 3C), the iBM-defined target corresponds to the fMRI activation area, but for less restrictive values than usual (p<0.01).
38
STEP 4. Clinical Observations after MCS On the first operative day, plain x-rays and a head CT scan are performed to confirm the position of the lead and to rule out any complications, such as epidural bleeding. The surgical procedure is followed by a test-period of 1-4 weeks before considering implantation of the IPG for long-term stimulation. During this interval, if the patient is a chronic pain sufferer, he or she is stimulated during sessions lasting one to four hours every 4 hours (monophasic square wave pulses; frequency 40 to 120 Hz; pulse duration 100 to 400 s; amplitude 1 to 5V), with many bipolar combinations; the negative pole is generally over MI, but a positive pole can also be set. It is rarely useful to stimulate above 6V, as this increases the risk of inducing a motor response or even a seizure; however, cases of benefit at 8-10V are on record (Canavero and Bonicalzi 2002, 2007a,b). Usually, the therapeutic effect appears at one third (2-5V) of the motor threshold (7-13V). Excellent and good responders are implanted with a subcutaneous stimulator for long-term stimulation. For movement disorders, the protocol is detailed in chapter 13b. In our experience, ECS induces pain relief within 10-15 minutes of the start of stimulation and lasting from 15 to 120 minutes after the stimulation is switched off (Pirotte et al 2005). In the first 3 days after implantation, the pain often disappears or is strongly attenuated by general anesthesia. This is the reason why we usually wait for 3 days before starting the test run (i.e. patients are scheduled for surgery on Thursday, stay 24 hours in the intensive care unit and rest in their room during the week-end until Monday morning, when stimulation begins) or at least wait until the pain has returned to its disabling level. Some patients report that pain relief can remain stable for more than 24 hours, if the stimulation period is longer than 4 hours (so-called after-effect: Canavero et al 2002, 2007a,b). Other patients describe that severe pain recurs when they switch the stimulation off for more than 2 days. The goal of the test-period is to test all parameters during a short period of time, in order to decrease the risk of infection due to the presence of transcutaneous wires. Pain intensity is measured using a visual analog scale (VAS) by specially trained nurses or physicians; a pain diary has also to be completed by the patient. During the test, the bipolar combinations for which an analgesic effect has been observed must be re-tested many times in a single-blind (blinded patient) manner in order to exclude a placebo effect (Carroll et al 2000, Rasche et al 2006). This last step is crucial to consider the test as positive. Indeed, many combinations may appear as good only once, without being confirmed at further stimulations. Blinding is attained by covering and shielding the external stimulation device in the area of the lead combinations. Continuous stimulation is used during the test trial, but after IPG implantation, intermittent stimulation is programmed to avoid habituation, at least initially. In practice, a reproducible analgesic effect is often observed within the first 10 days of stimulation. In several cases, the coupling configuration is critical. Some patients mention a transient painful sensation (reported as a painful click) centered on the craniotomy when stimulation is switched on, probably due to a direct stimulation of the dura. Electrical combinations that stimulate the parietal cortex, but also MI (Canavero and Bonicalzi 2002, 2007a), may increase burning or tingling sensations. We recommend that the attending clinician bears in mind the following: 1) neurogenic pain is highly fluctuating from one day to another; 2) the analgesic effect from ECS must show reproducibility to be considered as valid;
39
3) patients tested in a hospital do not behave like at home; 4) pain patients have such huge expectations that they strongly and very quickly jump to the conclusion that a slight variation in the pain level is predicting a successful outcome.
B-Improving Functional Targeting At least some failures may relate to inaccurate electrode positioning. Therefore, it is fair to say that the actual efficacy of ECS might be underestimated. Targeting can be improved in several ways. 1-Awake Surgery Awaking patients during the surgical procedure might significantly improve the quality of iBM by increasing the amplitude of evoked potentials, by improving the reproducibility of electrophysiological data and by reducing the sensitivity of iBM to electrical artifacts. Some authors, more simply, perform the procedure under local anesthesia and neuroleptoanalgesia (Canavero and Bonicalzi 2002, 2007a). In Canaveros experience, uncooperative patients are infrequent and most patients find it that having no screens in front of the eyes, including surgical cloths, makes the procedure much more tolerable (Canavero S, personal communication). In this case, the surgical field is prepared by folding the surgical sterile cloth on the patients forehead and hanging it to nearby posts providing unrestricted view of the surrounding environment. 2-Combining Independent Functional Targeting Techniques Neuronavigation allows integration of different targeting methods, including 3-D MRI (Herregodts et al 1995), fMRI, MEG and others. fMRI guidance can provide independent and highly accurate targeting information for improving the surgical procedure. We combined fMRI guidance with iBM (Pirotte et al 2005). IBM-guided ECS was performed under stereotactic image-guidance by using a frameless neuronavigation system and the data obtained by iBM and fMRI were compared intra-operatively. Correspondence between contours of fMRI activation area and iBM in the precentral gyrus was found in almost all (94%) patients. Moreover, fMRI revealed to be less altered by artifacts and to provide data which were more constantly unambiguous than those from iBM. fMRI was also possible in amputated patients in which iBM is not feasible. Unfortunately, fMRI is a time consuming technique requiring training and engineering (statistical parametric mapping, spatial and functional validation) prior to its application to stereotactic conditions. FMRI can be combined with MR tractography and diffusion tensor imaging (DTI) for optimal brain damage assessment, especially in cases of stroke (chapter 15). MEG has an excellent temporal resolution (1 second, real time acquisition) which is much better than that of fMRI (30 min, summation acquisition). Moreover, MEG records directly the magnetic field created by the electrical neuronal activity, while fMRI activation signals reflect indirectly transient changes in the rCBF secondary to the neuronal activity. We are currently assessing its possible value as applied to ECS.
40
C-Optimizing Cortical Stimulation 1-Electrode Orientation Two schools exist regarding electrode orientation: perpendicular versus parallel to the Rolandic fissure. Until a formal prospective study is conducted, it can only be stressed that results as reported in the literature are not different using either of these approaches (Canavero and Bonicalzi 2002, 2007a), as far as neurogenic pain is concerned. In movement disorders, electrodes are generally placed parallel to the fissure (Canavero and Bonicalzi 2007b). 2-Subdural Positioning Opening the dura might present different advantages, but is a more invasive approach. First, subdural iBM recordings provide more precise functional targeting by increasing the iSEP wave amplitude. Secondly, subdural electrode positioning allows reduction of the intensity of cortical stimulation. Thirdly, direct observation of the cortical gyri allows better control of the location of the CS by direct correlation with navigation images. However, since this approach does not increase the techniques success rate, but only the risk of seizures induced by stimulation, and might generate scar tissue, cortical lesioning with secondary gliosis and infectious empyema, most authorities discourage its use (chapter 2 and Canavero and Bonicalzi 2007a). 3-Large Electrode Design Using double or multiple electrodes (side by side) might increase the cortical surface to be stimulated and the chances to obtain a therapeutic effect. However, a tight contact with the dura might not be guaranteed in the center of the electrode. 4-Stimulation of the Healthy or Ipsilateral (to Symptoms) Hemisphere As regards central pain, a careful analysis of available data strongly suggest that the generator can be switched contralaterally and that, in ECS failures, the healthy hemisphere should be targeted (Canavero and Bonicalzi 2007c). As for stroke, the reader is referred to chapter 15. Bilateral stimulation for movement disorders is discussed in chapter 13.
PRINCIPLES OF COMPLICATIONS AVOIDANCE

Complications are possible in the setting of ECS. The expert surgeon must know them and anticipate their onset.
1-Infection As classically observed in all surgical disciplines, the risk of infection is increased by the presence of implanted materials. Infection can occur as a primary process, directly related to the electrode implantation and rapidly presents as epidural or subdural (if the dura has been
41
opened see above) empyema. Infection can also start at the cutaneous exit of the electrode. Finally, infection of the electrode may occur secondarily to incomplete healing of the skin at the level of the scar on the vertex (wound dehiscence). Erosion of the connector through the skin is possible. Placement of the connector in the soft tissues of the neck should be avoided, given the stress related to cervical motion and the connectors tendency to erode through the skin or break. In practice, the infection rate is not high. In our experience with about 50 ECS procedures so far, we experienced infections in 3 cases only (1 primary epidural empyema 15 days after surgery, 1 subdural empyema 10 days after surgery and 1 epidural empyema with cortical abscess secondary to incomplete healing of the scar 6 weeks after surgery ). Despite excellent analgesia, these 3 cases required surgical removal of the entire stimulating system, followed by 4 weeks of intravenous antibiotics. In order to avoid primary infection of the electrodes, aseptic measures must be carefully enacted throughout the procedure. Double gloving, avoiding implants manipulations, as well as systematic use of 24-hour-prophylactic intravenous antibiotics are mandatory. We (but not most other authors) insert systematically a drain in the epidural or subcutaneous space in order to reduce the risk of hematoma and subsequent inflammation or local infection. If epidural or subdural infection occurs, a cycle of intravenous antibiotics can be attempted, but, in almost all cases, explantation of the whole system is the only effective way to achieve remission. Local cutaneous infection at the exit of the temporary cables behind the ear is benign and rather infrequent, in spite of the cables lying externally for weeks (several centers keep the patients on antibiotics as long as the connection cables are externalized, though). Daily application of an antiseptic (e.g. Betadine) at the local site is usually sufficient to control local inflammation and impede progression of the infection to the epidural space. Secondary infection of the scar due to incomplete healing of the wound at the convexity should not be underestimated. It represents a major source of electrode infection that can lead to forced explantation. Indeed, an incomplete healing of the wound can contaminate the subcutaneous space, where the electrode wires can be affected. We experienced one case of secondary epidural infection of the electrodes complicated by a cortical abscess. Another patient developed an epidural empyema secondary to suboptimal wound healing and a cortical abscess just below the electrode. This abscess was visible on CT and MRI and lasted for 6 months before disappearing. During the 6-week-infection phase, this 30-year-old female developed partial motor seizures (motor strip involved), had transient motor worsening and complained of increased pain sensation, despite analgesia before infection. This case convinced us to modify our policy of skin incision. We took special care to draw a large skin incision on the convexity so that the wound would not overlie any bone fixation or any wire or material, the skin incision being far away from the edge of the craniotomy. Finally, skin closure must be technically perfect.
2-Induced Seizures /Involuntary Movements Stimulating the patient with bipolar stimulation as described above carries almost no risk of inducing involuntary movements. Stimulating above 9V, however, increases the risk of inducing a motor response or even a seizure (variously described as brief focal seizures during programming, prolonged focal seizures with postictal speech arrest, brief generalized
42
seizures during programming and generalized seizures after activating the stimulator). Practically, this occurs very rarely. However, a seizure, even short, can induce neuronal suffering, generalize into a grand mal crisis and even create an epileptic focus (only 2 cases of a single fit are on record after long-term stimulation). This is the reason why we avoid stimulation during one week in case of one seizure during the test-period. To minimize the risk of seizures, some centers put the patients on anticonvulsants perioperatively and for a while postoperatively (ranging from 2 weeks to 6 months). Should a seizure occur intraoperatively during the iBM, we flush the cortex with cold saline.
3-Increased Pain Sensation During the test-period, patients often complain that the pain sensation is enhanced and is associated with an increased sensation of burning or tingling in the painful limb. When analyzing the bipolar combinations tested, this observation was constantly correlated with a cathodic stimulation placed above the parietal cortex. However, parietal cortex stimulation with a parallel electrode can be analgesic in central pain (chapter 9 and Canavero and Bonicalzi 2002, 2007a).
4-Loss of Effect Loss of effect after an initial excellent analgesia has been reported by many investigators (Canavero and Bonicalzi 2002, 2007a), and we too experienced two cases. We noticed that this occurred in the course of continuous (rather than cyclical: e.g. 4 hours ON, then 4 hours OFF) mode stimulation. Changing to a cycling mode made burning sensations disappear within 2 hours and restored analgesia within a few hours. Although PW and frequency may be increased as follow-up progresses, particularly in cases in which effects dwindle, the efficacy of stimulation changes very little with such incremental increases. On the other hand, some authors find that loss of efficacy may be due to keeping parameter settings constant, suggesting that perhaps longer-term plastic changes eventually catch up with ECS to limit its efficacy. Loss of effect can also be due to extensive fibrosis below and around the contacts: removal thereof may restore benefit (Canavero and Bonicalzi 2002, Velasco et al 2008).
5-Electrode and Cable Damage Special attention and skill are required for the surgical implantation and connection of the different electrodes and cables. After anchoring the electrode to the dura, the surgeon must avoid to kink the cables while fixating the bone flap or perforating the tiny metallic contacts and wires into the electrode. Similarly, during re-opening of the supra-mastoid skin at the connection level in order to withdraw the temporary cable left in place and to connect the extension cable to the extremity of the ECS electrode, great care must be exercised not to damage the components and to make the connections watertight by using the cap that is
43
provided. Any kinking, perforation, or suture defect can lead to loss of power and to ECS inefficacy. In non-responders, technical failure must be immediately entertained. In all cases, the impedance (electrical resistance) of each electrode contacts and cable must be tested. If any disconnection or loss of power occurred, the recorded impedance spikes or drops well above 600 . The first step to verify the system for hardware breakage is to perform a standard X-ray of the whole implanted system.
6-Local Hyperesthesic Phenomena About 20 % of the patients, report painful sensations either at the scalp around the bone flap - or at the supramastoid region where the connection is located - or around the neurostimulator. Visual inspection, as well as CT, X-rays or MRI usually do not show anything wrong. Two of our patients reported a painful sensation along the subcutaneous trajectory of the extension cable in the neck as if it was stretched like a rope in rotatory movements of the head. Indeed, a fibrous reaction can occur around the extension cable, so that we recommend not to tunnel it too superficially in the neck in order for the platysma muscle to cover it completely. It should be borne in mind that normal sensory perception is generally disturbed in neurogenic pain patients, even in places located far from the painful limb or segment. This may cause minor but chronic and sometimes disabling complaints centered on the sites of implantation of the electrodes, cables or stimulators.
7-Comfort Issues Some patients complain of the site chosen to implant the pulse generator subcutaneously, generally the subclavicular or at the paraumbilical areas. This issue should be discussed with the patient prior to the surgical procedure. The paraumbilical region can be very convenient in fat persons, but sometimes painful in skinny patients because of frictions with trousers, belts or other clothes. Some young ladies may not accept subclavicular placement of the IPG for esthetic reasons. We have in some instances placed the stimulator 5 cm lower and 10 cm more laterally with a vertical incision above the breast, just on the pectoral muscle, with great satisfaction and no complaints so far.
REFERENCES
Canavero S, Bonicalzi V: Therapeutic extradural cortical stimulation for central and neuropathic pain: A review. Clin J Pain 2002; 18:48-55, 2002. Canavero S, Bonicalzi V: Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007a, pp 27-36.
44
Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, pp 223-232. Canavero S, Bonicalzi V. Central pain syndrome. Pathophysiology, diagnosis and management. New York: Cambridge University Press, 2007. Carroll D, Joint C, Maartens N, Shlugman D, Stein J, Aziz TZ. Motor cortex stimulation for chronic neuropathic pain: a preliminary study of 10 cases. Pain 2000; 84:431-437. Herregodts P, Stadnik T, De Ridder F, D'Haens J. Cortical stimulation for central neuropathic pain: 3-D surface MRI for easy determination of the motor cortex. Acta Neurochir Suppl (Wien) 1995; 64:132-135. Katayama Y, Fukaya C, Yamamoto T. Poststroke pain control by chronic motor cortex stimulation: neurological characteristics predicting a favorable response. J. Neurosurg. 1998; 89:585-591. Pirotte B, Neugroschl C, Metens T, et al. Comparison of Functional MRI-Guidance to Electrical Cortical Mapping for Targeting Selective Motor Cortex Areas in Neuropathic Pain: A Study Based on Intraoperative Stereotactic Navigation. AJNR Am. J. Neuroradiol. 2005; 26:2256-2266. Pirotte B, Voordecker P, Neugroschl C, et al. Combination of functional magnetic resonance imaging-guided neuronavigation and intraoperative cortical brain mapping improves targeting of motor cortex stimulation in neuropathic pain. Neurosurgery. 2008;62(6 Suppl 3):SHC941-56. Rasche D, Ruppolt M, Strippich C, Unterberg A, Tronnier VM. Motor cortex stimulation for long-term relief of chronic neuropathic pain: A 10 year experience. Pain 2006; 121: 4352. Roux FE, Ibarrola D, Tremoulet M, et al. Methodological and technical issues for integrating functional magnetic resonance imaging data in a neuronavigational system. Neurosurgery 2001; 49:1145-1156. Velasco F, Arguelles C, Carrillo-Ruiz JD, et al. Efficacy of motor cortex stimulation in the treatment of neuropathic pain: a randomized double-blind trial. J Neurosurg 2008; 108: 698-706.
Chapter 4
TRANSCRANIAL MAGNETIC AND DIRECT CURRENT STIMULATION: A PRIMER

Filippo Cogiamanian, Sara Marceglia, Lorenzo Rossi, Ernesto Della Torre and Alberto Priori
Clinical Center for Neuronanotechnology and Neurostimulation, Fondazione IRCCS Ospedale Maggiore Policlinico, Mangiagalli e Regina Elena, Dipartimento di Scienze Neurologiche, Universit di Milano, Italy.
INTRODUCTION
Transcranial magnetic (TMS) and transcranial direct current stimulation (tDCS) are non-invasive methods to stimulate the brain and to modulate its function (Priori 2003; Ridding and Rothwell 2007). In the past two decades, TMS and tDCS have provided a valuable tool for interventional neurophysiology applications and novel insights into the pathophysiology of neurological and psychiatric disorders (Fregni and Pascual-Leone, 2007; Sparing and Mottaghy, 2008). Moreover, TMS provides diagnostic and prognostic data. Non-invasive brain stimulation techniques have been proposed for the treatment of psychiatric disorders (chapter 17 and Boggio et al, 2008; Fregni et al, 2006), neurologic diseases (i.e. Parkinsons disease, dystonia, pain syndromes) (chapters 8 and 12) and neuro-rehabilitation (i.e. post-stroke recovery) (chapter 14). The aim of this chapter is to provide a primer on TMS and tDCS technology describing the physics and mechanisms of action as well as the basic design of devices and common stimulation techniques.
1A. TMS: BASIC PRINCIPLES

Electrical stimulation of the exposed human cortex has been introduced two centuries ago (Zago et al. 2008) and it is commonly used in neurosurgery to recognize the function of
46
Filippo Cogiamanian, Sara Marceglia, Lorenzo Rossi et al.
specific brain areas. Because of the high electrical resistance of the skull and of the scalp, electrical stimulation through the intact skull is however difficult. After the technique proposed by Gualtierotti (1953) that, though effective, was exceedingly painful, in the 80s Merton and Morton demonstrated that brief high-voltage electrical shocks from a special low-output-resistance stimulator, delivered through electrodes on the scalp, can directly excite the human motor cortex (Merton and Morton 1980; Merton et al. 1982). The pain elicited by electric currents delivered through the scalp has limited the clinical application of transcranial electric stimulation (TES) to some research applications (Mills et al. 1987, Maertens de Noordhout et al. 1989, Rothwell 1991). In 1985, Barker and colleagues showed that it was possible to stimulate brain using magnetic stimulation with no pain. TMS is based on the principle of electromagnetic induction discovered by Faraday in 1831: when a brief, high-current pulse flows in an insulated copper wire coil (encased in a plastic paddle-like wand), a magnetic field (up to about 2 Tesla) is induced (along with a popping sound) with lines of flux running perpendicularly to the plane of the coil. The magnetic field, penetrating the skull, reaches the brain (maximum range: 2-3 cm), where it induces an electric field able to excite neurons and axons 1/100000 as large as the primary current. The loops with the strongest current are near the outer circumference of the coil, whilst the current is weaker near the inner circumference of the coil, with no current at the center. Depending on stimulation parameters, TMS can up-regulate or down-regulate excitability to different extents in the neural structures under the stimulating coil. Because the skin, skull and meninges do not attenuate the magnetic field, the voltage needed to excite neuronal tissue can be reached without inducing pain. Conversely, because the intensity of an electric field is attenuated by the resistivity of these elements before reaching the cortex, the voltage applied outside the brain by TES is much greater and stimulates skin pain receptors. Moreover, because the intensity of the current induced by magnetic pulse is inversely proportional to the resistance of the tissue, the current induced in the high-resistance skin at the scalp is of low intensity and few (if any) pain receptors are activated (Ridding and Rothwell 2007).
1B. TDCS: BASIC PRINCIPLES

A constant flow of electric charge that does not change direction over time polarizes tissues. It is well known that weak direct current (DC) can change the excitability of human nerve through the skin (Kiernan and Bostock, 2000; Ardolino et al. 2005) and even stimulate the vestibular system through the skull (Day et al., 1997). tDCS refers to the application of a constant DC through a pair of surface electrodes (anode and cathode). Although the scalp has high impedance and, hence, most of the current applied is shunted and does not penetrate into the brain, the intracranial current flow is enough to produce changes in membrane resting thresholds beneath the electrode. Evidence that scalp DC can induce prolonged changes in brain excitability that persist long after their offset has been provided for motor (Priori et al. 1998; Nitsche and Paulus 2000, Nitsche and Paulus 2001), sensori-motor (Matsunaga et al. 2004), visual (Antal et al. 2001; Antal et al. 2004; Antal et al. 2006; Accornero et al. 2007), and cerebellar cortex (Ferrucci et al, 2008). Different technical issues, particularly the
Transcranial Magnetic and Direct Current Stimulation
47
orientation of the neural tissue with respect to the electrical field generated by stimulating electrodes, influence the net effect of DC (increase or decrease of cortical excitability) over the cerebral cortex. Whereas both tDCS and TES use an electrical source to influence the neural activity, the nature of the electrical stimulation differs. Commercially available stimulators for TES produce pulses up to 750 V with a fast rise and decay time constant of 50 to 100 sec. Conversely tDCS is based on time invariant constant amplitude low-intensity currents (1-2.5 mA) applied for long time intervals (5-20 minutes). TES uses smaller electrodes than tDCS scalp electrodes, leading to much larger stimulating current densities compared to tDCS (Priori 2003). Moreover, TES, as well as TMS, actively evokes action potentials from the underlying neural substrate (neurostimulation), whereas the application of tDCS alters the overall excitability of the neural tissue (neuromodulation). tDCS does not cause resting neurons to fire; it rather modulates the spontaneous firing rate of neurons by acting at the level of the membrane potential, thereby differentiating tDCS from other stimulation techniques- such as TMS/rTMS, conventional TES or electroconvulsive therapy- which excite neurons directly (Priori 2003). Whereas ample experimental data are available on brain polarization in animals, less is known about how weak DC applied through the scalp affects brain excitability in humans. The cellular mechanisms of DC stimulation were studied by Liebetanz et al (2002) who conducted experiments of pharmacological modulation of DC stimulation concluding that combination of glutamatergic mechanisms and of changes in membrane potentials are required for DC induced after-effects on cortical excitability. More recently, Ardolino et al (2005), evaluating tDCS-induced after-effects on the MEP amplitude evoked by TES, demonstrated that the after-effects of tDCS have a non-synaptic mechanism of action based upon changes in neural membrane function. They suggested that these changes apart from reflecting local changes in ionic concentrations, could arise from alterations in transmembrane proteins and from electrolysis-related changes in [H+] induced by exposure to constant electric field.
2A.TMS DEVICES AND STIMULATION PROTOCOLS

TMS stimulators consist of a capacitor discharged into an inductor (the stimulating coil) via an electric switch (thyristor). The energy for stimulation is derived from charging a bank of capacitors up to about 4 kV, which when discharged induces a current of up to 5,000 A that passes through the stimulating coil (consisting of one or more well insulated coils of copper) placed over the scalp, creating a brief magnetic field (pulsed fields of 14 Tesla in strength) peaking at about 160 sec (Ridding and Rothwell 2007); the pulse lasts about 1 ms. During TMS, the operator can control the intensity of the stimuli by changing the intensity of current flowing in the coil, thus changing the magnitude of the induced magnetic field and of the secondarily induced electrical field. Stimulus strength is usually expressed as a percentage of maximal output possible from the stimulator. In rTMS, a biphasic waveform is generally used. Unlike monophasic waves, in the biphasic mode, up to 60% of the original energy in the pulse is returned to the capacitor, rendering rTMS more energy efficient and
48
thus enabling the capacitors to recharge more quickly. The biphasic waveform also seems to require lower field intensities to induce a current in neural tissue. In fact, neurons are not perfect capacitors, they are leaky, and the quicker the rise to peak intensity of the magnetic field, the less time is available for the tissue to lose charge. Finally, a fast rise decreases the energy requirements of the stimulator and the heating of the coil. The second key hardware component of magnetic stimulators is the current carrying coil, which serves as the electromagnetic source during stimulation. Design of the coil is critically important because it is the only component that comes in direct contact with the subject undergoing stimulation. The coils shape directly influences the induced current distribution and, thus, the site and the focality of stimulation. Two types of TMS coils are commonly used for stimulation, the circular coil and the figure-of-8 coil. A circular coil activates a large volume of brain tissue. A figure-of-8 coil is composed of 2 adjacent circular coils that carry currents in opposite directions, with a summation field where they meet. The magnetic field of a figure-of-8 coil is focused at the point where the 2 circular coils meet, thereby providing stimulation to a focal area of brain tissue of approximately 1 cm3 in volume. Because currently available coils can induce very focal stimulation, the precise localization of the target area and of the coil position can be achieved by a frameless stereotactic system. These systems are based on the subjects head MRI data and the coil geometry and are able to digitally track the coil position relative to the subjects head registering the predicted stimulation location in MRI space, with an accuracy of 4-8 mm. Specificity of the area stimulated can be increased by making contact between only one arc of the coil and the scalp. A novel TMS device with controllable PW and near-rectangular pulse shape (cTMS) has been described. It uses an insulated gate bipolar transistor with appropriate snubbers to switch coil currents up to 6kA, enabling PW control from 5 to 100 s. The near-rectangular induced electric field pulses use 2-34% less energy and generate 67-72% less coil heating compared to matched conventional cosine pulses (Peterchev et al 2008). Magnetic impulses can be applied as single pulses every few seconds (single-pulse TMS), in pairs, with a 1-30 ms inter-pulse interval and >4 sec between pairs (paired-pulse TMS), and in series (repetitive TMS, rTMS). rTMS is characterized by the frequency of stimulation and can be divided into low-frequency or slow rTMS (<1 Hz) and high-frequency or rapid rTMS (>1 Hz, up to 50Hz).
A-Single-pulse TMS (sTMS) Most of our knowledge of the action of TMS on the human cortex comes from sTMSstudies of the primary motor cortex. Stimulation in this area evokes activity in muscles on the opposite side of the body, which is easily measured using electrophysiological methods (motor evoked potentials, MEPs) to investigate corticospinal tract properties and cortical excitability (i.e central motor conduction time, motor cortex threshold). By recordings of the descending corticospinal activity from the spinal cord, it has been demonstrated that a single magnetic stimulus over primary motor cortex evokes a burst of activity that lasts 5-10ms. This activity is thought to result from stimulation of intracortical
49
excitatory interneurons that gives rise to a series of discharges in the corticospinal tract that are known as I-waves (indirect waves) as they do not arise from direct stimulation of fastconducting pyramidal neurons (chapter 5). Differently from TES, sTMS can induce D-waves (direct waves from stimulation of the proximal axon of pyramidal neurons) only at high intensities and depending on the direction of the induced current. Motor responses evoked by TMS are followed by a long period (100200 ms), named silent period, during which the ongoing voluntary EMG activity is suppressed. The silent period duration linearly correlates to the stimulus intensity. Although the first part of the silent period is due in part to spinal cord refractoriness, the latter part is entirely due to cortical inhibition and it has been hypothesized that it is induced by a long-lasting GABA (-aminobutyric acid)-mediated inhibitory input that follows the initial discharge (Inghilleri et al. 1996, Werhahn et al. 1999).
B-Paired-pulse TMS (pTMS) These protocols are based on pairs of magnetic stimuli separated by a variable interstimulus interval (ISI). The interaction of a conditioning impulse and a second suprathreshold test impulse are evaluated. The net effect of the interaction depends on the intensity of the stimuli and on the ISI. Facilitatory effects of the conditioning TMS pulse on the test MEP can be observed at intervals of 720 ms using a subthreshold conditioning stimulus. Since the administration of an NMDA receptor antagonist reduces intracortical facilitation (IF) (Ziemann et al. 1998), glutamate is probably involved in mediating this effect. Conversely, two types of intracortical inhibition have been studied by paired TMS: short-interval intracortical inhibition (SICI) and long-interval intracortical inhibition (LICI). As of SICI, maximum inhibitory effects are found at short interstimulus intervals of 14 ms and conditioning stimuli of 6080% of the resting motor threshold. LICI is elicited by suprathreshold conditioning and test pulses applied at ISIs of approximately 50200 ms (Valls-Sol et al. 1992). Both SICI and LICI are mediated by cortical inhibitory interneurons and several lines of evidence suggest that they are related to different subtypes of GABAergic receptors: SICI may be mediated by GABAA receptors, LICI by GABAB receptors. The paired-pulse technique has been extensively used to investigate interhemispheric interactions. During transcallosal inhibition protocols a test stimulus is given over one hemisphere and a conditioning stimulus over the other. When the conditioning stimulus preceded the test stimulus by 5-30 ms, the response to the test stimulus is reduced. The inhibitory effect is directly affected by the intensity of the conditioning stimulus. Axons mediating interhemisperic inhibition arise from contralateral cortex and travel to the opposite hemisphere to exert their inhibitory effects by activating local inhibitory neurons which are shared with LICI (Daskalakis et al. 2002).
C-Repetitive TMS (rTMS) rTMS stimulates the brain with a series of magnetic pulses. The higher the stimulation frequency and intensity, the greater is the disruption of cortical function during the train of
50
stimulation. Moreover, repetitive magnetic pulses can induce powerful effects on cortical excitability that outlast the period of stimulation. This was first demonstrated in the motor system. In brief, slow rTMS (i.e. 1 Hz or less) can suppress excitability of MI (Chen et al. 1997), while rapid rTMS (5Hz or higher) will transiently enhance motor excitability, particularly at high intensities of stimulation (Pascual-Leone et al. 1994, Maeda et al 2000). The lasting modulation of cortical activity by rTMS is not limited to motor cortical areas. There is also evidence that these long-lasting effects of rTMS can be induced in areas outside MI and be associated with measurable behavioral effects at visual (Kosslyn et al. 1999), prefrontal (Mottaghy et al. 2002), and parietal cortical levels (Hilgetag et al. 2001), as well as in the cerebellum (Theoret et al. 2001). After-effects can last seconds (Berardelli et al 1998) to about 1 hour (Maeda et al 2000), but depends on some stimulation features, such as the number of pulses applied, the rate of application and the intensity of each stimulus. Aftereffects also depend on the pattern of the pulses applied. For example, Huang et al (2005) use a pattern of stimulation named theta burst stimulation (TBS) in which three 50Hz pulses are applied regularly 5 times per second at low intensities. In intermittent theta burst stimulation (iTBS), a 2 s train of TBS is repeated every 10 s for a total of 190 s (600 pulses). In continuous theta burst stimulation (cTBS), a 40 s train of uninterrupted TBS is given (600 pulses). cTBS yields a suppression of MI excitability, whereas iTBS is associated with an increase of MEP amplitudes (chapter 5). The mechanisms of the long-lasting effects of rTMS on cortical excitability are still unclear. Long-term potentiation and depression of cortical synapses or closely related neuronal mechanisms have been suggested as possible mechanisms to explain the effects of high and low-frequency rTMS, respectively. Animal studies suggest that modulation of neurotransmitters and gene induction may contribute to these long-lasting modulatory effects of rTMS (chapter 5). Application of rTMS results in skin sensations, or even muscle twitches, and a popping sound that require sham controls. Sham techniques include specially-designed coils producing noise without delivering any magnetic pulse, keeping the coil oriented perpendicularly to the target area, thereby minimizing current delivery to the brain and applying rTMS to uninvolved control cortical sites.
2B. TDCS DEVICES AND STIMULATION PROTOCOLS

The key hardware component for tDCS is a battery-driven constant current generator capable of delivering a constant electrical current flow of up to 2-3 mA. These stimulators do not incorporate complex circuitry and, hence, are significantly less expensive than TMS devices. The electrical current is delivered through two sponge electrodes soaked in saline solution. Typically, these electrodes have a relatively large surface of 2035 cm2 (e.g. 5x5 cm). Even if the wide electrode surface limits the focality of stimulation, it avoids the possible harmful effects of high current density. Moreover, apart from occasional, transient and short-lasting tingling and burning sensations below the electrodes, DC stimulation strength remains below the conscious cutaneous sensory threshold making this technique suitable for double blind sham-stimulation controlled study design. Anodal tDCS upregulates cortical excitability, whereas cathodal tDCS downregulates it. Typical tDCS protocols
51
include 5-20 minute-long single or repetitive sessions (i.e. daily session for more than two consecutive days). The electrodes montage can be bi-cephalic or mono-cephalic, according to the relative position of the active and the reference electrodes. For example, most of the tDCS studies on motor cortex used a bi-cephalic electrodes setup with one electrode (active) positioned on the scalp over the primary motor area and the other (reference) on the contralateral forehead. Conversely, in a mono-cephalic montage, the reference electrode is positioned far away from the scalp (i.e. over the deltoid). As modelling studies suggested that current flow is localized beneath the stimulating electrodes, both the anode and the cathode, the use of a bi-cephalic setup studying complex motor or non-motor behavior would generate confusion regarding the source of the observed effect, as it could originate from the active as well as the cephalic reference electrode. In this regard, a single cephalic electrode on the scalp helps to resolve this ambiguity, as the observed after-effects are unlikely to be related to the extracephalic reference electrode. Among critical technical points to consider in choosing the extracephalic electrode setup are safety criteria and possible charge flow to the brainstem (Nitsche et al. 2003a), which cannot a priori be excluded given the stimulation dipole geometry. Nevertheless, most of the current delivered by transcranial stimulation gets shunted through the scalp and the current density delivered to the cortex by the stimulation decreases rapidly with depth making unlikely to produce a current density able to induce any biological effect in the brainstem.
SAFETY ISSUES
Despite TMS and tDCS are non-invasive in nature and are commonly considered safe, caution is required.
A: Potential adverse effects of rTMS include localized warmth and erythema at the site of stimulation, headache and facial twitching. Transient hearing loss despite earplugs, sedation and sleep have been described in depressed patients (Lisanby et al 2002). Safety studies reported no deterioration in neuropsychological performance, no significant mean changes in auditory threshold, and no significant abnormality in electroencephalography (EEG) after two to four weeks of rTMS (Anderson et al 2006, Janicak et al 2008, Loo et al 2008). Light local sensations of pain during stimulation or transient headache after stimulation are reported by about 10-20% of stimulated patients. The main issue raised with rTMS has been the possibility of seizures. The degree of risk varies with the dosing parameters (i.e. high intensity, high frequency stimulation) and with individual subjects factors (i.e history of epilepsy, or intracranial abnormality, such as a previous stroke). rTMS can also induce bursts of electromyographic activity in several muscles contralateral to the stimulated hemisphere showing that trains of rTMS applied to MI induced a spread of cortical excitability. The spread of excitability depends on the intensity and frequency of the stimuli and probably constitutes an early epileptogenic effect of rTMS (Pascual-Leone et al 1993). Tissue from
52
two adult patients with medically intractable epilepsy who underwent temporal lobe resections 4 weeks after approximately 2000 stimuli, in a rTMS study, showed no pathology attributable to rTMS (Gates et al 1992). Nevertheless, the careful application of published safety guidelines have made TMS-induced seizures an exceptional event, 0.7% in large hemispheric strokes (Wassermann 1998, Chen et al 1997b)). As the magnetic coil can move or heat metallic hardware, the presence of metal anywhere in the head is generally a major contraindication to TMS. Wasserman suggests exclusion from rTMS studies of subjects with cardiac pacemakers and implanted medication pumps, without a clear potential benefit from the treatment. Pregnancy and childhood are considered relative contraindications to rTMS. The risk of TMS to the developing fetus is unknown and few data about the effects of TMS on pregnant women are available. Nahas et al (1999) reported the uncomplicated inclusion of a depressed woman in her second trimester of pregnancy in a depression treatment trial. She received 14 daily treatments at 5 Hz and 100% motor threshold, with marked clinical improvement. A review (Frye et al 2008) analyzed 84 studies including > 800 normal children and >300 neurologically abnormal children. Although TMS has not been associated with any serious side effects in children and appears to be well tolerated, the authors underscored the need of general safety guidelines for use in the pediatric population.
B: The use of tDCS in protocols to date has not resulted in significant adverse effects, but safety issues regarding tDCS are still a matter of controversy. Nietsche et al (2003a) suggest that the appropriate variables for determining safety limits for tDCS should be current density (CD, mA/cm2) and total charge (TC, C/cm2). Data from the literature suggests that tissue damage occurs at a TC of 216 C/cm2 (Yuen et al., 1981) and that a CD below 25 mA/cm2 (McCreery et al., 1990) does not induce tissue damage. Notably, the stimulation parameters commonly used are a thousand-fold lower than these limits. tDCS does not elevate serum neuron-specific enolase levels (Nitsche et al. 2003b), nor does it induce brain edema or alterations of cerebral tissue detectable by conventional MRI or MRI spectroscopy (Nitsche et al., 2004, Rango et al. 2008). tDCS often elicits short-lasting tingling sensations at the beginning and end of the stimulation period, rarely accompanied by redness under the electrode sites. Exceptionally, small skin ulcerations have been described under cathodal electrode. Furthermore, no adverse cognitive effects were noted following a treatment protocol effective in alleviating depression (Fregni et al., 2006). Careful monitoring for adverse effects is indicated, as safety issues with tDCS may only emerge with larger studies or using novel stimulation protocols with repetitive daily sessions.
CONCLUSIONS
TMS and tDCS are techniques of non-invasive neuromodulation. Whereas TMS is based on the application of a magnetic field outside the scalp that induces an electric field able to
53
excite neurons and to produce action potentials, tDCS consists in the application of weak constant electric currents that modulate brain excitability, without inducing action potentials. Devices for TMS and tDCS are, therefore, different, also in terms of costs (Table 1). Although careful monitoring for adverse effects is mandatory, these techniques can be considered safe. Table 1. Relative comparison of stimulation methods
(Courtesy of Dr. Kim H-I)
Safety Focality of stimulation Accuracy Network Stimulation Therapeutic convenience Need for hospital visit/medical professionals Cumulative After-effects Invasiveness Cost
*+/-, +, ++, +++: minimal to maximal.
TMS +++ + ++ + +++ + +/++
tDCS +++ + + ++ ++ + +/+
ECS +++ +++ +++ +++ +++ +/+++ ++ +++
REFERENCES
Accornero N, Li Voti P, La Riccia M, Gregori B. Visual evoked potentials modulation during direct current cortical polarization. Exp Brain Res. 2007;178:261-6. Anderson B, Mishory A, Nahas Z, et al. Tolerability and safety of high daily doses of repetitive transcranial magnetic stimulation in healthy young men. J Ect 2006;22:49-53. Antal A, Nitsche MA, Kruse W, Kincses TZ, Hoffmann KP, Paulus W. Direct current stimulation over V5 enhances visuomotor coordination by improving motion perception in humans. J Cogn Neurosci. 2004;16:521-7. Antal A, Nitsche MA, Paulus W. External modulation of visual perception in humans. Neuroreport 2001;12:3553-5. Antal A, Nitsche MA, Paulus W. Transcranial direct current stimulation and the visual cortex. Brain Res Bull. 2006; 68:459-63. Ardolino G, Bossi B, Barbieri S, Priori A. Non-synaptic mechanisms underlie the aftereffects of cathodal transcutaneous direct current stimulation of the human brain. J Physiol. 2005; 568:653-63. Barker, A. T., Jalinous, R., Freeston, I. L. Noninvasive magnetic stimulation of human motor cortex. Lancet 1985, 11061107. Berardelli A, Inghilleri M, Rothwell JC, at al. Facilitation of muscle evoked responses after repetitive cortical stimulation in man. Exp Brain Res. 1998; 122, 79-84. Chen R, Classen J, Gerloff C, et al. Depression of motor cortex excitability by low-frequency transcranial magnetic stimulation. Neurology 1997;48: 1398-403.
54
Chen R, Gerloff C, Classen J, et al. Safety of different inter-train intervals for repetitive transcranial magnetic stimulation and recommendations for safe ranges of stimulation parameters. Electroencephalogr Clin Neurophysiol 1997b; 105:415421. Daskalakis ZJ, Christensen BK, Fitzgerald PB, Roshan L, Chen R. The mechanisms of interhemispheric inhibition in the human motor cortex. J Physiol. 2002; 543:317-26. Day BL, Sverac Cauquil A, Bartolomei L, Pastor MA, Lyon IN. Human body-segment tilts induced by galvanic stimulation: a vestibularly driven balance protection mechanism. J Physiol. 1997;500:661-72. Ferrucci R, Marceglia S, Vergari M, et al. Cerebellar Transcranial Direct Current Stimulation Impairs the Practice-dependent Proficiency Increase in Working Memory. J Cogn Neurosci 2008 (in press). Fregni, F., Boggio, P.S., Nitsche,M.A., Rigonatti, S.P., Pascual-Leone, A. Cognitive effects of repeated sessions of transcranial direct current stimulation in patients with depression. Depress Anxiety 2006; 23, 482484. Fregni F, Pascual-Leone A. Technology insight: noninvasive brain stimulation in neurologyperspectives on the therapeutic potential of rTMS and tDCS. Nat Clin Pract Neurol 2007;3:383-93. Frye RE, Rotenberg A, Ousley M, Pascual-Leone A. Transcranial Magnetic Stimulation in Child Neurology: Current and Future Directions. J Child Neurol 2008; 23; 79-96. Gates JR, Dhuna A, Pascual-Leone A. Lack of pathologic changes in human temporal lobes after transcranial magnetic stimulation. Epilepsia 1992; 33, 5048. Gualtierotti T. Stimulation of the cerebral cortex with electrodes applied outside the skull. J Physiol. 1953;122(Suppl):17-8P. Hilgetag CC, Theoret H, Pascual-Leone A. Enhanced visual spatial attention ipsilateral to rTMS-induced virtual lesions of human parietal cortex. Nat Neurosci 2001; 4: 95357. Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta burst stimulation of the human motor cortex. Neuron. 2005;45(2):201-6. Inghilleri M, Berardelli A, Marchetti P, Manfredi M. Effects of diazepam, baclofen and thiopental on the silent period evoked by transcranial magnetic stimulation in humans. Exp Brain Res. 1996;109:467-72. Janicak PG, O'Reardon JP, Sampson SM, et al. Transcranial magnetic stimulation in the treatment of major depressive disorder: a comprehensive summary of safety experience from acute exposure, extended exposure, and during reintroduction treatment. J Clin Psychiatry 2008;69:222-232. Kiernan MC, Bostock H. Effects of membrane polarization and ischaemia on the excitability properties of human motor axons. Brain 2000;12:2542-51. Kosslyn SM, Pascual-Leone A, Felician O, et al. The role of area 17 in visual imagery: convergent evidence from PET and rTMS. Science 1999; 284: 16770. Liebetanz D, Nitsche MA, Tergau F, Paulus W. Pharmacological approach to the mechanisms of transcranial DC-stimulation-induced after-effects of human motor cortex excitability. Brain 2002;125:2238-47. Lisanby SH, Kinnunen LH, Crupain MJ. Applications of TMS to therapy in psychiatry. J Clin Neurophysiol. 2002;19:344-60.
55
Loo CK, McFarquhar TF, Mitchell PB. A review of the safety of repetitive transcranial magnetic stimulation as a clinical treatment for depression. Int J Neuropsychopharmacol 2008;11:131-147. Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A Modulation of corticospinal excitability by repetitive transcranial magnetic stimulation. Clin Neurophysiol 2000; 111, 800805. Maertens de Noordhout A, Rothwell JC, Day BL, Thompson PD, Delwaide PJ, Marsden CD. Percutaneous electric and magnetic stimulation of the motor cortex in man. Physiological aspects and clinical applications. Rev Neurol (Paris) 1989;145:1-15. Matsunaga K, Nitsche MA, Tsuji S, Rothwell JC. Effect of transcranial DC sensorimotor cortex stimulation on somatosensory evoked potentials in humans. Clin Neurophysiol. 2004;115:456-60. McCreery DB, Agnew WF, Yuen TG, Bullara LA. Charge density and charge per phase as cofactors in neural injury induced by electrical stimulation. IEEE Trans Biomed Eng 1990; 37:9961001. Merton PA, Hill DK, Morton HB, Marsden CD. Scope of a technique for electrical stimulation of human brain, spinal cord, and muscle. Lancet 1982;2:597-600. Merton PA, Morton HB.Stimulation of the cerebral cortex in the intact human subject. Nature 1980;285:227. Mills KR, Murray NM, Hess CW. Magnetic and electrical transcranial brain stimulation: physiological mechanisms and clinical applications. Neurosurgery 1987;20:164-8. Mottaghy FM, Gangitano M, Sparing R, Krause BJ, Pascual-Leone A. Segregation of areas related to visual working memory in the prefrontal cortex revealed by rTMS. Cereb Cortex 2002; 12: 36975. Nahas Z, Bohning D, Molloy M, Oustz J, Risch C, George M. Safety and feasibility of repetitive transcranial magnetic stimulation in the treatment of anxious depression in pregnancy: a case report. Journal of Clinical Psychiatry 1999; 60, 5052. Nitsche MA, Liebetanz D, Lang N, Antal A, Tergau F, Paulus W. Safety criteria for transcranial direct current stimulation (tDCS) in humans. Clin Neurophysiol. 2003a;114:2220-2. Nitsche MA, Nitsche MS, Klein CC, Tergau F, Rothwell JC, Paulus W. Level of action of cathodal DC polarisation induced inhibition of the human motor cortex. Clin Neurophysiol. 2003b;114:600-4. Nitsche MA, Niehaus L, Hoffmann KT, et al.. MRI study of human brain exposed to weak direct current stimulation of the frontal cortex. Clin Neurophysiol. 2004;115:2419-23. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J Physiol 2000; 527.3:6339. Nitsche MA, Paulus W. Sustained excitability elevations induced by transcranial motor cortex stimulation in humans. Neurology 2001;57:1899901. Pascual-Leone A, Houser CM, Reese K, et al. Safety of rapid-rate transcranial magnetic stimulation in normal volunteers. Electroencephalogr Clin Neurophysiol. 1993; 89:12030. Pascual-Leone A, Valls-Sole J, Wassermann, E. M, Hallett, M. Responses to rapid-rate transcranial magnetic stimulation of the human motor cortex. Brain 1994; 117, 847858.
56
Peterchev AV, Jalinous R, Lisanby SH. A transcranial magnetic stimulator inducing nearrectangular pulses with controllable pulse width (cTMS). IEEE Trans Biomed Eng 2008; 55: 257-266. Priori A, Berardelli A, Rona S, Accornero N, Manfredi M. Polarization of the human motor cortex through the scalp. Neuroreport 1998;9:225760. Priori A. Brain polarization in humans: a reappraisal of an old tool for prolonged noninvasive modulation of brain excitability. Clin Neurophysiol. 2003;114:589-95. Rango M, Cogiamanian F, Marceglia S, Arighi A, Biondetti P, Priori A. Myoinositol Content in the Human Brain Is Modified by Transcranial Direct Current Stimulation in a Matter of Minutes: A 1H-MRS Study. Magnetic Resonance in Medicine 2008; 60:782-789 Rothwell JC. Physiological studies of electric and magnetic stimulation of the human brain. Electroencephalogr Clin Neurophysiol Suppl 1991;43:29-35. Ridding MC, Rothwell JC. Is there a future for therapeutic use of transcranial magnetic stimulation? Nat Rev Neurosci. 2007;8:559-67. Sparing R, Mottaghy FM. Noninvasive brain stimulation with transcranial magnetic or direct current stimulation (TMS/tDCS)-From insights into human memory to therapy of its dysfunction. Methods 2008;44:329-37. Theoret H, Haque J, Pascual-Leone A. Increased variability of paced finger tapping accuracy following repetitive magnetic stimulation of the cerebellum in humans. Neurosci Lett 2001; 306: 2932. Valls-Sol J, Pascual-Leone A, Wassermann EM, Hallett M. Human motor evoked responses to paired transcranial magnetic stimuli. Electroencephalogr Clin Neurophysiol 1992;85:355364. Wassermann EM. Risk and safety of repetitive transcranial magnetic stimulation: report and suggested guidelines from the International Workshop on the Safety of Repetitive Transcranial Magnetic Stimulation, June 57, 1996. Electroenceph Clin Neurophysiol 1998;108:1-16. Werhahn, K. J., Kunesch, E., Noachtar, S., Benecke, R, Classen, J. Differential effects on motorcortical inhibition induced by blockade of GABA uptake in humans. J. Physiol. 1999;517, 591597. Yuen TG, Agnew WF, Bullara LA, Skip Jaques BS, McCreery DB. Histological evaluation of neural damage from electrical stimulation: considerations for the selection of parameters for clinical application. Neurosurgery 1981; 9: 292299. Zago S, Ferrucci R, Fregni F, Priori A. Bartholow, Sciamanna, Alberti: Pioneers in the Electrical Stimulation of the Exposed Human Cerebral Cortex. Neuroscientist. 2008; 14:521-528 Ziemann U, Chen R, Cohen LG, Hallett M . Dextromethorphan decreases the excitability of the human motor cortex. Neurology 1998;51:13201324.
Chapter 5
MECHANISMS OF ACTION
Vincenzo Di Lazzaro1,2, Paolo Profice1, Fabio Pilato1, Michele Dileone1, Federico Ranieri1, Riccardo Di Iorio1 and Pietro A. Tonali1,2
1
Institute of Neurology, Catholic University, Rome, Italy; 2 Fondazione Don C.Gnocchi, Rome, Italy.
INTRODUCTION
Ever since the introduction of transcranial electrical (TES), transcranial magnetic stimulation (TMS) and extradural cortical stimulation (ECS), there has been a considerable number of attempts to define the physiological basis and the structures of the human brain activated by these techniques. Most of our knowledge comes from studies of the primary motor cortex (MI) and is mainly based on the evaluation of the final output produced by MI stimulation, that is the surface recorded muscle response or single motor unit potentials evoked by cortical stimulation. However, motor responses are the result of changes in excitability produced by the stimulus both at cortical and subcortical level. Therefore, the evaluation of muscle responses evoked by cortical stimulation yields limited information about the structures activated. The direct recording of the activity of the corticospinal tract can provide insights about the physiological basis of the motor effects produced by the different techniques of cortical stimulation. This has been first realized in animal preparations (Adrian and Moruzzi 1939; Patton and Amassian 1954; Kernell and Wu 1967). Experimental studies have shown that, in response to a single electrical stimulus to MI, an electrode placed in the medullary pyramid or on the dorsolateral surface of the cervical spinal cord records a series of high frequency waves (Patton and Amassian 1954; Kernell and Wu 1967). The earliest wave that persisted after cortical depression and after cortical ablation was thought to originate from the direct activation of the fast pyramidal tract neuron (PTN) axons and was termed D. The later waves, that required intact grey matter, were thought to originate from the indirect trans-synaptic activation of PTNs and were termed I waves (Patton and
58
Vincenzo Di Lazzaro, Paolo Profice, Fabio Pilato et al.

Amassian 1954). Recordings from individual PTN axons showed that a given axon may give both a D and a subsequent I wave discharge (Patton and Amassian 1954). We recorded directly the corticospinal volley evoked by different forms of cerebral cortex stimulation from the cervical extradural space of conscious patients with chronically implanted spinal electrodes. The ability to record descending corticospinal activity in conscious humans provides a very useful insight into the physiological basis of transcranial and extradural cerebral cortex stimulation, since the synchronous neural volleys are a direct measure of the synaptic activity within central motor circuits evoked by cortical stimulation. These recordings provide detailed insights about the physiological basis of the excitatory and inhibitory phenomena produced by single pulse and repetitive transcranial and extradural stimulation of the human brain.
SINGLE PULSE STIMULATION

1-Transcranial Electrical Stimulation Electrical (anodal) stimulation of MI at active motor threshold intensity (the minimum stimulus intensity that produces a liminal response during voluntary contraction of the tested muscle) evokes a single negative wave with a latency of 2.0-2.6 ms in cervical extradural recordings (Di Lazzaro et al 1998a). The short latency of this wave suggests that it originates from direct activation of corticospinal axons just below the grey matter, and that it is the equivalent of the D wave described by Patton and Amassian (1954) after direct stimulation of the exposed MI in monkeys. Voluntary contraction has no effect on the amplitude of the descending wave evoked by anodal stimulation (Di Lazzaro et al 1999a). The fact that this wave is not modified by changes in cortical excitability produced by voluntary contraction supports the hypothesis that it is due to activation of corticospinal axons in the subcortical white matter at some distance from the cell body. It presumably results from excitation of corticospinal axons at a particular low threshold point (Amassian et al 1992).
2-Transcranial Magnetic Stimulation A-Posterior-anterior Induced Current in the Brain (Figure-of-eight Coil) Using a focal coil (figure-of-eight coil) and a monophasic stimulus, the lowest threshold for evoking a motor evoked potential (MEP) from the hand area of MI is obtained with a stimulus that induces a posterior to anterior (PA) current across the central sulcus in the brain. This lowest threshold volley has a latency 1.0-1.4 ms longer than the volley recruited by electrical anodal stimulation (Di Lazzaro et al 1998a). This volley increases in size and is followed by later volleys as the intensity of stimulation increases. Since the earliest volley elicited by electrical stimulation is probably a D wave, the later volleys recruited by magnetic stimulation represent I waves. The inter-peak interval between I waves is about 1.4 ms, which indicates a discharge frequency of about 700 Hz. At a magnetic stimulus intensity of about 180-200% of the active motor threshold (MT), an earlier wave of small amplitude appears. This wave has the same latency as the D wave evoked by electrical stimulation at threshold.
Mechanisms of Action
59
Changes in cortical excitability produce changes in the output of MI to TMS. An increase in cortical excitability produced by voluntary contraction of the target muscle results in an increase in the output of MI (Di Lazzaro et al 1998b). Voluntary contraction increases the size and number of extradural volleys evoked by a given intensity of magnetic stimulation and the amplitude of the descending waves is higher during activity (particularly during maximum contractions) than at rest. This effect can be substantial: maximum contraction can increase the total amplitude of the waves by 50%. This large effect on the size of the descending corticospinal volleys is not paralleled by a comparable effect on the threshold for evoking recognizable activity after TMS. Voluntary activation of MI has only a small effect on the threshold of descending activity, and only in a minority of subjects. This suggests that the elements activated by PA magnetic stimulation of the upper limb area of MI have a relatively constant threshold. The likely explanation is that magnetic stimulation activates axons at some distance from the cell body so that threshold is unaffected by synaptic activity. These axons cannot belong to corticospinal neurons, since no D wave is elicited at threshold intensity. Presumably they are cortico-cortical axons projecting onto corticospinal neurons. The opposite effect on the output of MI is observed after pharmacological enhancement of inhibitory GABAergic activity. Following administration of lorazepam, the output of MI to high intensity TMS is reduced (Di Lazzaro et al 2000). Neither the threshold nor the amplitude of the lowest threshold volley evoked by TMS is affected by lorazepam. In contrast, a pronounced suppression of later waves (the waves following the I1 wave) is observed. This sensitivity of the later I waves evoked by TMS to interventional manipulation of cortical excitability suggests that they are generated within cortical networks and that they are pre-synaptic in origin. B-Latero-medial Induced Current in the Brain (Figure-of-eight Coil) When the orientation of the figure-of-eight coil is changed, so that monophasic currents in the brain are induced in a lateral to medial (LM) direction, the MEP tends to have the same latency as the one evoked by electrical anodal stimulation (Werhahn et al 1994). At liminal stimulus intensities, TMS with LM induced current in the brain usually recruits two waves (Di Lazzaro et al 1998a): the earliest wave has the same latency of the D-wave evoked by electrical anodal stimulation and the second one has the same latency of the I1 wave evoked by TMS with PA induced current in the brain. This suggests that TMS with LM induced current in the brain excites PTNs directly at their axons, and also trans-synaptically. With LM stimulation at suprathreshold intensity, the I1 wave may be relatively small, whereas the I3 wave may be much larger. A similar behavior is described in the experimental studies of Kernell and Wu (1967) on anodal stimulation of the exposed monkey MI. This suggests that, after direct activation of the corticospinal axons, the stimulus spreads both orthodromically and antidromically, making the axon itself partially refractory to the I1 wave. When, with larger stimuli, most of the axons are activated directly, there are no axons that can respond to indirect activation of PTNs in the latency range of the I1 wave. C-Anterior-posterior Induced Current in the Brain (Figure-of-eight Coil) When reversing the direction of the induced current in the brain from the usual PA direction to the anterior-posterior (AP) direction, MEPs in small hand muscles have an onset
60
latency that is about 3 ms longer than the one seen with PA stimulation (Sakai et al 1997). It has been suggested, therefore, that stimulation with AP induced current can, at least in some individuals, preferentially recruit I3 waves in the pyramidal tract (Sakai et al 1997). The direct recording of the output of MI produced by TMS with AP induced current shows that reversing the direction of the stimulating current does not simply reverse the order of recruitment of descending activity (i.e. an I3 wave before an I1 wave) in all subjects, but leads to a more complex pattern of activation (Di Lazzaro et al 2001a). In some subjects, AP stimulation recruits a later I wave rather than an I1 wave, while in other subjects, I1 or D waves occur, but the latency of these waves is slightly longer than the corresponding waves evoked by PA magnetic stimulation. Also, at suprathreshold intensity, the descending volleys evoked by AP stimulation have slightly different peak latencies and/or longer duration than those seen after PA stimulation (Di Lazzaro et al 2001a). These findings suggest that AP and PA stimulation activate different sites of the same cortical elements, and perhaps even slightly different populations of cortical neurons. Indeed, the folding of the cerebral cortex means that there will be many neurons with axons bent in a particular orientation. Some of these will be best stimulated by one direction of current flow, whereas others will be best stimulated by another current direction. D-Biphasic Magnetic Stimulation (Figure-of-eight Coil) When using biphasic magnetic stimulation, the output is less consistent compared to monophasic stimulation (Di Lazzaro et al 2001b). At active threshold intensity, PA-AP stimulation evokes either an I1 wave, that has the same latency as the I1 wave after monophasic PA, or a delayed I1 wave together with a delayed D wave, or a delayed I3 wave, that have the same latency as the I3 wave evoked by AP stimulation. At intermediate stimulus intensities, the recruitment of additional waves is observed: some of the waves are similar to those observed with monophasic AP stimulation and some have the same latency of the waves evoked by PA stimulation. At high stimulus intensities, the I waves evoked by PA-AP stimulation are similar to those evoked by monophasic PA stimulation. At all intensities, the latency of the D wave is longer than the latency of the D wave from anodal stimulation. Using AP-PA stimulation, active threshold is lower than the one to PA-AP stimulation, and the pattern of recruitment of D and I waves at increasing stimulus intensities resembles the one with monophasic PA stimulation more than with AP stimulation. Therefore, the direct recording of extradural volleys suggests that both phases of the biphasic pulse are capable of activating descending motor output and that biphasic magnetic pulses produce a more complex pattern of corticospinal activation than monophasic pulses (see also Taylot and Loo 2007). Both phases of the stimulus pulse can activate descending pathways, but the precise combination of neural elements activated by AP and PA directions depends on their relative threshold and amplitude. E-Non-focal Stimulation of the Brain (Circular Coil) Initial studies in monkey (Edgley et al 1990) suggested that the lowest threshold volley evoked by a circular coil centered over the vertex originates from the direct activation of the corticospinal axons at their initial segment. In anesthetized human subjects, Burke et al (1993) confirmed that, using a circular coil centred over the vertex and recording from the
61
spinal cord, the D wave is the component of lowest threshold. Our recording of extradural volleys shows different results in different subjects (Di Lazzaro et al 2002). The earliest volley evoked by a circular coil centered over the vertex is a descending wave that has a latency about 0.2 ms longer than the LM or anodal D wave. Either this wave or a wave corresponding to the I1 wave evoked by PA magnetic stimulation can be the lowest threshold volley using a circular coil. At supra-threshold intensities, later waves can be recruited: these waves, in some cases, may have a latency that is outside the periodicity of I waves evoked by PA magnetic stimulation. At higher intensities, the I waves evoked by non focal magnetic stimulation are similar to those evoked by monophasic PA stimulation. Maximum voluntary contraction increases the amplitude of all descending volleys including the earliest (D) wave. These data show that there is a major difference between non-focal stimulation of the hand area with a circular coil and focal stimulation with a figure-of-eight coil, even if the induced current flows in both cases in a posterior-anterior direction. Non-focal stimulation is more likely to evoke a D wave than PA focal stimulation. Moreover, the early (D) volley recruited by non-focal stimulation has a slightly longer latency than the LM or anodal D wave and it is also facilitated by voluntary contraction. These features suggest that it is initiated closer to the cell body of the pyramidal neurons than the conventional D wave evoked by anodal or LM magnetic stimulation, perhaps at the initial segment rather than at some distance down the axon.
3-Extradural Cortical Stimulation There have been few direct comparisons between transcranial and direct stimulation of MI in humans. Katayama et al (1988) recorded descending volleys from the spinal extradural space after direct stimulation of the exposed cortex in anesthetized patients during neurosurgery. They described recruitment of probable D and I waves, with D waves having the lowest threshold for anodal stimulation. The comparison of the pattern of activation produced by TMS and extradural electrical stimulation of MI was performed in a single patient who had both a cortical extradural electrode implanted over the motor area for treatment of pain, and an extradural electrode inserted into the spinal extradural space (Di Lazzaro et al 2004). The recordings showed that extradural MI stimulation can evoke multiple descending waves: an I1 and an I2 wave, with the same latency as magnetic PA stimulation, together with a delayed I3 wave. This suggests that some of the circuits activated by TMS with PA induced current in the brain and by extradural MI stimulation are the same, while other circuits responsible for later I wave generation may be different. ECS also evoked an earlier probable D wave that had a slightly later onset than that after LM stimulation. It might be that the D wave evoked by ECS originates closer to the cell body of the pyramidal neurons than the conventional D wave, perhaps at the initial segment rather than at some distance down the axon. Thus, extradural stimulation of MI can produce repetitive excitation of corticospinal neurons. The order of recruitment of the volleys, and the latency of the D and I3 waves may be slightly different to that seen after TMS. This suggests that there may be subtle differences in the populations of neurons activated by the two forms
62
of stimulation, presumably related to differences in the focality and distribution of the electric field produced by the two methods of stimulation.
REPETITIVE STIMULATION OF THE CEREBRAL CORTEX

In recent years, several authors have attempted to produce changes in the effectiveness of synapses between cortical neurons in the human brain by using techniques of repetitive cerebral cortex stimulation. The hope is that stimulation-induced long term depression (LTD) or long term potentiation (LTP)-like changes of synaptic connections of the brain could promote plasticity in cortical circuits damaged by an acute or a chronic lesion (see chapters 14-16). On the other hand, human studies (Angelucci et al 2004; Zanardini et al 2006; Yukimasa et al 2006) have demonstrated that repetitive stimulation of the cerebral cortex may modulate plasma levels of Brain-Derived Neurotrophic Factor (BDNF), a potent survival factor for neurons. A preemptive neuroprotective and antiexcitotoxic effect of repetitive cortical stimulation has also been suggested (Post et al 1999, Fujiki et al 2003). Several investigators have used rTMS to produce changes of the excitability of the corticospinal system that outlast the period of stimulation. For instance, 900 stimuli of 1 Hz rTMS decrease the amplitude of the MEPs elicited by single pulse TMS for the following 30 minutes (Chen et al 1997), whereas higher rTMS frequencies may increase MEPs amplitude (Berardelli et al 1998). Since spinal H-reflexes are unaffected, it is usually assumed that these rTMS after-effects are due to changes in neural circuits within the cortex, perhaps involving processes such as LTD or LTP at cortical synapses. Several studies provided direct evidence for the cortical origin of these rTMS effects by means of extradural recordings of the corticospinal volleys from the spinal cord (Di Lazzaro et al 2002b,c; Di Lazzaro et al 2005; Di Lazzaro et al 2006; Di Lazzaro et al 2007). It was found that suprathreshold 5 Hz rTMS of MI is accompanied by a gradual increase in the size and number of descending corticospinal volleys evoked by each TMS pulse and that this effect parallels the increase in MEP amplitude (Di Lazzaro et al 2002b). Subthreshold 5 Hz rTMS (total of 50 stimuli at an intensity of active motor threshold) has no effect on MEP amplitude, but reduces paired pulse intracortical inhibition (Di Lazzaro et al 2002c), suggesting that low-intensity rTMS at 5 Hz can selectively modify the excitability of GABAergic networks in the human MI. Huang et al (2005) have described a rapid method to modulate excitability in MI, termed theta burst stimulation (TBS). The protocol uses short bursts of low intensity (80% of active motor threshold), high frequency (50 Hz) pulses, repeated at 5 Hz, i.e. the frequency of the theta rhythm in the EEG. Different patterns of delivery of TBS (continuous versus intermittent) produce opposite effects on synaptic efficiency of the stimulated MI (Huang et al 2005). The paradigm named continuous theta-burst stimulation (cTBS) produces a consistent LTD-like effect, causing a prolonged decrease of MI excitability. Forty seconds of cTBS reduce the amplitude of MEPs significantly for about one hour (Huang et al 2005). Recordings of MI output in conscious subjects with cervical spinal electrodes showed that cTBS leads to a pronounced decrease in the excitability of cortical circuits generating the I1
63
wave, whilst later I waves are much less affected (Di Lazzaro et al 2005). This suggests that cTBS has its major effect on the synapse between the interneurons responsible for the I1 wave and the corticospinal neurons. The paradigm named intermittent theta-burst stimulation (iTBS) produces a consistent LTP-like effect, causing a prolonged increase of MI excitability (Huang et al 2005). The effects of iTBS were evaluated in a chronic stroke patient who had a dorsal extradural electrode, where iTBS determined a pronounced enhancement of the corticospinal descending activity evoked by lower limb area stimulation (Di Lazzaro et al 2006). Another method of increasing excitability in the corticospinal system is based on repetitive paired stimulation at I wave periodicity (iTMS). This protocol uses paired TMS stimuli of equal intensity with a 1.5 ms interstimulus interval, delivered for several minutes at a rate of 0.2 Hz (Thickbroom et al 2006). During iTMS MEP amplitude increases steadily and, moreover, single-pulse MEP amplitude is increased for several minutes after the end of stimulation. This after-effect of iTMS is considered to take place at cortical level because MEPs evoked by cervico-medullary junction stimulation are not affected by iTMS (Hamada et al 2007). The effects of iTMS on extradural activity were evaluated in a single patient with high cervical extradural electrode (Di Lazzaro et al 2007). In this patient a pronounced increase in MEP amplitude was paralleled by only a slight increase in the amplitude of extradural volleys (Di Lazzaro et al 2007). It has been suggested that the I waves do not represent all the descending activity evoked by TMS and that there might be additional activity that is more dispersed and not apparent in the extradural recordings. If this additional activity is increased by iTMS, then the MEP would be larger, even though the increase in the I wave activity is rather limited. There is some evidence to support the idea that dispersed descending activity could contribute to the facilitation of MEPs: a similar phenomenon is observed with the facilitation of MEPs produced by paired magnetic stimulation at interstimulus intervals of 10-25 ms. This form of intracortical facilitation is also not associated with an increase in corticospinal output as evaluated through extradural activity recording, though it is cortical in origin (Di Lazzaro et al 2006b). Thus, the same phenomenon could be produced by iTMS.
CONCLUSIONS (FIGURE 1)
The different techniques of cortical stimulation activate axons of cortical motoneurons at variable distance from the cell body. It can be hypothesized that electrical anodal stimulation activates directly the axons of PTNs at some distance from the cell body in the white matter, as demonstrated by the fact that the output of MI evoked by TES is not influenced by the level of cortical excitability. This is compatible with a site of activation located downstream to the cerebral cortex. In case of TMS with a posterior-anterior induced current in the brain, the fact that the threshold for evoking descending activity is scarcely influenced by the level of cortical excitability again suggests that the axons at some distance from the cell body are the site of excitation. However, in contrast with the output produced by TES, the output of MI to TMS is highly sensitive to the level of cortical excitability. Indeed, increase or decrease in cortical excitability are paralleled by an increase or a decrease of the output. This
64
suggests that, in this case, the activated axons are pre-synaptically located with respect to the PTNs and that TMS activates the PTNs mainly indirectly through the axons of excitatory systems projecting upon them. The orientation of the electric field induced by TMS is crucial for the activation of cortico-cortical or cortico-spinal axons: when the direction of the current induced in the brain is changed to latero-medial, both cortico-spinal and cortico-cortical axons are activated. The output is even more complex when the direction of the induced current is changed to anteriorposterior.
Figure 1. Diagrammatic representation of possible sites of direct and indirect activation of corticospinal cells using different techniques of transcranial (TES and TMS) and extradural cortical (ECS) stimulation. The triangular neuron is a corticospinal cell, the circular neurons are first and second order excitatory interneurons. A subcortical D wave activation at some distance from the cell body of the pyramidal cell axon is preferentially produced by TES and LM TMS (focal coil). Note that D wave activation at the proximal part of the pyramidal cell axon can be produced by non-focal TMS (circular coil), biphasic TMS, and ECS. Recruitment of I waves by TMS (both focal and circular coil) and by ECS is due to activation of the axons of excitatory interneurons.
In certain circumstances, the site of impulse initiation after TMS moves closer to the cell body. This is observed in the case of MI stimulation using a large circular coil centered over the vertex and in case of biphasic magnetic stimulation. Extradural stimulation of MI also produces repetitive excitation of corticospinal neurons. However, the order of recruitment of the volleys, and the latency of the D wave and of later I waves may be slightly different to that seen after TMS, suggesting that there may be subtle differences in the populations of neurons activated by the two forms of stimulation. In order to improve clinical effects of
65
rTMS, changing the orientation of the coil by 90 (from postero-anterior to latero-medial) may make it more ECS-like and thus more effective. Repetitive stimulation leads to changes in the excitability of cortical mechanisms generating corticospinal volleys that depend on the specific paradigm used. Several rTMS parameters and, in particular, the frequency, the intensity and the pattern of stimulation, influence the direction of its effect (enhancement or suppression of cortical excitability). Most of the protocols modulate the excitability of cortical circuits generating later I waves. This is the case for iTBS and 5 Hz rTMS, but it should be considered that this latter protocol can also affect the excitability of the corticospinal axons. The intensity of stimulation is crucial: 5 Hz rTMS at subthreshold intensity does not modify the excitability of the excitatory circuits, but selectively affects the excitability of the inhibitory circuits, as demonstrated by changes in paired pulse cortical inhibition. cTBS preferentially affects the amplitude of the earliest I wave and not of later I waves. In contrast with the other protocols, iTMS does not affect the mechanisms generating the later I waves, but involves circuits in addition to those involved in the generation of these waves. The implication is that populations of cortical excitatory and inhibitory neurons are differentially sensitive to the different protocols of repetitive stimulation. This may be of importance when testing the therapeutic potential of these methods in patients with neurological disorders. On the other hand, the effects of ECS remain poorly understood (chapters 6 and 7).
REFERENCES
Adrian ED, Moruzzi G. Impulses in the pyramidal tract. J Physiol 1939;97:153-99. Amassian, VE, Eberle, LP, Maccabee, PJ, Cracco, RQ. Modelling magnetic coil excitation of human cerebral cortex with a peripheral nerve immersed in a brain shaped volume conductor: the significance of fiber bending in excitation. Electroenceph clin Neurophysiol 1992;85:291-301. Angelucci F, Oliviero A, Pilato F, et al. Transcranial magnetic stimulation and BDNF plasma levels in amyotrophic lateral sclerosis. Neuroreport. 2004;15:717-20. Berardelli A, Inghilleri M, Rothwell JC, et al. Facilitation of muscle evoked responses after repetitive cortical stimulation in man. Exp Brain Res. 1998;122:79-84. Burke D, Hicks R, Gandevia SC, Stephen J, Woodforth I, Crawford M. Direct comparison of corticospinal volleys in human subjects to transcranial magnetic and electrical stimulation. J Physiol. (Lond) 1993;470:383-93. Chen R, Classen J, Gerloff C, Celnik P, Wassermann EM, Hallett M, Cohen LG. Depression of motor cortex excitability by low-frequency transcranial magnetic stimulation. Neurology. 1997;48:1398-403. Di Lazzaro V, Dileone M, Profice P, et al. Direct demonstration that repetitive transcranial magnetic stimulation can enhance corticospinal excitability in stroke. Stroke. 2006a;37:2850-3. Di Lazzaro V, Oliviero A, Berardelli A, et al. Direct demonstration of the effects of repetitive transcranial magnetic stimulation on the excitability of the human motor cortex. Exp Brain Res 2002b;144, 549-53.
66
Di Lazzaro V, Oliviero A, Mazzone P, et al. Comparison of descending volleys evoked by monophasic and biphasic magnetic stimulation of the motor cortex in conscious humans. Exp Brain Res. 2001b;141:121-7. Di Lazzaro V, Oliviero A, Mazzone P, et al. Short-term reduction of intracortical inhibition in the human motor cortex induced by repetitive transcranial magnetic stimulation. Exp Brain Res 2002c;147, 108-13. Di Lazzaro V, Oliviero A, Meglio M, et al. Direct demonstration of the effect of lorazepam on the excitability of the human motor cortex. Clin Neurophysiol. 2000;111:794-9. Di Lazzaro V, Oliviero A, Pilato F, et al. Descending volleys evoked by transcranial magnetic stimulation of the brain in conscious humans: effects of coil shape. Clin Neurophysiol. 2002a;113:114-9. Di Lazzaro V, Oliviero A, Profice P, et al.Effects of voluntary contraction on descending volleys evoked by transcranial electrical stimulation over the motor cortex hand area in conscious humans. Exp Brain Res 1999a; 124:525-8. Di Lazzaro V, Oliviero A, Profice P, et al. Comparison of descending volleys evoked by transcranial magnetic and electric stimulation in conscious humans. Electroencephalogr. Clin. Neurophysiol 1998a;109:397-401. Di Lazzaro V, Oliviero A, Saturno E, et al. The effect on corticospinal volleys of reversing the direction of current induced in the motor cortex by transcranial magnetic stimulation. Exp Brain Res. 2001a;138:268-73. Di Lazzaro V, Oliviero A, Pilato F, et al. Direct recording of the output of the motor cortex produced by transcranial magnetic stimulation in a patient with cerebral cortex atrophy. Clin Neurophysiol. 2004;115:112-5. Di Lazzaro V, Pilato F, Oliviero A, et al. Origin of facilitation of motor-evoked potentials after paired magnetic stimulation: direct recording of epidural activity in conscious humans. J Neurophysiol. 2006b;96:1765-71. Di Lazzaro V, Pilato F, Saturno E, et al. Theta-burst repetitive transcranial magnetic stimulation suppresses specific excitatory circuits in the human motor cortex. J Physiol. 2005;565:945-50. Di Lazzaro V, Thickbroom GW, Pilato F, et al. Direct demonstration of the effects of repetitive paired-pulse transcranial magnetic stimulation at I-wave periodicity. Clin Neurophysiol. 2007;118:1193-7. Di Lazzaro, V., Restuccia, D., Oliviero, A., et al. Effects of voluntary contraction on descending volleys evoked by transcranial stimulation in conscious humans. J Physiol (Lond) 1998b;508:625-633. Edgley, SA., Eyre, JA., Lemon, RN, Miller, S. Excitation of the corticospinal tract by electromagnetic and electric stimulation of the scalp in the macaque monkey. J of Physiol (Lond) 1990;425:301-320. Fujiki M, Kobayashi H, Abe T, Kamida T. Repetitive transcranial magnetic stimulation for protection against delayed neuronal death induced by transient ischemia. J Neurosurg. 2003;99:1063-9. Hamada M, Hanajima R, Terao Y, et al. Origin of facilitation in repetitive, 1.5ms interval, paired pulse transcranial magnetic stimulation (rPPS) of the human motor cortex. Clin Neurophysiol. 2007;118:1596-601.
67
Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta burst stimulation of the human motor cortex. Neuron 2005;45:201-6. Katayama Y, Tsubokawa T, Maejima S, Hirayama T, Yamamoto T. Corticospinal direct response in humans: identification of the motor cortex during intracranial surgery under general anaesthesia. J Neurol Neurosurg Psychiatry 1988;51:5059. Kernell D, Chien-Ping WU. Responses of the pyramidal tract to stimulation of the baboon's motor cortex. J Physiol. 1967;191:653-72. Patton, H. D. & Amassian, V. E. Single- and multiple-unit analysis of cortical stage of pyramidal tract activation. J Neurophysiol 1954;17:345-363. Post A, Mller MB, Engelmann M, Keck ME. Repetitive transcranial magnetic stimulation in rats: evidence for a neuroprotective effect in vitro and in vivo. Eur J Neurosci. 1999;11:3247-54. Sakai K, Ugawa Y, Terao Y, Hanajima R, Furubayashi T, Kanazawa I. Preferential activation of different I waves by transcranial magnetic stimulation with a figure-of-eight-shaped coil. Exp Brain Res. 1997;113:24-32. Taylor JL, Loo CK. Stimulus waveform influences the efficacy of repetitive transcranial magnetic stimulation. J Affect Dis 2007; 97: 271-276. Thickbroom GW, Byrnes ML, Edwards DJ, Mastaglia FL. Repetitive paired-pulse TMS at Iwave periodicity markedly increases corticospinal excitability: a new technique for modulating synaptic plasticity. Clin Neurophysiol. 2006;117:61-6. Werhahn, K.J., Fong, J.K.Y., Meyer, B.U., et al. The effect of magnetic coil orientation on the latency of surface EMG and single motor unit responses in the first dorsal interosseous muscle. Electroenceph clin Neurophysiol 1994; 93:138-146. Yukimasa T, Yoshimura R, Tamagawa A, et al. High-frequency repetitive transcranial magnetic stimulation improves refractory depression by influencing catecholamine and brain-derived neurotrophic factors. Pharmacopsychiatry. 2006;39:52-9. Zanardini R, Gazzoli A, Ventriglia M, et al. Effect of repetitive transcranial magnetic stimulation on serum brain derived neurotrophic factor in drug resistant depressed patients. J Affect Disord. 2006;91:83-6.
Chapter 6
OPTIMIZATION OF PARAMETER SELECTION AND ELECTRICAL TARGETING

Amorn Wongsarnpigoon and Warren M. Grill
Department of Biomedical Engineering, Duke University, Durham NC USA.
INTRODUCTION
Presently, it is unclear how the stimulation parameterspolarity, pulse width, frequency, electrode location, etc.influence the efficacy of ECS. Until we know the exact mechanisms of how ECS treats specific disorders, we cannot say for certain what the optimal placement and geometry of the electrode are, nor do we know what the best combination of stimulation parameters is. If we do not know which neurons are responsible for the therapeutic effects, then it is unclear how to set the stimulation parameters or position the electrode to stimulate them selectively. In addition, it is not known whether the target neurons should be activated, inhibited, or tonically polarized. This lack of understanding might prevent patients from receiving maximum benefit from ECS. In addition, clinicians may have difficulty selecting the best stimulus parameters for a patient, a challenge also found in deep brain stimulation (DBS) (Kuncel and Grill 2004), where 12964 combinations of voltage, pulse width and frequency plus 65 combinations of electrode geometry are possible. Stimulation parameters are selected through a time-consuming trial and error process that does not guarantee optimal benefit. Improper selection of stimulation parameters can reduce therapeutic effects, shorten battery life and/or increase side effects. During ECS, the electrode is placed over the area of the cortex corresponding to the disorder being treated. For example, during treatment of movement disorders, stimulation is delivered to the area of the cortex known as the primary motor cortex (MI), which is responsible for executing and controlling motor functions (Canavero and Bonicalzi 2007a,b). MI contains a motor map of the entire body; similarly, the body can be mapped to areas of the sensory cortex (SI), which is responsible for processing tactile sensation
Correspondence concerning this article should be addressed to: Dr. Warren M. Grill, Ph.D. e-mail: warren.grill@duke.edu.
70

(Box 1 for a description of cortical cytoarchitecture). A basic understanding of the fundamental principles governing ECS is required for the rational selection of stimulation parameters. The goal of this chapter is to introduce these principles and how they relate to ECS.
BOX 1. Neuronal cytoarchitecture of the cortex

The human cerebral cortex is physically convoluted, with numerous folds and grooves. The ridges in the cortex are known as gyri (sing. gyrus), while the crevices separating the gyri are known as sulci (sing. sulcus). On the surface of the cortex is a layer of gray matter, usually about 2-4 mm thick. The gray matter can be subdivided into six layers with layer I being the most superficial layer and layer VI the deepest layer. These layers are distinguished by the presence of specific types of cells and neural elements. The laminar structure of the gray matter can be seen in all regions of the cortex, although some regions of the cortex vary slightly. For instance, layer IV is not present in the motor cortex, but is quite prominent in the visual cortex. Beneath the gray matter is the white matter, which forms the bulk of the cerebral hemispheres.Although neurons have the same basic componentssoma, dendrites, and axonsneurons can vary greatly in shape, size, and function, particularly in the cortex. Also, cortical neurons are located at different depths within the cortex. These differences among cortical neurons make it difficult to predict which neurons are stimulated during ECS and what the effect(s) of stimulation will be. Cortical neurons can be divided into two general groups: pyramidal cells and local interneurons. Pyramidal cells represent 75-80% of the neurons in the cortex and are located only at specific depths within the cortex. Local interneurons, on the other hand, can be found throughout the cortex and make up the remaining 20-25% of the neurons in the cortex. Also known as projection neurons, pyramidal cells send excitatory signals out from the cortex to subcortical targets. While some interneurons are also excitatory, most are inhibitory and only target nearby cortical cells. The neurons of the cortex can be classified by their shape. All pyramidal cells have the same basic structure: a conical soma, a single apical dendrite that rises up towards the surface of the cortex and branches along the way, basal dendrites that project laterally or down from the soma, and axons that project down from the soma away from the surface (Figure 1a-b). Among the interneurons, the morphologies of the somas, dendrites, and axons are highly variable, but the interneurons can be classified through their axonal features (Figure 1c-e). For example, one type of interneuron known as a basket cell is known for its axon arborizations that seem to form baskets around the somas of pyramidal cells. Within and among classes of cortical neuron, the sizes of the neurons vary. The somas of pyramidal cells can have diameters that range from 12 m all the way up to 100 m (Feldman 1984; Nolte 2002). The largest pyramidal cells are known as Betz cells and can be found in layer V of the motor cortex. The diameters of the somas of interneurons also vary, though not quite as much as among pyramidal cells. For example, the diameters of the somas of basket cells range from 15-30 m (Jones and Hendry 1984). In general, diameters of the somas of all interneurons range from 10-20 m (Markram et al. 2004). The diameters of the axons and dendrites can also vary. The diameters of the axons of pyramidal cells range from 0.5-1 m, and the diameter is directly proportional to the diameter of the soma. The diameters of the apical dendrites of pyramidal axons are slightly larger, ranging from 1-3 m (Feldman 1984). Even larger are the dendrites of basket cells, which can range from 5-10 m in diameter (Jones and Hendry 1984). Although the diameters of the somas, axons, and dendrites of some interneurons are greater than or equal to those of the pyramidal cells, the lengths of the dendrites and axons are usually much greater in pyramidal cells. For example, the apical dendrites of Betz cells span almost the entire width of the cortex (2-4 mm) while the axons are even longer, connecting the cortex to the spinal cord (about 1 m long). On the other hand, with a few exceptions, the dendrites and axons of local interneurons are generally much shorter. The synaptic inputs to the different classes of cortical neurons vary greatly. Most often, excitatory inputs are received at dendrites (Sloper 1973; Ichikawa et al. 1985; Kaneko et al. 1994; Kimura et al. 1994; Kaneko et al. 2000) while inhibitory inputs are received at the somas and axons (Peters and Kaiserman-Abramof 1970; DeFelipe et al. 1985; Peters et al. 1991). Pyramidal cells often deliver excitatory inputs to one another, for instance, layer III pyramidal cells to layer V pyramidal cells (Kaneko et al. 2000), or from one functional cortical area to another (DeFelipe and Farinas 1992). However, interneurons rarely make direct connections onto one another (Markram et al. 2004). A great deal of connectivity exists between pyramidal cells and interneurons. Pyramidal cells receive inhibitory input from many different classes of interneurons and deliver excitatory inputs to a greater number of classes (Markram et al. 2004). Finally, cortical neurons receive inputs from subcortical structures. Pyramidal cells can receive excitatory inputs from the thalamus (Ichikawa et al. 1985) as well as from the claustrum (DeFelipe and Farinas 1992), and interneurons can receive inputs from the thalamus (spiny stellate neurons) (Lund 1984). In sum, the geometry of the cortex is complex and inhomogeneous at both the macro (e.g. gyri and sulci, laminar structure) and micro (i.e. cellular) levels, which makes predicting the effects of ECS difficult.
Optimization of Parameter Selection and Electrical Targeting
71
Figure 1. Neurons in the cortex vary greatly in terms of shape, size, and location in the cortex. A few different cortical neurons are depicted, including a) layer 5 pyramidal cell, b) layer 3 pyramidal cell, c) chandelier cell, d) basket cell, and e) double bouquet cell. Thicker lines represent dendrites while thinner lines represent axons. a) and b) are projection neurons, whereas c)-e) are local interneurons.
THE TWO-PIECE PROBLEM

Determining the response of cortical neurons to ECS can be thought of as a two-piece problem. First, the distributions of electric field and current density generated in the brain during ECS must be calculated. Second, the effects of the generated electric fields on different neurons in the cortex must be determined. It is important to understand both pieces of the problem to optimize the delivery of ECS.
1-Generation of Electric Fields (Box 2) If current only passed through a single medium with uniform conductivity during ECS, then calculating potentials in the cortex would be a fairly trivial matter. In reality, current flows through many different media. Consider how current reaches the brain during monopolar anodic (i.e., positive current or voltage) stimulation. Current leaves the anode, passes through the dura mater, enters the cerebrospinal fluid (CSF), and flows into the cortex. Also influencing the flow of current is the anatomy surrounding the cortex, including the skull and white matter (chapter 7). To determine the electric fields generated during ECS, the conductive properties of the different media in the head must be analyzed. A-Conductive Properties of Anatomical Structures Involved in ECS The different structures in and around the cortex have inhomogeneous and anisotropic electrical properties (Table 1).
72
Amorn Wongsarnpigoon and Warren M. Grill BOX 2. Principles of electromagnetism
During neural stimulation, an electric field is generated by the electrode and exerts forces on charged particles (ions) in the body. The electric field indicates the magnitude and direction of the force that is experienced by a unit of charge. Another measure known as current density describes the rate and direction of the flow of charge. The relationship between the electric field and current density in a conductive medium is known as Ohms Law and is one of the most basic principles of electromagnetism: J = E (1) where J is the current density, is the conductivity of the medium in which the electric field is applied, and E is the electric field. The conductivity of a medium is directly proportional to how easily charge can flow through it. Every material in the bodyblood, bone, fat, etc.has a unique conductivity. Consider a current source located at a point (point source) in a conductive medium. The current source produces a constant current I, and the medium has conductivity . Whenever current passes through a conductive material, there is a change in electric potential, . If is assumed to be 0 at a point infinitely far away from the current source, then we can derive from Ohms Law an expression for anywhere in space: (2) 4r where r is the distance from the current source. We can make three observations from this expression: 1) as I is increased, will also increase, 2) will increase as we get closer and closer to the current source (i.e., as r gets smaller), and 3) will increase if the conductivity decreases. These relationships would also hold if a voltage source were used instead of a current source. When there is more than one current source, can be found by using the principle of superposition. That is, can be found at any point in the conductive medium by calculating the potential generated by each current source separately and then adding them together. Now we will consider a current source located in inhomogeneous media, in other words, a volume composed of different media with different conductivities. In this case, equation (2) no longer applies because the current and electric field in each medium are influenced by the presence of the other media. To determine , we must analyze what happens at the boundaries between media. Consider the simple case of a volume composed of two media: medium 1 with conductivity 1 and medium 2 with conductivity 2. The potential at a point on the boundary between the media must be the same regardless of whether we approach it from medium 1 or medium 2, i.e., the potential must be continuous. If were discontinuous, then according to Ohms Law a current would flow across the boundary. However, since the boundary has zero thickness, this current would have to be infinitely large, which is physically impossible. Another condition that must hold at any point on the boundary is continuity of current. In other words, the normal component of the current that flows into a point on the boundary from one medium must equal the normal component of the current that flows out of the point in the other medium. These boundary conditions can be used to derive the equations for the potential in either medium, which are quite different from equation (2). With additional media, each boundary between media must be analyzed in the same way, resulting in increasingly complicated equations for the potential.
= I
Table 1. Conductivities of Materials in the Body

Material Skull (bone) Dura mater CSF Gray matter White matter (parallel to fibers) White matter (perpendicular to fibers) Conductivity (S/m) 0.00625 0.065 1.7 0.2 1.1 0.13 Source (Haueisen et al. 2002) (Manola et al. 2005) (Geddes and Baker 1967) (Ranck 1963; Li et al. 1968) (Nicholson 1965)
The structure in the vicinity of the cortex with the lowest conductivity is the skull. The conductivity is so low that very little current travels through it during ECS, and it acts essentially as an electric insulator. Below the skull is the dura mater, which is much more
73
conductive than the skull, but still less conductive than the deeper layers in the head. The CSF, composed primarily of water and ions, is highly conductive. The cerebral hemispheres are inhomogeneous; the gray matter and white matter have different conductive properties due to their different compositions. Like the other layers, the gray matter is isotropic, meaning that the conductivity is the same in all directions. The white matter, on the other hand, is anisotropic because the axons that make up the white matter are highly oriented perpendicular to the surface of the cortex. Since current prefers to flow parallel to these fibers, the conductivity in the white matter is greater in this direction than in other directions. So far, the capacitive and inductive properties of the tissue have been ignored. Equation (2) in Box 2 implies that, if I changes over time, then can still be described by the equation at any instant in time. This is known as the quasi-static approximation. When the capacitive and inductive components are taken into account, the conductivity and, consequently, becomes dependent on the frequency of I. The frequency in this context refers not to the rate at which each pulse is delivered (stimulation frequency), but rather to the frequency content of each individual pulse. For instance, a single rectangular pulse can be decomposed into an infinite number of sinusoidal waves that span the entire spectrum of frequencies. Because the current delivered during ECS and other forms of neural stimulation are almost never constant (d.c.) signals, it is not obvious whether the quasi-static approximation is valid. If the approximation is inaccurate and the capacitive and inductive effects must be taken into account, then calculating potentials becomes much more arduous. Fortunately, it has been shown using typical stimulation parameters under physiologically realistic conditions that the potentials generated during neural stimulation are estimated well by the quasi-static approximation (Bossetti et al. 2008). B-Properties of Epidural Electrodes Up to this point, we have only considered point sources, which only represent well microelectrodes with sharp tips. Now the properties of electrodes with realistic geometries will be examined. The shape, size, and configuration of the electrodes have a significant effect on the electric fields that are generated. Epidural electrodes are composed of a thin insulating substrate in which one or more metal contacts are embedded. Electrical stimulation is delivered through these contacts, usually thin metal discs arranged in various configurations in the substrate. The discs generate potentials in a spatial pattern that is different from that of a point source (Wiley and Webster 1982) (Figure 2). In fact, any unique contact shape will produce a distinct spatial pattern of potential. Related to the spatial pattern of potential is the distribution of current density. Current flows away from an anode (or towards a cathode) and decreases in magnitude with increasing distance. The magnitude of the current density is greatest beneath the perimeter of the electrode, a phenomenon known as edge effect (Rubinstein et al. 1987). The current density becomes more uniform further from the electrode, and the size and shape of the electrode determine how far beneath the electrode the edge effects can be seen. C-Electrode-tissue Interface Charge is carried by electrons in electrical circuits, while in the body it is carried by ions. During neural stimulation, a transition takes place from the flow of electrons in the pulse
74
generator to the flow of ions in the body, which is accomplished via electrochemical reactions that occur at the electrode-tissue interface. A simple model of these reactions, which are seen at both the working and return electrodes, is a resistor and capacitor in parallel, representing faradaic reactions and double layer capacitance, respectively.
Figure 2. Isopotential plots for a) point electrode, and b) disc electrode.
D-Regulated Current vs. Regulated Voltage Stimulation If a simple circuit model of ECS is analyzed, then the effects of the electrode-tissue interface on the output of the electrode can be observed (Figure 3). We are interested in the current that passes through the tissue, Itissue, since this current is the source that generates potentials in the cortex (i.e., equation (2)) and causes neuronal polarization.
Figure 3. Circuit model of electrode-tissue interface.
Neural stimulators generate stimuli that are either voltage-regulated or current-regulated. Regardless of the electrode-tissue interface and impedance of the tissue, if the stimulus is voltage-regulated, then the voltage from the stimulator, Vstim, can be controlled precisely.
75
Likewise, the current output, Istim, can be controlled when using current-regulated stimuli. Although the difference between these methods of regulating stimulation does not appear to be significant, in fact, they have a substantial effect on Itissue. Suppose a square pulse is delivered using current-regulated stimulation (Figure 4a). A simple analysis of the circuit reveals that Istim = Itissue, i.e., the same square pulse will pass through the tissue. Now suppose a square pulse is delivered using voltage-regulated stimulation (Figure 4b). Because of the electrode-tissue interface, the pulse is no longer square when it reaches the tissue. Instead, as a result of the charging and discharging of the capacitance at the electrode tissue interface(s), Itissue jumps on the rising and falling phases of the square pulse and decays to a steady-state value. This lack of control of Itissue when using voltage-regulated stimulation is a reason why many neural stimulators employ current-regulated stimulation.
Figure 4. Current-regulated stimulation (Istim) vs. voltage-regulated stimulation (Vstim) through electrode-tissue interface.
E-Simulation of ECS Although it is easy to quantify electric fields in idealized situations (e.g. a point source in a homogeneous medium) (BOX 2), it is difficult to make anything, but qualitative predictions in more complex situations such as in ECS. To quantify the electric fields generated during ECS, we used the principles described thus far to construct a computer-based, threedimensional extruded slab model of the precentral gyrus, the surrounding anatomical structures, and an epidural electrode (Figure 8). Each region of the model was assigned a value of conductivity corresponding to the structure it represented (Table 1) and the model was solved numerically using the finite element method (Wongsarnpigoon and Grill 2008).
76
Amorn Wongsarnpigoon and Warren M. Grill Box 3. Electrical circuit models of neuronal fibers and the activating function
Axons and dendrites are often modeled as simple fluid-filled fibers using the core-conductor model, also known as the cable model. In this model, axial symmetry is assumed: the potential inside and outside the fiber only varies along the length of the fiber and does not vary with the angle from the axis of the fiber. As a result, the cable model reduces to a one-dimensional model. At each point within the fiber, current can flow through the fiber (intracellular current) as well as through the membrane and out of the fiber (membrane current). An equivalent electrical circuit to the membrane is a resistor, representing the ion channels, in parallel with a capacitor, representing the lipid bilayer. The intracellular fluid can be represented as a resistor as well. Using these equivalent electrical circuits, a distributed electrical circuit model can be constructed, where we assume that the fiber is composed of a series of interconnected compartments that represent different points along the length of the fiber (Figure 5). By analyzing this distributed electrical circuit model, an equation can be derived that describes the membrane potential along the entire length of the fiber (Plonsey and Barr 2000): 2 2 Vm Vm Ve Vm 1 = ( + ) (3) 2 2 rm c m ri c m x t x where Vm is the transmembrane potential, Ve is the potential along the outside of the fiber resulting from extracellular stimulation, x is the length along the fiber, t is time, ri is the intracellular resistance per unit length, rm is the membrane resistance per unit length, and cm is the membrane capacitance per unit length. If there is no extracellular stimulation, then we assume that Ve is a constant and 2Ve/x2 = 0. The complexity and non-linearity of (3) make it impossible to predict the exact membrane voltage along the entire fiber for any arbitrary point in time. However, equation (3) can be used to approximate the response of the neuron to extracellular stimulation. Before extracellular stimulation is applied, the membrane along the entire fiber is at resting potential, which in equation (3) means that Vm=0 and 2Vm/x2 = 0. The instant after stimulation is turned on, it is still assumed that Vm=0, and 2Vm/x2 = 0 since the capacitance of the membrane prevents an instantaneous change in membrane voltage. Consequently, equation (3) becomes: 2 Vm 1 Ve = (4) 2 ri c m x t This equation states that the change in membrane potential is proportional to the second spatial derivative of the potential along the fiber, which has been termed the activating function (Rattay 1989). Where the activating function is positive, Vm/t is also positive, which indicates that the membrane is depolarized. Similarly, the membrane will be hyperpolarized where the activating function is negative. Consider a point current source in an extracellular medium. The equation for the potential in the medium is given by equation (2). To calculate the voltage along a straight fiber, we substitute r = (d2 + x2)1/2 into the equation to get
= Ve =
I 4 d 2 + x 2
(5)
where x is the position along the fiber, x = 0 is the point closest to the current source, and d is the distance between the current source and x = 0. Then, the equation for the second spatial derivative of Ve is 2 Ve I 2 2 2.5 2 2 (d + x ) ( 2 x d ) (6) = 2 4 x The plots of Ve(x), Ve(x)/x and 2Ve(x)/x2 are shown for cathodic stimulation (negative I) in Figure 6a. The activating function depends on the spatial distribution of the potential. For instance, if a disc electrode were used instead of a point source, then the profiles of Ve(x), Ve(x)/x and 2Ve(x)/x2 change (Figure 6b). Notice the triphasic shape of the activating function in Figure 6a, which predicts a depolarized region on the fiber beneath the current source flanked by two hyperpolarized regions, known as virtual anodes. As a result, during cathodic stimulation, the region beneath the electrode is the most likely site for the initiation of action potentials. If anodic stimulation were delivered instead, all of the plots in Figure 6 would simply be flipped over the x-axis. Then, action potentials could be initiated at the flanking regions, which in this case are known as virtual cathodes. However, because the amplitude of the flanking peaks is not as great as the amplitude of the central peak, the threshold amplitude during anodic stimulation is greater than during cathodic stimulation.
Optimization of Parameter Selection and Electrical Targeting Box 3. (Continued)
77
The predictions of the activating function make sense when the flow of current is considered. Again, consider a point source delivering cathodic stimulation. Current is drawn into the cathode from the fiber, which causes depolarization of the membrane. However, the current drawn from the middle of the fiber must come from adjacent regions of the fiber, which in turn draws some current from the extracellular space and causes hyperpolarization (Figure 7). Thus, the triphasic shape of the activating function is an expected phenomenon. Although the activating function has been used with great success to predict the relative membrane voltage along fibers during extracellular stimulation, it must be used carefully. The activating function is the driving force of equation (3), not the solution. As a result, the prediction made by the activating function becomes worse as the duration of stimulation increases (Warman et al. 1992). Another reason to be careful when using the activating function is that it was calculated under the assumption that the fiber was infinitely long. If a finite-length fiber is considered, then the activating function still approximates the response well over most of the fiber. However, if the ends of the fiber are located near the electrode, then the voltages at the ends of the fiber are approximated more accurately with the first spatial derivative of the potential (Rattay 2008). This must be taken into account when using the activating function to make predictions for ECS since many axons and dendrites terminate beneath the electrode. Finally, when calculating the activating function, bends in the axon must be considered. Cortical fibers typically bend and meander, especially close to the boundary between the gray matter and white matter, where the fibers curve towards the direction of the fiber tracts in the white matter. A bend in an axon can affect the threshold stimulus amplitude and the location of action potential initiation (Tranchina and Nicholson 1986; Struijk et al. 1993; Schnabel and Struijk 1999). To account for bends in an axon, the activating function must be calculated along the paths of the axon. Despite its limitation, the activating function still provides a quick and useful approximation to the response of fibers to extracellular stimulation without actually modeling the fibers.
Figure 5. Cable model of neuronal fiber.
Figure 6. Comparison of membrane voltage and spatial derivatives for a) point electrode and b) disc electrode delivering cathodic stimulation.
78
Figure 7. Flow of current during extracellular cathodic stimulation of an axon.

2D cross-section
Skull
Gray matter
3D model
Dura Mater CSF
Substrate of lead Electrode
Precentral Gyrus
Precentral Sulcus
Central Sulcus
12 cm
Electrode
White matter
15 mm 14 mm
12
cm
6 cm
3.75 mm
40 mm
Figure 8. 3D extruded slab model of ECS. a) 2D cross section of precentral gyrus and surrounding anatomy. b) Cross section was extruded into 3D. c) Electrode (Northstar Neuroscience, Inc. Seattle) was modeled and placed above the dura mater.
First, we will examine the results of monopolar anodic stimulation applied directly over the middle of the gyrus. The outputs from the model included the distributions of electric potential and current density in the cortex. The distributions of current density describe where current goes and how much current flows through the cortex (Figure 9). Current flowed out of the anode, through the dura, and into the CSF. From the CSF, the current either entered the gyrus or continued through the CSF. The current in the gyrus either continued further down into the brain or exited the sides of the gyrus, across the sulcus, and back into the cortex on the other side. When current reached the white matter, it tended to flow down to the bottom of the model, a result of the anisotropic conductivity of the white matter. The magnitude of the current density was greatest in the crown of the gyrus beneath the electrode, while the current density was much lower along the banks of the sulcus. From the distributions of the electric potential in the cortex, the distributions of the activating function were calculated in different directions. Recall that the activating function is the second spatial derivative of the extracellular potential along a fiber. The activating function predicts the response of fibers oriented in the same direction in which it is calculated. Thus, the activating function was calculated in the directions in which many of the fibers in the cortex are oriented: perpendicular and parallel to the surface of the cortex. Neurons oriented perpendicular to the cortical surface include those with axons projecting from one cortical layer to another or from the cortex to subcortical structures, and parallel
79
neurons include cortico-cortical interneurons that synapse onto neurons within the same layer.
Figure 9. Distribution of current density generated by monopolar anodic stimulation. Direction of the flow of current is indicated by the arrows. The magnitude of the current density is represented by the color scale.
Figure 10. Distribution of activating function generated by monopolar anodic stimulation. Activating function calculated a) perpendicular and b) parallel to surface.
The activating function calculated in the perpendicular direction beneath the anode was positive (Figure 10a). This indicates that perpendicular fibers beneath the electrode are predicted to be depolarized by anodic stimulation. Beyond the perimeter of the electrode, the activating function was negative. On the other hand, the activating function calculated in the
80
parallel directions was negative beneath the electrode (Figure 10b). The activating function reached a minimum directly beneath the stimulating contact on the surface of the precentral gyrus, while on either side of this minimum the activating function was positive. The triphasic response of these distributions resembles that of the activating function generated by a point source described earlier (Figure 6a). Since the model is linear with respect to voltage, switching from anodic to cathodic stimulation would flip the sign of the activating function. Then, instead of being hyperpolarized, parallel fibers beneath the cathode would be depolarized. Thus, the model predicts that monopolar anodic stimulation would depolarize perpendicular fibers beneath the anode and monopolar cathodic stimulation would depolarize parallel fibers beneath the cathode. F-Influence of Electrode Parameters It is important to analyze how changes in the placement and geometry of the electrode affect the distributions of current density and activating function. First, we will consider how bipolar stimulation differs from monopolar stimulation. From the first model, it was shown that monopolar anodic stimulation depolarized perpendicular fibers beneath the anode and monopolar cathodic stimulation depolarized parallel fibers beneath the cathode. When bipolar stimulation was delivered, both orientations of fibers were depolarized: perpendicular fibers beneath the anode and parallel fibers beneath the cathode (Figure 11c-d). If the primary targets of ECS are neurons oriented in a particular direction, then bipolar stimulation may be ineffective.
Figure 11. Bipolar stimulation. a) Slices were made through the length of the electrode. b) Current flowed from anode to cathode, as indicated by arrows. The magnitude of the current density (color scale) was greatest beneath the two contacts. c)-d) Activating function was calculated perpendicular (c) and parallel (d) to surface of cortex (indicated by arrows). Areas of greatest stimulation were directly beneath the contacts, not between the contacts.
81
A common misconception is that bipolar stimulation will target the area of the cortex between the two contacts because current flows directly from anode (+) to cathode (-) and through this region. Although current flowed between the anode (+) and cathode (-) (Figure 11b), the areas where the activating function was greatest in magnitude were directly beneath the contacts (Figure 11 c-d): exactly in the middle of the anode and cathode, the activating function was 0. These results can be thought of in terms of superposition: the distribution of the activating function during bipolar stimulation was the sum of the distribution generated from monopolar anodic stimulation through one contact and the distribution generated from monopolar cathodic stimulation through the other contact. Between the anode and cathode, the distributions of the activating function were opposite in sign, causing destructive interference and reducing the overall magnitude of the activating function. If bipolar stimulation caused a reduction in the activating function, then one might think that delivering the same polarity of stimulation through both contacts might increase the magnitude of the activating function and possibly target the region between the contacts. However, the distance between the contacts was so large that little interaction occurred between the distributions of activating function generated by each contact. As a result, the regions where the activating function was greatest in magnitude were still directly beneath the contacts. These results indicate that the primary targets of stimulation during ECS will likely be directly beneath the stimulating contacts, and stimulation should not be delivered through a contact unless the area beneath it is an intended target.
Figure 12. Electrode placed perpendicular to gyrus, bipolar stimulation: distributions of activating function calculated in perpendicular direction (inset). a) With anode directly over precentral gyrus, adjacent gyrus was stimulated by cathode. b) With contacts on electrode straddling gyrus, neurons on adjacent gyri received more stimulation than those on precentral gyrus.
Stimulation can also be delivered with the electrode oriented perpendicular to the precentral gyrus. One technique that has been used during ECS is to straddle the target area of the cortex between the anode and cathode during bipolar stimulation (Canavero and Bonicalzi 2007a,b). The simulation of ECS with the electrode in this position demonstrated that the adjacent gyri were stimulated but the precentral gyrus was largely unaffected (Figure
82
12b), which was expected considering the results of bipolar stimulation described previously. Instead of straddling the gyrus, the electrode could be positioned with the anode directly over the gyrus (Canavero and Bonicalzi 2007a,b). In this case, the distribution of the activating function in the precentral gyrus was virtually identical to the distribution in the baseline model (Figure 12a). However, the adjacent gyrus also received significant stimulation. Thus, not only would there be no apparent benefit over the baseline model, but there would be unknown consequences from stimulating an unintended target on the adjacent gyrus.
2-Properties of Neuronal Stimulation Now we turn to the second part of our two-piece problem: the effect of the fields generated by ECS on cortical neurons. A-Pulse Duration The stimulus amplitude required to trigger an action potential in a neuron depends on the duration of the stimulus pulse. The relationship between the threshold current (Ith) and the pulse width (PW) is called the strength-duration relationship and can be described by the following equation (Weiss 1901):
I th = I RH (1 +
TCH ) PW
(7)
where IRH is the rheobase current, the threshold current for a pulse of infinite duration (in practice, the minimum current for excitation at long PWs); and TCH is the chronaxie, or the PW at which the threshold current is twice IRH (Figure 13).
Figure 13. Strength-duration curve.
Cell bodies have chronaxies in the 1-10 s range, large myelinated fibers in the 30-200 s range and small myelinated fibers in the 200-700 s range (Ranck 1975). However,
83
chronaxie may not be sensitive to the neural element stimulated and is similar for axons of passage and local cells during extracellular stimulation, since this latter results in action potential initiation in the axon even for electrodes positioned near the cell body (McIntyre and Grill 2000). B-Efficiency of Stimulation Clearly, two important factors to consider when selecting stimulation parameters for implantable stimulators are the degree of therapeutic effect and the severity of side effects. However, there are other important considerations, such as the power efficiency, charge efficiency, and energy efficiency. Power efficiency, or the maximum power required per stimulus pulse, should be considered since the maximum power that can be generated by a power source (battery) is directly proportional to its size. The charge efficiency, or charge delivered per pulse, is another factor to consider, because excessive charge and charge density can lead to neural or tissue damage (Pudenz et al. 1975; Yuen et al. 1981; McCreery et al. 1990). In addition, energy efficiency is a significant consideration; the amount of energy consumed per pulse affects the battery life of the pulse generator, and the replacement of the pulse generator requires expensive and invasive surgery. The strength-duration relationship (7) can be used to analyze how these three measures of efficiency vary with PW. The power is proportional to the square of the current, so the threshold power is
Pth I RH (1 +
TCH 2 ) PW
(8)
According to the equation, the power required to reach threshold decreases as PW increases. In other words, to be power efficient, long pulse widths should be used. The charge delivered during a rectangular pulse is equal to the current multiplied by the pulse width, and the charge-duration relationship is Qth = I RH ( PW + TCH ) (9) The equation predicts that charge is minimized when PW = 0. Obviously, this is nonsensical since PW cannot be 0, but we can presume that charge efficiency (and the therapeutic window) increases as PW decreases. The energy delivered during a rectangular pulse is equal to the power multiplied by PW, and from equation (8) the energy-duration relationship is
E th I RH (1 +
TCH 2 ) PW PW
(10)
An analysis of this function reveals that energy is minimized when PW = TCH (Kroll 1993). If stimulation is most charge-efficient with short PW, most power efficient with long PW, and most energy efficient with PW = TCH, then what PW should be used? To make this choice, one must consider what is most important in each situation. For example, if charge efficiency is the most significant factor, then a short PW should be used. If energy and charge efficiency are equally important, then a PW somewhere between 0 and TCH should be used. In the following section we will show that PW not only influences efficiency, but also influences the selectivity of stimulation.
84
C-Spatial and Fiber Diameter Selectivity One of the goals of neural stimulation is to stimulate a specific population of neurons, grouped by either spatial location or fiber diameter, without activating other populations. This ability is known as selectivity, and using principles that have already been discussed, we will analyze how to achieve spatial and fiber diameter selectivity. During ECS, it is often desirable to activate neurons only in a specific region of the cortex, for example, the hand area of the motor cortex. To achieve this spatial selectivity, it is important to understand the relationship between threshold current (Ith) and electrode-fiber distance (r):
I th = I 0 + Kr 2
(11)
where Io and K are constants (Ranck 1975). There are two reasons for this currentdistance relationship. First, according to equation (2), the potential in a conductive medium is inversely proportional to the distance from the electrode. Second, as the electrode-fiber distance increases, the potential along the length of the fiber becomes more uniform. Both factors contribute to the conclusion that a fiber located close to the electrode will be activated before a fiber of equal diameter located further away. Since different classes of cortical neurons have fibers with varying diameters, it may also be desirable to target selectively fibers with different diameters. The relationship between Ith and fiber diameter (D) is
I th = I D +
a D
(12)
where ID and a are constants. This current-diameter relationship is due to the differences in the internodal spacing between fibers of different diameters: as fiber diameter increases, the spaces between nodes of Ranvier also increase (McNeal 1976). As a result, the difference in potential between two nodes increases, and more importantly, the second difference (i.e. activating function) also increases. Thus, if two fibers are equidistant from the electrode, the fiber with the larger diameter will be activated first. How do spatial and fiber-diameter selectivity translate to ECS? An initial reaction might be that neurons closest to the surface would be activated before neurons located in deeper layers because of the current-distance relationship. However, neurons with larger diameters, such as those of pyramidal cells, are found in deeper layers of the cortex, so they may actually have lower thresholds than more superficial neurons with thinner fibers. Thus, it is unclear from the current-distance and current-diameter relationships which neurons are targeted during ECS. D-Electrode Polarity In the 3D model of ECS, polarity had different effects on perpendicular and parallel fibers. Specifically, monopolar anodic stimulation depolarized perpendicular fibers beneath the anode, and monopolar cathodic stimulation depolarized parallel fibers beneath the cathode. However, these predictions were made using the activating function, the limitations of which were described previously in this chapter. Nonetheless, the therapeutic results of cortical stimulation have been shown to vary with polarity (Canavero and Bonicalzi 2007a,b),
85
indicating that different populations of neurons were activated for anodic and cathodic stimulation. E-Biphasic Pulses and Prepulses Rather than deliver a single anodic or cathodic pulse (monophasic stimulation), a stimulation pulse will often be preceded or followed by an additional pulse of opposite polarity (biphasic stimulation). One reason for the additional pulse is to achieve charge balance. That is, any charge that was transferred during the first pulse can be returned by the second pulse, thus reducing the risk of electrode corrosion and tissue damage. Another purpose for using biphasic pulses is to affect selectivity. As described in the previous section, anodic stimulation will likely activate different populations of neurons than cathodic stimulation. If the prepulse is symmetric to the main pulse (equal in PW and amplitude, opposite polarity), then selectivity is expected to decrease; whatever selectivity achieved through the anodic pulse is negated by the cathodic pulse, and vice-versa. In contrast, asymmetric prepulses (longer PW and smaller amplitude than main pulse) should improve selectivity. Neural elements that are hyperpolarized by the prepulse will become hyperexcitable, while those that are depolarized will accommodate. Also, the change in excitability for fibers with lower thresholds (i.e. closer to the electrode, larger diameters) will be greater than the change for fibers with larger thresholds. Then, when the main pulse is delivered, the difference in thresholds for the different populations of cortical neurons will increase, and thus, selectivity is improved. These effects were demonstrated in studies of stimulation in the peripheral nervous system (Gorman and Mortimer 1983; Grill and Mortimer 1994) as well as in the central nervous system (McIntyre and Grill 2000). F-Frequency The frequency of stimulation is the rate at which stimulus pulses are delivered (Hz, or pulses/sec). Different neuronal populations may be recruited at different frequencies. Frequencies between 30-100Hz are usually considered excitatory, above 100 Hz likely inhibitory. The optimal frequency during ECS likely varies with the disorder being treated. In deep brain stimulation (DBS) for the treatment of movement disorders, stimulation at high frequencies (>100 Hz) is more efficacious than at lower frequencies and is believed to mask the information content of pathological signals (Grill et al. 2004). The effectiveness of ECS on movement disorders may also depend on frequency in the same manner. However, since other disorders have different etiologies, the efficacy of ECS for other disorders may depend on frequency differently. Further, frequency may differentially affect indirect (synaptic) and direct (field-mediated) responses to stimulation. A common misconception is that the frequency of stimulation somehow affects the spatial spread of current. For example, when an increase in frequency results in an increase in some response, it may be attributed to the current directly stimulating a population of neurons located further away. This view is incorrect because the potentials generated in the body by a stimulus pulse do not rely on the previous state of the potentials in the body (quasi-static approximation). The discussion in this chapter has focused primarily on direct stimulation of neurons. That is, it has been assumed that the only sources of stimulation were the electric fields
86
generated by an electrode. However, the neurons of the brain are highly interconnected and can be stimulated indirectly by synaptic inputs. As a result, if the goal of ECS is to activate a specific population of neurons, then this goal may be achieved by stimulating the population directly, by activating the presynaptic neurons to stimulate the population, or a combination of direct and indirect effects. Although the frequency of stimulation does not affect the direct stimulation of the neurons in the cortex, it can affect indirect stimulation through a process known as temporal summation. Suppose a single stimulus pulse generates an action potential in a neuron, which causes an excitatory postsynaptic potential (EPSP) in the postsynaptic target. Postsynaptic potentials generated by individual synapses in the brain are far too small to elicit an action potential, and they are also transient, lasting tens of milliseconds or less. If another stimulus pulse is delivered before the effects of the first EPSP have disappeared, then the second EPSP will add onto the first EPSP, generating a cumulative postsynaptic potential that is greater than an individual EPSP. As the frequency of stimulation increases, more and more EPSPs sum up and can lead to tonic subthreshold depolarization of the target (Figure 14). Although tonic depolarization does not lead to activation of the target directly, it can facilitate activation indirectly. A class of receptors known as NMDA receptors produces EPSPs when activated, but the response is dependent on voltage. When the postsynaptic membrane is at resting potential, the channel of the NMDA receptor is blocked by magnesium ions. As a result, EPSPs cannot be produced by the NMDA receptor. However, if the membrane is depolarized, then magnesium is forced out of the channels, and the NMDA receptors can generate EPSPs. If frequency is increased even further, then temporal summation can generate a cumulative EPSP exceeding threshold, and an action potential is initiated. However, increasing the frequency requires more energy and depletes batteries more quickly. Temporal summation also applies to inhibitory postsynaptic potentials (IPSP), which can lead to tonic hyperpolarization of the target neuron. Thus, increasing the frequency of stimulation can affect a greater number of neurons without stimulating additional neurons directly.
Figure 14. Temporal summation. Each line on the top set of graphs represents a presynaptic action potential. Every time a presynaptic axon fires an action potential, a postsynaptic potential is generated. Increasing frequency allows postsynaptic potentials to add together, creating a larger aggregate postsynaptic potential.
87
CONCLUSION
Even without knowing the exact mechanisms of the effects of ECS, there are guidelines that can be followed to make ECS more effective and efficient. 1. The electrode should be positioned close to the targeted area of the cortex. 2. The stimulation should not be delivered through a contact located near a part of the cortex that is not the intended target. For instance, straddling the target gyrus between the anode and cathode and delivering bipolar stimulation will not target the neurons in the gyrus, but, rather, the neurons in the adjacent gyri will receive greater stimulation than the target. 3. If more than one combination of stimulation parameters produces identical therapeutic effects, then the choice of stimulation parameters should be made using an analysis of efficiency. Depending on the application of ECS, the importance of power, charge, and energy efficiency will vary. 4. One must be mindful of the geometry of the anatomy and electrode. The optimal stimulation parameters for ECS delivered over one area of the cortex may not be optimal when delivered over a different part of the cortex or with a different electrode.
REFERENCES
Bossetti, C. A., M. J. Birdno, et al. Analysis of the quasi-static approximation for calculating potentials generated by neural stimulation. J Neural Eng 2008: 5: 44-53. Canavero S, Bonicalzi V. Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007a, 27-36. Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, 223-232. DeFelipe, J. and I. Farinas. The pyramidal neuron of the cerebral cortex: morphological and chemical characteristics of the synaptic inputs. Prog Neurobiol 1992; 39: 563-607. DeFelipe, J., S. H. Hendry, et al. Variability in the terminations of GABAergic chandelier cell axons on initial segments of pyramidal cell axons in the monkey sensory-motor cortex. J Comp Neurol 1985; 231: 364-84. Feldman, M. L. Morphology of the Neocortical Pyramidal Neuron. Cerebral Cortex. A. Peters and E. G. Jones. New York, Plenum Press. 1984. Geddes, L. A. and L. E. Baker .The specific resistance of biological material--a compendium of data for the biomedical engineer and physiologist. Med Biol Eng 1967; 5: 271-93. Gorman, P. H. and J. T. Mortimer. The effect of stimulus parameters on the recruitment characteristics of direct nerve stimulation. IEEE Trans Biomed Eng 1983; 30: 407-14. Grill, W. M. and J. D. Mortimer. Effect of stimulus pulse-width on spatial selectivity of neural stimulation. Proc. 16th Int. Conf. IEEE-EMBS,1994.
88
Grill, W. M., A. N. Snyder, et al. Deep brain stimulation creates an informational lesion of the stimulated nucleus. Neuroreport 2004; 15: 1137-40. Haueisen, J., D. S. Tuch, et al. The influence of brain tissue anisotropy on human EEG and MEG. Neuroimage 2002; 15: 159-66. Ichikawa, M., K. Arissian, et al. Distribution of corticocortical and thalamocortical synapses on identified motor cortical neurons in the cat: Golgi, electron microscopic and degeneration study. Brain Res 1985; 345: 87-101. Jones, E. G. and S. H. Hendry.Basket Cells. Cerebral Cortex. A. Peters and E. G. Jones (Eds) New York, Plenum Press, 1984. Kaneko, T., M. A. Caria, et al. Information processing within the motor cortex. II. Intracortical connections between neurons receiving somatosensory cortical input and motor output neurons of the cortex. J Comp Neurol 1994; 345: 172-84. Kaneko, T., R. Cho, et al. Predominant information transfer from layer III pyramidal neurons to corticospinal neurons. J Comp Neurol 2000; 423: 52-65. Kimura, A., M. A. Caria, et al. Long-term potentiation within the cat motor cortex. Neuroreport 1994; 5: 2372-6. Kroll, M. W. A minimal model of the monophasic defibrillation pulse. Pacing Clin Electrophysiol 1993; 16(4 Pt 1): 769-77. Kuncel, A. M. and W. M. Grill. Selection of stimulus parameters for deep brain stimulation. Clin Neurophysiol 2004; 115: 2431-41. Li, C. L., A. F. Bak, et al. Specific resistivity of the cerebral cortex and white matter. Exp Neurol 1968; 20: 544-57. Lund, J. S. Spiny Stellate Neurons. Cerebral Cortex. A. Peters and E. G. Jones. New York, Plenum Press, 1984 Manola, L., B. H. Roelofsen, et al. Modelling motor cortex stimulation for chronic pain control: electrical potential field, activating functions and responses of simple nerve fibre models. Med Biol Eng Comput 2005; 43: 335-43. Markram, H., M. Toledo-Rodriguez, et al. Interneurons of the neocortical inhibitory system. Nat Rev Neurosci 2004; 5: 793-807. McCreery, D. B., W. F. Agnew, et al. Charge density and charge per phase as cofactors in neural injury induced by electrical stimulation. IEEE Trans Biomed Eng 1990; 37: 9961001. McIntyre, C. C. and W. M. Grill. Selective microstimulation of central nervous system neurons. Ann Biomed Eng 2000; 28: 219-33. McNeal, D. R. Analysis of a model for excitation of myelinated nerve. IEEE Trans Biomed Eng 1976; 23: 329-37. Nicholson, P. W. Specific impedance of cerebral white matter. Exp Neurol 1965; 13: 386401. Nolte, J. The Human Brain: An Introduction to Its Functional Anatomy. St. Louis, Mosby, 2002 Peters, A. and I. R. Kaiserman-Abramof. The small pyramidal neuron of the rat cerebral cortex. The perikaryon, dendrites and spines. Am J Anat 1970; 127(4): 321-55. Peters, A., S. L. Palay, et al. The Fine Structure of the Nervous System: Neurons and Their Supporting Cells. New York, Oxford University Press, 1991
89
Plonsey, R. and R. C. Barr. Bioelectricity: A Quantitative Approach. New York, Kluwer Academic/Plenum Publishers, 2000 Pudenz, R. H., L. A. Bullara, et al. Electrical stimulation of the brain. III. The neural damage model. Surg Neurol 1975; 4: 389-400. Ranck, J. B., Jr. Specific impedance of rabbit cerebral cortex. Exp Neurol 1963; 7: 144-52. Ranck, J. B., Jr. Which elements are excited in electrical stimulation of mammalian central nervous system: a review. Brain Res 1975; 98: 417-40. Rattay, F. Analysis of models for extracellular fiber stimulation. IEEE Trans Biomed Eng 1989; 36: 676-82. Rattay, F. Current distance relations for fiber stimulation with pointsources. IEEE Trans Biomed Eng 2008; 55: 1122-7. Rubinstein, J. T., F. A. Spelman, et al. Current density profiles of surface mounted and recessed electrodes for neural prostheses. IEEE Trans Biomed Eng 1987; 34: 864-75. Schnabel, V. and J. J. Struijk. Magnetic and electrical stimulation of undulating nerve fibres: a simulation study. Med Biol Eng Comput 1999; 37: 704-9. Sloper, J. J. An electron microscope study of the termination of afferent connections to the primate motor cortex. J Neurocytol 1973; 2: 361-8. Struijk, J. J., J. Holsheimer, et al. Excitation of dorsal root fibers in spinal cord stimulation: a theoretical study. IEEE Trans Biomed Eng 1993; 40: 632-9. Tranchina, D. and C. Nicholson. A model for the polarization of neurons by extrinsically applied electric fields. Biophys J 1986; 50: 1139-56. Warman, E. N., W. M. Grill, et al. Modeling the effects of electric fields on nerve fibers: determination of excitation thresholds. IEEE Trans Biomed Eng 1992; 39: 1244-54. Weiss, G. Sur la possibilite de rendre comparables entre eux les appareils servant a l'excitation electrique. Arch. Ital. Biol. 1901; 35: 413-446. Wiley, J. D. and J. G. Webster. Analysis and control of the current distribution under circular dispersive electrodes. IEEE Trans Biomed Eng 1982; 29: 381-5. Wongsarnpigoon A, Grill WM. Computational modeling of epidural cortical stimulation. J Neural Eng 2008; 5: 443-454. Yuen, T. G., W. F. Agnew, et al. Histological evaluation of neural damage from electrical stimulation: considerations for the selection of parameters for clinical application. Neurosurgery 1981; 9: 292-9.
Chapter 7
A NEUROENGINEERING MODEL OF EXTRADURAL CORTICAL STIMULATION

Ljubomir Manola1 and Jan Holsheimer2
2
Boston Scientific International, Nanterre, France; Biomedical Signals & Systems, University of Twente, Enschede, The Netherlands.
INTRODUCTION
In this chapter, a computer model of how ECS might work is presented and discussed. A word of caution is warranted. Models like the present one are still too coarse and rely on poorly defined electrical-geometrical and cytomorphometric values in the human cortex (e.g. fiber diameter distribution, fiber orientation, degree of myelination, ion channel density, neuron position,), to which a large inter/intra-subject variability may be added. Duration and amplitude of stimuli can vary markedly during the parameter-search stage and these models offer nothing in this regard. Finally, since the gyri are curvilinear, results of excitation change along the course of the folding pattern.
THE ECS STIMULATING SYSTEM

An ECS stimulating system comprises one or two paddle electrodes with 4 (or more) contacts each, extension cables and an implantable pulse generator (IPG). The IPG generates an electrical stimulus pulse which is conducted via the extension cables and directed towards the cortical surface through those contacts programmed as a cathode or an anode. Usually, a 4-contact electrode array (circular electrode contacts, 4 mm diameter, 10 mm center-to-center spacing, insulating paddle about 2 mm thick) is placed extradurally, but two 4-contact electrode arrays may be needed if a larger body area is affected by pain.
92
Ljubomir Manola and Jan Holsheimer
The electrical stimulus comprises a sequence of equally-spaced pulses. In chronic stimulation, the cathode-anode contact combination delivering the pulses, the amplitude, pulse rate and pulse width are chosen empirically, based on patients reports of symptom relief and side effects. Generally, a bipolar contact combination with the cathode over the MI target area and the anode over CS, or farther in MI or the primary somatosensory area (SI), is used. The array can be positioned fully on SI (Canavero et al 2002, 2007a). The amplitude is usually set at less than 50% of the motor threshold. The reported ranges of stimulation parameters are 1-10 Volts (pulse amplitude), 5-130 Hz (pulse rate) and 60-450 sec (pulse width) (Canavero and Bonicalzi 2002, 2007a,b). A cycling stimulation mode is often applied to extend battery longevity and, theoretically, to prevent accommodation to the stimulus, but a continuous mode may be necessary in some patients. The OFF cycling phase is set based on the duration of the residual therapeutic response when the stimulation is stopped.
Figure 1. Anterior-posterior cross-section through precentral gyrus and surrounding precentral and postcentral sulci. Anatomical compartments are labeled. Mean values for compartment dimensions as used in the model are given in mm (with permission from Elsevier).
MODELING METHOD
Our model of MI ECS (Figure 3) consists of: a 3-dimensional volume conductor model within which the stimulus-imposed electrical potential field is calculated and a nerve cell model which predicts the cellular response to the imposed electrical field.
A Neuroengineering Model of Extradural Cortical Stimulation 1. Volume Conductor Model
93
The volume conductor model represents a 3D model of the relevant structures of the motor cortex region (Box 1). BOX 1. Anatomy of the motor cortex
The brain is protected by the skull, a thin layer of extradural fat, dura mater, arachnoid and a layer of cerebrospinal fluid (CSF), while its surface is intimately encapsulated by a very thin, but highly vascular pia mater. The outer layer of the brain, the cortex, is the gray matter (GM), mainly consisting of cell bodies, dendrites and synapses. Underneath is the white matter (WM), consisting of cortical afferent and efferent nerve fibers. In order to increase its surface and thereby its functional capacity, the human cortex is extremely curved and folded, creating gyri (convex folds) and sulci (fissures). The primary motor cortex (MI: Brodmanns area 4, BA4) is hosted by the precentral gyrus (PCG). This gyrus is delineated by the precentral sulcus (PCS), the central sulcus (CS), the lateral and the interhemispheric fissure on the anterior, posterior, inferior and superior sides respectively (see Figure 1 for a schematic cross-section of the PCG along the antero-posterior axis and corresponding average dimensions). At the inferior side, near the lateral fissure, BA4 is confined almost completely to the anterior wall of the CS. More superiorly, MI extends over PCG with maximal extension superiorly, near the interhemispheric fissure. Brodmann area 6 (premotor cortex) constitutes the complementary part of PCG. MI has a columnar organization, each column being perpendicular to the surface of the cortex (Mountcastle 1997). The Penfield motor homunculus depicts the topographic organization in a simplified way (Penfield and Boldrey 1937). The face is represented most inferiorly in the motor cortex, succeeded superiorly by the hand representation, the upper limb, trunk and finally the lower limb, the latter often being represented in the wall of the interhemispheric fissure. The cortical GM is fairly complex and has a laminar (layered) composition, with the laminae retaining a parallel orientation in respect to the (folded) surface of the cortex. Each of the six MI laminae has a different composition (Manola et al 2005). Nerve fibers from ventro-anterior (VA) and ventro-lateral (VL) thalamic nuclei ascend up to lamina IV of MI where they bifurcate into the laminar plane, thus parallel to the cortical surface. These nerve fibers ascending into the cortex are named thalamo-cortical fibers. Conversely, cortico-thalamic nerve fibers originate in pyramidal cells in lamina V and VI and project onto the VA-VL complex of the thalamus, thus closing the corticothalamic neural loop. The largest pyramidal cells (a.k.a. Betz cells) are exclusively found in BA4 lamina V. Their myelinated nerve fibers (axons) descend and constitute the cortico-spinal tract. All pyramidal cells are oriented such as to follow the curvature of the cortex, ensuring that they project their long apical dendritic trees perpendicular to the cortical layers, up to lamina I. The dendritic trees serve as a converging point for inputs from various cortical layers. Connections are quite abundant within a cortical column and far less in between columns (Gatter and Powell 1978). A scheme of afferents and efferents of MI is shown in Figure 2. Little documentation is available about morphometry (caliber, length, myelination etc.), synaptic connectivity, ionic channels distribution and other features of cortical neural elements. The largest caliber of nerve fibers observed in the human pyramidal tract was 13m (von Keyserlingk and Schram 1984). Because Betz cells are the largest, it is expected that their axons, forming the cortico-spinal tract, would have the largest diameter. Since those Betz cells innervating the lower limb are larger than e.g. those innervating the face, their axons are most likely larger in diameter. A better understanding of these aspects is essential to improve the predictive value of modeling as will be discussed (see relevant chapters in Paxinos and Mai 2004 for more anatomical details). In sum, MI represents a complex and essential information processing and transmitting network, closely connected to several cerebral structures.
A cross-section through the model in the anterior-posterior direction is shown in Figure 4. All anatomical structures represented in Figure 1 are identified in this cross-section, including PCG and CS occupying the center of the model. Dimensions of the model compartments were chosen to match their mean measured values (see Figure 1 and Manola et al 2005). The physical size of the model is 66x43x57 mm in the anterior-posterior, depth and superior-inferior direction, respectively. The electrical properties of the anatomical structures are represented by their specific conductivities (see Manola et al 2005).
94
Ljubomir Manola and Jan Holsheimer Table 1. Modeled contact combinations Polarity and voltage of the contact over PCG - 1.0 Volt + 1.0 Volt n/a n/a - 0.5 Volt + 0.5 Volt Polarity and voltage of the contact over CS n/a n/a - 1.0 Volt + 1.0 Volt + 0.5 Volt - 0.5 Volt
Name Cathode (-) over PCG Anode (+) over PCG Cathode (-) over CS Anode (+) over CS Bipole w/ cathode over PCG Bipole w/ anode over PCG
Type Monopolar Monopolar Monopolar Monopolar Bipolar Bipolar
Figure 2. Weigert staining of the cortical gray matter showing myelinated nerve fibers and an approximate delination of the 6 cortical laminae. On the right-hand side, neural connections between different cortical laminae, ventro-anterior, ventro-lateral thalamic nuclei and spinal cord are shown. Triangles represent pyramidal cells of lamina V and VI. Arrows indicate natural (orthodromic) direction of information propagation. Cortico-cortical nerve fibers are not shown: they follow a similar course as thalamo-cortical fibers, except that they are not incoming via thalamus (with permission from Elsevier).
The cross-sections along the model axis perpendicular to the plane shown in Figure 4 are identical. This is based on the assumption that the geometry of PCG and the surrounding sulci does not vary from the superior to the inferior side, except where the electrode is located. These model planes also contain the electrode contacts and an insulation layer (electrode paddle). The effect of the additional mass of the paddle on the local thickness of CSF layer is also incorporated in the model (Figure 4 represents such a cross-section). One or two electrode contacts were modeled as a cathode or an anode (see Table 1).
A Neuroengineering Model of Extradural Cortical Stimulation
95
One contact was centered over PCG and the other over CS (see Figure 4 for electrode contact positions). The combinations defined in Table 1 allow to test effects of all contact polarity combinations including a bipole with the axis perpendicular to CS. Given the voltage applied to an electrode contact (see Table 1), the electrical potential field created in the surrounding tissues can be calculated using numerical methods (Box 2). The electrical field represents the means by which a neurostimulator interacts with the neural tissue.
Figure 3. A sequence of steps followed to obtain modeling results.
Box 2. Calculation of the electrical potential field

The model space comprises a 3D rectangular grid with variable spacing between grid points. The sum of all grid spacings corresponds to the total size of the model. Our MI ECS model contains 122x74x81 grid points in the three orthogonal directions, giving a total model size of 66x43x57 mm. The cubic elements defined by the grid are assigned a specific conductivity value [S/m] corresponding to the tissue they represent. The grid points (vertices of the cubes) are those locations in the model where the solution of the electrical potential field () is calculated. The initial equation to solve the electrical potential field is Laplace equation: = 0 This equation can be discretized using the finite difference method (Venner and Lubrecht 2000):
x +1, y , z 2 x , y , z + x 1, y , z x , y +1, z 2 x , y , z + x , y 1, z +y + 2 2 ( x ) ( y )
x , y , z +1 2 x , y , z + x , y , z 1 = 0 2 ( z )
+z where:
x,y,z represent local electrical conductivities in three orthogonal directions (different values if the medium is anisotropic) x,y,z represents the unknown electrical potential at a grid point with (x,y,z) coordinates denominators represent squared values of the local grid resolution in the three orthogonal directions The result of discretization for each grid point can be written as a matrix system of equations:
96
A = d
with matrix A containing conductivity and grid spacing information, unknown electrical field potentials at grid vertices and d Dirichlet boundary conditions. The Dirichlet boundary conditions define potentials at the model boundary and electrode contact points to be 0 Volts and an assigned contact voltage, respectively. The system of equations is solved using an iterative Gauss-Seidel relaxation method (Venner and Lubrecht 2000). As an initial guess for , an under- and an overestimated value is assumed and the system of equations is solved independently for each of the two initial guesses. When the average difference between the two solutions at the grid points is below 0.0001 Volt, the iterations are stopped and the solution for considered to be found.
Figure 4. An anterior-posterior cross-section of the model with anatomical compartments labeled. Locations of the electrode contacts over PCG and CS are also shown. The dashed line indicates a supposed location of the motor cortex on top of PCG and in the anterior wall of CS (with permission from Elsevier)
Figure 5. Modeled nerve cells: A a myelinated nerve fiber parallel to cortical laminae, E1, E2 and E3 pyramidal cells on top of PCG, in the lip of CS and in the anterior wall of CS, respectively. Each pyramidal cell comprises an apical dendrite, a cell body and a downward projecting myelinated nerve fiber (with permission from Elsevier).
A Neuroengineering Model of Extradural Cortical Stimulation 2. Models of the Neural Elements
97
We modeled myelinated nerve fibers (axons) as well as complete nerve cells (neurons) consisting of an apical dendritic tree, a cell body and a myelinated nerve fiber. The axon models represent afferent fiber collaterals parallel to the cortical surface (parallel fibers) and are found primarily in lamina I, IV, V and VI of MI (see Figure 2). The neurons represent pyramidal cells, whose cell bodies are found primarily in lamina V. The complex apical dendritic trees are represented by a simple dendrite extending up to lamina I perpendicular to the laminae and their axons projecting downward either towards thalamus or spinal cord or to other cortical areas. The orientation of these cells is variable to accommodate the curvature of the cortex, always remaining perpendicular to the cortical laminae (Manola et al 2005). For this reason, we modeled three pyramidal cells representing three distinct locations: on top of PCG, in the anterior lip of CS and in the anterior wall of CS. Figure 5 shows the relative location of a parallel nerve fiber (A) and three pyramidal cells (E1, E2, E3) with standard positions with respect to PCG and the electrode contacts. Depending on the shape and location of the electrical field with respect to a neural element, its intensity and duration, the neural membrane will get locally hyperpolarized or depolarized when a stimulus pulse is applied. Local hyperpolarization of the cell membrane indicates that this part cannot be excited under the influence of the imposed electrical field. Conversely, those parts of the membrane that get depolarized are likely candidates to respond with excitation (generation of an action potential, AP), if the given electrical field becomes strong (or lasts long) enough. The model of a neural element describes the resting state properties and the ion-channel kinetics of each membrane compartment and predicts their compound behavior as the neuron is exposed to the external electrical field. With the imposed electrical potential field (calculated in the volume conductor model) and a fixed duration of this field (pulse width), the nerve cell model calculates the time course of the polarization of each membrane compartment and predicts which part of the neuron membrane will get excited and at what stimulus intensity (excitation threshold). Box 3 gives more information on the calculations needed to predict the response of the nerve cell.
MODELING RESULTS
In this section, the main modeling results obtained so far (Manola et al 2005, 2007) are summarized. According to the modeling sequence presented in the previous section, the results are divided into those related to the stimulus-imposed electrical field and those related to the response of the modeled neurons.
1. Stimulus-imposed Electrical Field The stimulus-imposed electrical field exists during the stimulus pulse everywhere in the volume conductor surrounding the active electrode contacts. The magnitude of the field is largest in the region near the active contact and decays with increasing distance from it. This
98
field can be visualized by iso-potential or iso-current density lines. The iso-lines connect those points in the volume conductor cross-section with the same potential [V] or with the same magnitude of the current density vector [A/mm2], respectively. Box 3. Neuron model
A nerve cell is separated from the environment by a semi-permeable membrane. The membrane contains ion channels transporting specific ions from one side to the other if certain conditions are met e.g. sufficient local (de)polarization of the membrane in case of voltage-sensitive channels. In addition to these dynamic ion channels the membrane consists of a capacitance representing the double lipid layer. A myelinated nerve fiber (axon) is insulated by a myelin sheath, except at regular intervals by annular patches of membrane - nodes of Ranvier (see Figure I1). At these nodes, rich in ionic channels, ions are exchanged between the inside and the outside of the axon. The nodal current consists of a capacitive component (accounted for by a capacitor, Cmem, in the membrane model) and a resistive component (accounted for by a conductance, Gmem) (see Figure B1). Gmem of the node represents the conductance of the voltage-sensitive channels. Its value varies depending on the polarization state of the membrane that it models. The resistive current flowing across Gmem (Iresistive) comprises a sodium, a potassium and a leakage current, and is described by the Hodgkin-Huxley (1952) equation:
I resistive = I Na + + I K + + I Leak
This model of the node of Ranvier corresponds structurally to that of Frankenhaeuser and Huxley (1964), applied in McNeals model. The equations and parameters describing ionic currents were adapted to mimic human nerve fiber characteristics as measured by Schwarz et al. (1995) and implemented by Wesselink et al. (1999). In our model, we assumed that the internodal membrane is perfectly insulated by the myelin sheath. The internodal part of the nerve fiber has, therefore, been modeled as a simple resistor (Rint) corresponding to the axial resistance of the internodal segment. Alternating nodal and internodal compartments constitute the cable-like, McNeal (1976)-type nerve fiber model (see Figure B1). To construct a model of a (pyramidal) nerve cell, a model of the apical dendrite and cell body is attached to the myelinated axon model by a model of the axonal initial segment. Models of the cell body and apical dendrite are similar in structure to the cable-like model of the nerve fiber. The difference is that their membrane is considered to be passive (only few voltage-sensitive ionic channels). Therefore, the resistive current is reduced to a single component: I resistive = I Leak Conversely, the axonal initial segment was assumed to have the same ionic channel density and composition as a node of Ranvier. The difference between the two is the larger length of the initial segment. The schematics and the equivalent electrical circuit of a pyramidal neuron are shown in Figure B1. Based on Kirchoffs law, a differential equation linking the outward current over a membrane compartment with the intracellular axial currents originating from the adjacent compartments is derived for each membrane compartment along the cable-like neuron model:
C mem,i where:
dVmem ,i dt
+ I resistive ,i =
Vint,i +1 Vint,i R int,i:i +1
Vint,i 1 Vint,i R int,i:i 1
indices i-1, i, i+1 denote three adjacent membrane compartments of the model, Cmem, Rint and Iresistive represent the membrane compartment capacitance, the intracellular resistance between adjacent compartments and the resistive current component of the membrane compartment Vint represents the intracellular voltage (V) of the compartment location and is calculated as: Vint = Vext + Vrest + Vmem
with: Vext being the electrical field potential at the compartment location determined by the stimulus-imposed electrical field, Vrest being the resting membrane potential (equal for all compartments) and Vmem being the membrane potential deflection from the resting value. We used Matlab (Mathworks Inc., Natick, MA, USA) to calculate the electrical field potentials (Vext) at the exact
99
location of the membrane compartments in the model. This is done by applying a spline interpolation on the calculated potentials at nearby grid points of the volume conductor model. Next, the system of differential equations written for nerve cell model compartments is solved using numerical method ode15s incorporated in Simulink (Matlab). This yields the time course of Vmem at each membrane compartment. A typical time course of Vmem for several membrane compartments is given in Figure B1. Generation of an AP and its propagation to adjacent membrane compartments is demonstrated. By scaling the stimulus voltage, and thereby the stimulus-imposed electrical field, the excitation threshold (lowest voltage exciting the neuron) is determined for each neuron model. By comparing the Vmem response of all compartments it can also be determined at which compartment the initial excitation takes place (AP generation). In our models, we considered the site of initial excitation to be the compartment exhibiting the largest depolarization at the termination of the stimulus pulse.
Figure B1. Upper part: The morphometry of a pyramidal nerve cell including a myelinated nerve fiber. Bottom part: The corresponding electrical equivalent. Cmem capacitance of a membrane compartment, Gmem conductance of a membrane compartment, Rint axial resistance between two adjacent membrane compartments, Vext external voltage potential at the location of the corresponding membrane compartment. Arrows indicate an active membrane compartment, whereas their lack a passive one (cell body and dendrite) (with permission from Elsevier).
There is a layer of CSF in the vicinity of the electrode, just on the opposite side of the dura mater. Since the conductivity of the CSF is much higher than those of the other anatomical compartments, most current flows via the CSF from the anode to the cathode and only a small portion of the injected current invades the gray (and white) matter. The amount of current injected through an electrode to activate neurons directly is proportional to the square of the distance between the neurons and the electrode. This is expressed as: I =K2, where I is the current level, is distance, and K is the excitability constant (Tehovnik 1996).
100
Figure B2. Time course of the membrane voltage deflection from a resting value for several membrane compartments of a single nerve cell. During the pulse some compartments get depolarized (positive voltage deflection), others hyperpolarized (negative deflection). The compartment whose membrane voltage reaches threshold for excitation at the end of the pulse is considered site of excitation. Following that event, an action potential is generated and propagated to the adjacent compartments (with permission from Elsevier).
Therefore, if the CSF space is small, the current will penetrate deep into the cortex and underlying white matter, while increasing the load of impedance. On the other hand, if the CSF space is larger, it can reduce the current density in the cortex and white matter, while decreasing the load impedance. We estimated that, with the thickness of the CSF set at 3.1mm (being the mean value calculated from MRI scans), less than 60% of the injected current enters the cortex (Manola et al 2005). Figure 6A shows iso-current density lines in CSF and PCG below a single extradural electrode contact centered over the PCG and programmed as a cathode at -1V. Due to the inter subject variability of the CSF layer thickness and the expected effect of the mass of the extradural electrode on this thickness, a wide range of CSF thicknesses is expected. Figure 6A shows the current-density fields calculated in models with various CSF thicknesses below the electrode. With the applied voltage kept constant and the thickness of the highly-conductive CSF layer increased, the impedance [Ohms] decreased and the impressed current increased. The overall penetration of the current into PCG, however, was reduced dramatically (Figure 6A). More in detail, when the CSF space was increased from 0 to 2.5mm, the load impedance decreased by 28%, while the stimulation amplitude increased by 6.6 V for each millimeter of CSF. As a result, if the CSF layer thickness is greater than 2.5 mm, 1V cathodal cortical stimulation is not likely to affect the deep cortex: an increase of stimulation amplitude is required, though not desirable.
101
Figure 6. Iso-current density lines for different: a) CSF thickness (value indicated in mm) b) electrode contact location along the anterior-posterior axis. The stimulation amplitude was -1 V and the resulting current is given for each plot. The iso-current density lines were drawn in the range 4-75 A/mm2 at equal intervals. CS - central sulcus (with permission from Elsevier).
Similarly, Figure 6B depicts iso-current density lines relating to different electrode contact positions: centered over PCG, slightly posterior and centered over CS. As shown, the location of maximum current penetration into the motor cortex changes with the electrode contact location. Notably, the percentage current invading the cortex also changes, despite the constant voltage applied. Whereas the thickness of CSF and the contact location influence the percentage current penetrating PCG, we found no significant influence of the width and depth of the sulci, which were varied to match the range that had been measured in our laboratory (Manola et al 2007).
2. Activating Functions To predict which part(s) of a nerve cell membrane gets depolarized and which gets hyperpolarized in a particular stimulus-imposed field, the activating functions were calculated along the lines approximating the course of a fiber parallel to and perpendicular to an electrode over the PCG (Figure 5). They essentially represent the second derivative of the electrical potential along the fiber trajectory (Rattay 1986), assuming a uniform structure along the nerve fiber (as was assumed in Manola et al 2005). A positive value of the activating function indicates membrane depolarization, while a negative value indicates membrane hyperpolarization. The concept of the activating function (chapter 6) predicts that a parallel nerve fiber located in PCG (parallel to the plane of the extradural electrode) would get depolarized directly under a cathodal electrode contact, while a perpendicular nerve fiber (pointing towards a cathode) would get hyperpolarized. Conversely, a nerve fiber perpendicular to an extradural anode over PCG would get depolarized and could get excited at a sufficient stimulus magnitude (Manola et al 2005).
102
The predictions obtained by applying the concept of the activating function are in accordance with those obtained by calculations of the responses of neuron models to the stimulus-imposed field, as shown below.
Figure 7. Excitation thresholds [V] for A, E1-E3 nerve cells with varied diameter of the nerve fiber (5-15m) and for different active electrode contact combinations, as indicated on each plot (with permission from Elsevier).
3. Neuronal Responses Those parts of the neuron models where excitation takes place were identified and the excitation thresholds were determined using the method outlined in Box 3. The dependency of the excitation threshold on the diameter of myelinated nerve fibers (5-15 m range) is shown in Figure 7 for a parallel nerve fiber located 1.4 mm below the cortical surface (A) and three distinct pyramidal cells (E1-E3). The stimulus was applied by the contact combinations listed in Table 1 and with a fixed rectangular pulse duration (210sec). The
103
thickness of the CSF below the electrode was set at 1.1 mm, corresponding to the mean value of 3.1 mm as measured on MRI scans minus the thickness of the insulating electrode paddle (2.0 mm). As shown in Figure 7 a lower voltage is needed to excite a particular neuron if the diameter of its nerve fiber is larger. It is also shown that by varying the contact combination, polarity or location of the active contact(s), the excitation threshold of a nerve cell varies relative to other neurons. For all six contact combinations modeled, Table 2 lists the type of neuron being most and least excitable, corresponding to the lowest and highest excitation threshold, respectively. The same fiber caliber was assumed for all neuron types. Table 2. Most and least excitable nerve cells per modeled contact combination. Equal diameter of nerve fiber assumed Name Cathode over PCG Anode over PCG Cathode over CS Anode over CS Bipole w/ cathode over PCG Bipole w/ anode over PCG Most excitable A, E1,A A E2,A A,E2,E3 A,E1 Least excitable E1 E2,E3 E1,E2,E3 E1 E1 E2-E3
The site of initial excitation was always located at a node of Ranvier of the axon, regardless of the stimulation conditions and the type of neuron. In Figure 8, arrows and brackets are used to indicate the sites of excitation. Brackets are used to show that some adjacent nodes of Ranvier often have a similar level of polarization at the end of the stimulus pulse, making it difficult to decide on the exact node of initial excitation.
DISCUSSION
1. Control of Stimulus-imposed Electrical Field The modeling results of ECS underscore the significance of the CSF layer thickness below the electrode, equaling the distance to the cortical surface in our model. The thickness of the CSF layer varies among patients and is likely reduced by the thickness of the lead placed between the dura and the skull (from 3.1 mm to 1.1 mm). Having the distinctly highest value of electrical conductivity, the CSF determines the main direction of the impressed current flow. With a larger CSF thickness, the impressed current may spread into all directions within the CSF, with a modest portion flowing directly towards the cortical surface (Figure 6A-3). Conversely, a larger portion of the current flows towards and into the cortex with a reduced distance between the electrode and the cortical surface (Figure 6A-1). Modeling showed a nearly linear dependency of the excitation threshold of some fiber types on the CSF thickness (Manola et al 2005). Not only do the thresholds increase with
104
increasing CSF thickness, they also increase at a different rate for various nerve fiber types. Assuming that all other anatomical and stimulus parameters are fixed, this means that the CSF thickness may determine which nerve fibers would get preferentially excited. In addition, the obvious influence of the CSF thickness on energy consumption should be taken into account. Modeling also shows that the location and depth of penetration of the electrical field into the motor cortex is dependent on the anterior-posterior position of the active electrode contact (Figure 6B). As a single electrode contact was moved in a posterior direction over PCG towards CS, the excitation threshold of different neuron types changed differently and, interestingly, did not necessarily increase or decrease consistently (unpublished results). As shown, the position of an electrode contact could even determine whether a neuron type could be excited at all (compare Figures 7A vs 7C and 7B vs 7D). Because the cortical surface is not exposed during the surgical placement of an extradural electrode, its position with respect to CS cannot be determined precisely. In combination with the rather large spacing between electrode contacts (10 mm center-to center with a standard electrode), this may not only result in high stimulus amplitudes, but also change the preferential excitation towards one of the neuron types. A combination of an IPG that can electronically steer the current between electrode contacts (Bradley 2006) and a smaller contact spacing may potentially provide more control and flexibility over the preferential excitation of some neuron types and the required stimulus amplitude. The shape of the electrical field is also influenced by the combination of active electrode contacts (monopole, bipole, etc.) and by reversal of the polarity of a contact (cathode, anode). Figure 7 shows that changing the contact combination, while all other model parameters are kept constant, changes the excitation thresholds and thus the recruitment order of different neural populations.
2. Which Nerve Cells of the Motor Cortex are Affected by MI ECS? Some populations of cortical neurons point towards an extradural electrode contact. This is the consequence of their orientation in the PCG. As demonstrated by the modeling results, varying the orientation of the neurons with respect to the electrode contact renders them more or less excitable depending on the contact location and polarity. Predictions based on the concept of the activating function (Rattay 1986) were confirmed by calculation of the excitation threshold: - a cathode depolarizes, and at sufficient stimulus magnitude, excites those parts of a neuron having their main directional component parallel to the cathodal surface. Conversely, a cathode hyperpolarizes and cannot readily excite those parts of a neuron with their main directional component perpendicular to the cathodal surface (Figures 7A, 7C) - an anode elicits opposite responses: it depolarizes or hyperpolarizes those parts of a neuron perpendicular or parallel to the anodal surface, respectively (Figures 7B, 7D). In other words, the intracortical segments of fibers in the anterior wall of the central sulcus are almost parallel to the extradural disc electrode and need less current for their activation when the electrode is a cathode. The perpendicular fibers in the anterior lip of the
105
central sulcus have an intermediate orientation and may be activated by both anodal and cathodal stimulation. Due to the favorable orientation of its proximal segment, wall fibers would need less cathodal current for their excitation than lip fibers, despite being farther from the electrode. Modeling, therefore, predicts that in ECS, unlike in e.g. spinal cord stimulation, an anode cannot be considered an indifferent electrode contact, incapable of producing excitation (Manola et al 2005,2007). In fact, an anode could, under certain conditions, result in a similar or even lower excitation threshold than the same contact programmed as a cathode (cf. excitation thresholds of A in Figure 7A and E1 in Figure 7B). This observation has been confirmed empirically as will be discussed later in this section. Modeling has also shown that, in some cases, nerve cells distant from an electrode contact may have a lower excitation threshold than the same cell type of the same caliber at a smaller distance (compare E2 vs E3 in Figure 7A). The dominant contact in bipolar MI ECS is most likely the one centered on PCG. The bipolar excitation threshold curve resembles the one of the corresponding monopole over PCG (compare Figures 7A and 7E, Figures 7B and 7F). The main difference compared to a monopole over PCG is that adding an anode or a cathode over CS decreases or increases, respectively, the excitation threshold of the neurons pointing at that contact. Figure 7E shows excitation thresholds of the electrode contact combination most often applied in ECS. This combination excites most easily A fibers, followed by E2 and E3 neurons, respectively (assuming equal fiber diameters). Due to their orientation perpendicular to the cathode excitation of E1 neurons is unlikely. Knowing which MI neurons are modeled would help to identify the pathways involved in ECS benefit. Therefore, the following assumptions were made: - E models represent neurons with axons perpendicular to the cortical laminae, in particular pyramidal cells with their efferent fibers being part of e.g. the cortico-thalamic, cortico-pontine, cortico-spinal or cortico-cortical tract. - A models represent nerve fibers parallel to the cortical laminae, being intrinsic cortical fibers or bifurcations/collaterals of ascending cortical afferents (e.g thalamo-cortical fibers). These assumptions will be used to demonstrate how modeling predictions may lead to a hypothesis on mechanisms of ECS.
3. Comparison with Empirical Data An important step of any modeling procedure is to validate its predictions. Although there are few clinical studies that could be used to validate this model of human ECS, many animal studies are available. 3.1. Experimental Data Excitation thresholds: When stimulating motor cortex in animals, two types of responses can be recorded from the cortico-spinal tract( Patton and Amassian 1954): - D-wave (direct response), attributed to direct excitation of pyramidal axons and therefore having a short latency following the stimulus pulse, and
106
Figure 8. Parts of the nerve cell where excitation occurred for different active electrode contact combinations. Note that brackets indicate a number of adjacent nodes of Ranvier where excitation could take place (with permission from Elsevier).
- I-wave (indirect response) attributed to excitation of cortical fibers, including interneurons projecting onto pyramidal cells via one or more synapses, thus resulting in a larger latency. Excitation of A fibers would likely result in I-waves, whereas excitation of E neurons would result in D-waves. The existence of D- and I- waves has been well documented by intra- and extra-cellular recordings (Amassian et al 1987, Philips and Porter 1962, Purpura and McMurtry 1965, Rosenthal et al 1967). Both cathodal and anodal stimuli can evoke D- and I-waves. Cathodal stimuli usually evoke I-waves at a lower amplitude than a D-wave, while anodal stimuli generally evoke them in a reversed order (Amassian et al 1987, Gorman 1966). Basically, less current is needed to excite pyramidal cells (E model) than parallel fibers (A model) in
107
anodal stimulation, whereas the opposite holds for cathodal stimulation. Modeling predictions presented in Figures 7A and 7B are in accordance with these empirical data, although the difference in threshold stimulus between anodal and cathodal stimulation may not be large when fiber diameters are equal (e.g. A vs E1 in Figure 7B). In the experiments comparing anodal versus cathodal stimulation, the former required least current (Amassian et al 1987, Hern et al 1962, Patton and Amassian 1954), with a ratio of 1.5-5 (Hern et al 1962). A cathodal stimulus resulted in a more complex and delayed response, suggesting a trans-synaptic response (excitation of parallel A fibers). This accords with modeling results, when excitation threshold of A-fiber in Figure 7A is compared with the excitation threshold of E1-neuron in Figure 7B, assuming that E1 has twice the diameter of A fiber. When only the I-wave response was observed in experiments, anodal stimulation had a higher threshold than cathodal stimulation (Gorman 1966), which is in accordance with modeling predictions (cf. A-fiber thresholds in Figures 7A vs 7B and 7C vs 7D). The opposite was observed in experiments for D-wave response and is confirmed by modeling (cf. E-neuron thresholds for anodal vs cathodal stimulation in Figure 7). Site of initial excitation: Regarding the local polarizations within a neuron during an anodal stimulus pulse, intracellular recordings have shown that the axonal part was depolarized, whereas the soma-dendritic part was hyperpolarized if the axis of the neuron was pointing to the anode (dendrite proximally). The polarization effects were opposite with a cathodal pulse (Amassian et al 1987, Basser and Roth 2000, Rothwell 1997). These observations are in accordance with predictions from our model, as well as other mathematical models of neurostimulation (Ranck 1979, Rattay 1998). The models predict that the initial excitation occurs always in the axon (Figure 8). In addition, the threshold stimulus pulse (210 sec) hardly affects the polarization of the somadendritic membrane compartments, as can be predicted by their large membrane time constant (about 10 msec) and their passive character (few ion channels). In empirical studies, nodes of Ranvier were also identified as the excitation sites in extracellular stimulation of the cortex (Nowak and Bullier 1998, Porter 1963). In those cases in which a pyramidal cell model could be excited by both anodal and cathodal stimulation, the site of AP initiation was at a node more distally in the axon (deeper in WM) when anodal stimulation was applied (Figure 8A vs 8B). Similar results were obtained empirically, with a latency of 0.1 msec between the anode- and cathode-induced Dwaves (Gorman 1966). We calculated a similar latency by dividing the distance between the initial anodal and cathodal excitation sites by the AP propagation velocity (assuming a fiber diameter of 10 m) (Manola et al 2007). 3.2. Clinical Data In contrast to numerous animal experiments, we found only few clinical data that could be used to validate the model. In some studies, spinal activity evoked by stimulation of the motor cortex was recorded in the cervical cord with an extradural electrode. These studies confirmed the existence of D- and I-waves in man. In accordance to animal experiments and model predictions, the stimulation amplitude needed to elicit a D-response in anodal stimulation was lower than in cathodal stimulation (Katayama et al 1988). It was also
108
observed in awake subjects that a muscular response had a lower (or similar) threshold in cathodal as in anodal stimulation (Hanajima et al 2002, Libet et al 1964). This agrees with results from animal studies (Hern et al 1962). It should be noted that the D-response of a single pyramidal cell is an insufficient condition to elicit a motor response (normally observed in human experiments) (Patton and Amassian 1951). Therefore, excitation of more than just a single pyramidal cell seems to be necessary. This differs from the concept of the excitation threshold which relates to activation of a single neuron. Since it is not known which cells and how many need to be excited to elicit a motor response, we only modeled excitation thresholds i.e. the excitability at the single cell level. With this limitation, Figure 7 is inconclusive regarding the similarity of cathodal and anodal motor thresholds. Clinical data seem to agree with model prediction regarding the site of excitation: by means of chronaxie and refractory period measurements, myelinated axons were identified as those cortical elements initially activated in MI ECS (Hanajima et al 2002). Although the clinical data that could be used to validate the model are scarce, they generally seem to comply (qualitatively) with animal data which were demonstrated to be in accordance with model predictions.
4. Model Limitations As demonstrated in the previous section, the model predictions comply qualitatively with empirical results. However, in a quantitative sense, predictions (e.g. excitation thresholds) may well differ from those observed clinically. The pulse amplitudes used in chronic stimulation range from 1 to 10V. Excitation thresholds predicted by the model may be out of this range if smaller fiber diameters are considered (Figure 7). In fact, some neurons in Figure 7 have excitation thresholds beyond the output range of the IPG (10.5 V). Also, modeling was performed with fairly wide pulses (210 sec), whereas shorter pulses (60-90 s) would additionally increase the calculated thresholds. Because the soma-dendritic compartments exhibit little influence on the excitation threshold and because the excitation site resides always in the axon, it is an imperative to understand the properties of human cortical nerve fibers well. A model of the human peripheral sensory nerve fiber (Wesselink et al 1999) was used in the current model to describe the behavior of cortical fibers, although their characteristics do not necessarily match. Data on electrical and geometrical properties of human cortical nerve fibers are mostly not available, although they are essential to predict more reliably data such as those presented in Figure 7. Similarly, parameters of the volume conductor, such as the CSF thickness under the electrode, the transition between GM and WM etc., contribute to uncertainty when predicting excitation thresholds and must be better known.
CONCLUSIONS: HOW DOES MI ECS WORK?

In chronic stimulation, the bipolar stimulus amplitude is set below the motor threshold to cause therapeutic effects, while avoiding motor responses. According to some evidence, the
109
anodal contact(s) giving the largest motor evoked potentials should be programmed as cathodes to obtain maximum pain relief (Holsheimer et al 2007). Another study correlated maximal D-responses elicited by anodal stimulus with therapeutic effects (Yamamoto et al 2007). Based on modeling predictions, it is expected that substitution of an anode by a cathode will influence excitability of the nerve cells in the underlying MI. The model of a bipole with a cathode over the PCG corresponds best to those combinations normally programmed for chronic stimulation. As shown by modeling, this combination has a potential to excite both A and E2-E3 group of nerve fibers. Since no motor effects are elicited in chronic stimulation, it is not expected that many E-fibers descending into the cortico-spinal tract are excited, but this cannot be wholly excluded based on modeling predictions of excitation thresholds. However, a direct excitation of the cortico-thalamic fibers is less likely due to their generally smaller diameter relative to pyramidal tract fibers. On the other hand, activation of A-fibers would send signals antidromically into VA-VL thalamic nucleus. Ascending thalamo-cortical fibers might also be excited, presumably anodally, and result in the equivalent effect on VA-VL nuclei. Importantly, a paddle can be applied to SI, wholly or partially, and this calls for further modeling. Finally, although ECS preferentially excites nerve fibers, we cannot exclude an influence of polarizations exhibited on soma-dendritic membranes in the cortex during the pulse. This may alter the synaptic gain of E-neurons. To sum it up, anodal extradural stimulation of any part of MI results in a low threshold direct (non synaptically mediated) MEP of the corresponding muscle(s), whereas cathodal stimulation at the same location and a 50-70% lower magnitude results in analgesia. Extradural anodal stimulation immediately activates descending corticospinal nerve fibers originating from primarily large pyramidal cells in Layer 5 of MI. Instead, cathodal stimulation activates most probably large myelinated fibers parallel to the cortical layers (i.e. collaterals of specific thalamocortical projections, collaterals of cortico-cortical projections, particularly from SI and premotor areas, intrinsic cortical connections in parallel to cortical layers I, IV,V). These parallel nerve fibers activated at a low stimulus level (20-50% of the motor threshold) constitute most likely the initial link of the neuronal chains resulting in analgesia. Action potentials elicited by electrical stimulation will generally propagate both ortho and antidromically. In conclusion, the main modeling predictions so far are: an anode cannot be considered an indifferent contact in ECS as it can excite nerve fibers pointing at it, often at a low stimulus amplitude; only axons can be excited, not cell bodies/dendrites; the CSF thickness below the electrode has a major influence on cortical excitation thresholds with a potential influence on the order of excitation; the contact combination may influence the stimulation threshold and the order of excitation; antidromic propagation of stimulus-induced APs by thalamo-cortical fibers might play a role in the ECS therapeutic mechanism.
110
REFERENCES
Amassian VE, Stewart M, Quirk GJ and Rosenthal JL. Physiological basis of motor effects of a transient stimulus to cerebral cortex. Neurosurg 1987; 20:74-93. Basser PJ and Roth BJ. New currents in electrical stimulation of excitable tissues. Ann Rev Biomed Eng 2000; 2:377-397. Bradley K. The technology: The anatomy of a spinal cord and nerve root stimulator: the lead and the power source. Pain Medicine 2006; 7:27-34. Canavero S, Bonicalzi V. Therapeutic extradural cortical stimulation for central and neuropathic pain: A review. Clin J Pain 2002; 18:48-55, 2002 Canavero S, Bonicalzi V: Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007a, 27-36 Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, pp 223-232 Frankenhaeuser B and Huxley AF. The action potential in the myelinated nerve fibre of Xenopus Leavis as computed on the basis of voltage clamp data. J Physiol 1964; 171:302-315. Gatter KC and Powell TPS. The intrinsic connections of the cortex of area 4 of the monkey. Brain 1978; 101:513-541. Gorman ALF. Differential patterns of activation of the pyramidal system elicited by surface anodal and cathodal cortical stimulation. J Neurophysiol 1966; 29:547-564. Hanajima R, Ashby P, Lang AE and Lozano AM. Effects of acute stimulation through contacts placed on the motor cortex for chronic stimulation. Clin Neurophysiol 2002; 113:635-641. Hern JEC, Landgren S, Phillips CG and Porter R. Selective excitation of corticofugal neurones by surface-anodal stimulation of the baboon's motor cortex. J Physiol 1962; 161:73-90. Hodgkin AL and Huxley AF. A quantitative description of membrane currents and its application to conduction and excitation in nerve. J Physiol 1952; 117:500-544. Holsheimer J, Lefaucheur J-P, Buitenweg JR, Goujon C, Nineb A and Nguyen J-P. The role of intra-operative motor evoked potentials in the optimization of chronic cortical stimulation for the treatment of neuropathic pain. Clin Neurophysiol 2007; 118:22872296. Katayama Y, Tsubokawa T, Maejima S, Hirayama T and Yamamoto T. Corticospinal direct response in humans: identification of the motor cortex during intracranial surgery under general anaesthesia. J Neurol Neurosurg Psychiatry 1988; 51:50-59. Libet B, Albert WW, Wright EW, Delattre LD, Levin G and Feinstein B. Production of threshold levesl of conscious sensation by electrical stimulation of human somatosensory cortex. J Neurophysiol 1964; 27:546-578. Manola L, Holsheimer J, Veltink P and Buitenweg JR. Anodal vs Cathodal stimulation of motor cortex: a modeling study. Clin Neurophysiol 2007; 118:464-474.
111
Manola L, Roelofsen BH, Holsheimer J, Marani E and Geelen J. Modelling motor cortex stimulation for chronic pain control:electrical potential field, activating functions and responses of simple nerve fibre models. Med Biol Eng Comput 2005; 43:335-343. McIntyre CC and Grill WM. Selective microstimulation of central nervous system neurons. Ann Biomed Eng 2000; 28:219-233. McNeal DR. Analysis of a model for excitation of myelinated nerve. IEEE Trans Biomed Eng 1976; 23:329-337. Mountcastle VB. The columnar organization of the neocortex. Brain 1997; 120:701-722. Nowak LG and Bullier J. Axons but not cell bodies are activated by electrical stimulation in cortical gray matter. I. Evidence from chronaxie measurements. Exp Brain Res 1998; 118:477-488. Patton HD and Amassian VH. Single- and multiple-unit analysis of cortical stage of pyramidal tract activation. J Neurophysiol 1954; 17:345-363. Paxinos G, Mai JK. The human nervous sytem. 2nd ed. Amsterdam: Elsevier-Academic Press, 2004 Penfield W and Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937; 60:389-443. Phillips CG and Porter R. Unifocal and bifocal stimulation of the motor cortex. J Physiol 1962; 162:532-538. Porter R. Focal stimulation of hypoglossal neurons in the cat. J Physiol 1963; 169:630-640. Purpura DP and McMurtry JG. Intracellular activities and evoked potential changes during polarization of motor cortex. J Neurophysiol 1965; 28:166-185. Ranck JB. Which elements are excited in electrical stimulation of mammalian central nervous system: A review. Brain Res 1975; 98:417-440. Rattay F. Analysis of models for external stimulation of axons. IEEE Trans Biomed Eng 1986; 33:974-977. Rattay F. Analysis of the electrical excitation of CNS neurons. IEE Trans Biomed Eng 1998; 45:766-772. Rosenthal J, Waller HJ and Amassian VE. An analysis of the activation of motor cortical neurons by surface stimulation. J Neurophysiol 1967; 30:844-858. Rothwell JC. Techniques and mechanisms of action of transcranial stimulation of the human motor cortex. J Neurosci Methods 1997; 74:113-122. Schwarz JR, Reid G and Bostock H. Action potentials and membrane currents in the human node of Ranvier. Eur J Physiol 1995; 430:283-292. Tehovnik EJ. Electrical stimulation of neural tissue to evoke behavioral responses. J Neurosci Methods 1996;65:1-17. Venner CH and Lubrecht AA. Multilevel methods in lubrication. Elsevier, 2000. von Keyserlingk DG and Schram U. Diameter of axons and thickness of myelin sheaths of the pyramidal tract fibres in the adult human medullary pyramid. Anat Anz, Jena 1984; 157:97-111. Wesselink WA, Holsheimer J and Boom HBK. A model of the electrical behaviour of myelinated sensory nerve fibres based on human data. Med Biol Eng Comput 1999; 37:228-235.
112
Yamamoto T, Katayama Y, Obuchi T, Kano T, Kobayashi K, Oshima H, Fukaya C and Kakigi R. Recording of corticospinal evoked potential for optimum placement of motor cortex stimulation electrodes in the treatment of post-stroke pain. Neurol Med Chir (Tokyo) 2007; 47:409-414.
CLINICAL CONDITIONS
CHRONIC PAIN
Chapter 8
NONINVASIVE STIMULATION FOR CHRONIC PAIN

Sergio Canavero, Massa Micon Barbara and GianLuca Zollino
Turin Advanced Neuromodulation Group, Turin, Italy.
INTRODUCTION
As highlighted in chapter 4, rTMS and dCS are a noninvasive strategy with a benign side effect profile and are easy to administer. Patients are awake, alert, sit in a comfortable chair during treatment sessions, which are roughly 30 minutes in duration. The cortical targets of noninvasive brain stimulation are access points for modulating the activity of specific neural networks in both hemispheres at cortical and subcortical levels. Over the past 15 years, several studies assessed the value of TMS and DCS as applied to chronic pain, including central pain, peripheral neuropathic pain, visceral pain, fibromyalgia, trigeminal neuralgia, migraine, back pain and osteomyelitis, and these are reviewed in this chapter. Alterations in sensory thresholds and pain thresholds by rTMS in experimental pain in healthy subjects may not relate adequately to pathophysiologic processes involved in the maintenance and exacerbations of chronic pain and these studies are not herewith reviewed (see Leo and Latif 2007).
1-EFFICACY
A recent metaanalysis concluded that 64% (54.6% at follow-up) of chronic pain patients responded to extradural cortical stimulation versus 40% of patients responding to noninvasive stimulation (weighted responder rate of 72.6% for ECS versus 45.3% for
Correspondence concerning this article should be addressed to: Dr. Sergio Canavero, MD (US FMGEMS), Founder and Director, +39 3494717819; e-mail: sercan@inwind.it.
118
Sergio Canavero and GianLuca Zullino
rTMS/tDCS) (Lima and Fregni 2008). Globally, brain central pain was the least responsive to rTMS with a figure-of-eight coil. A review of all studies (Table 1) shows that several patients did not deviate from baseline Visual Analog Scale (VAS) ratings. In some studies, reductions in pain ratings were also noted among patients in the sham conditions, and, in some trials, did not differ significantly from pain-rating reductions noted with active treatment: this may have been due to chance, subject expectations, a powerful placebo effect, or some other unknown confounding variable. In several studies, the sham condition involved application of the stimulating coil to an area of the head approximating the treatment conditions of real rTMS. However, the coil needs to be sufficiently tilted so that magnetic pulses are incapable of inducing neurophysiologic changes that would be expected in rTMS. Inadequate tilting of the coil in sham conditions may produce unintended cortical stimulation sufficient to produce analgesic effects and may have contributed to the observation that sham conditions indeed produced some analgesia.
2-PARAMETERS OF STIMULATION
Migita et al (1995) were the first to apply repeated TMS pulses at 0.2 Hz in two patients with central pain, using a nonfocal, circular coil over the motor cortex (MI), contralateral to the painful side (Migita et al 1995). In 1996, Canavero applied the same protocol in 9 patients suffering central and peripheral neuropathic pain (Canavero et al 1998, Canavero and Bonicalzi 2005). Later, he also applied repeated TMS pulses at 0.2 Hz in a series of patients with chronic pain secondary to stroke or spinal cord lesion, using a figure-of-eight coil for arm area stimulation or a double-cone coil for leg area stimulation (Canavero et al 2002,2003). The frequency of stimulation applied in these two studies was very low (0.2 Hz) compared with the frequencies used in chronic MI ECS (20-55 Hz) (Canavero and Bonicalzi 2002, 2005, 2007). Soon, other groups began applying rTMS and found that frequencies above 5 Hz (10-20 Hz) were more effective than lower frequencies (see Lefaucheur 2008). It has been argued that too low a frequency of stimulation (1-5 Hz) may explain the lack of efficacy of MI rTMS to induce significant analgesic effects in some studies (Lefaucheur 2008). Yet, high frequency rTMS achieved statistically significant, but clinically irrelevant effects in other reports (e.g. Andr-Obadia et al 2006), in which the threshold of efficacy was set at 10% improvement on a VAS! Although it is said that high-frequency stimulation (>5 Hz) is excitatory, while low-frequency stimulation (=/<l Hz) causes inhibition (implying that excitation is the basis of TMS-mediated analgesia), experience with MI ECS suggests that there is no uniformly effective frequency of stimulation, and that both high and low frequencies can be effective. Each individual shows variable cortical excitability, and this must be assessed at the individual level.
Table 1. rTMS and DCS studies of chronic pain

Authors/ date Stimulation site Type of coil Parameters of stimulation Study Design Number of patients Pain etiology Outcome measures Outcome Comparison with placebo Effect duration 30% reduction of pain; 0%
Migita et al. (1995)
M1 corresponding to pain distribution
0,2 Hz (circular coil) 1 train of 16,7 min (200 pulses) 80% stimulator output
No sham
Cerebral palsy + talamotomy (1), putamen hemorrrhage (1)
VAS
1 r (50%), effect duration: 1 hour same effect seen after extradural cortical stimulation Anesthesia dolorosa: only left SI afforded 30% relief CP: 1 responder (33%) PNP: 2 responders (40%) Propofol matched TMS results Extradural cortical stimulation (MI or SI) matched propofol and TMS results (one patient not implanted surgically) 42% reduction of pain (no sham)
Canavero et al. (1998)
MI and SI corresponding to pain distribution
0.2 Hz (circular coil) 200 pulses 65-100% stimulator output
No sham propofol
1 anesthesia dolorosa, 3 CP (2 CPSP, 1 postsurgery BCP), 5 peripheral neuropathic pains
VAS
Reid and Pridmore
Left dorsolateral prefrontal cortex
20 Hz
Teeth removal (1)
Authors/ date
Stimulation site Type of coil
Parameters of stimulation
Study Design
Number of patients
Pain etiology
Outcome measures
Outcome Comparison with placebo Effect duration
(2001) coil not reported Canavero et al. (2002) M1 corresponding to pain distribution
30 trains of 2 s (1200 pulses), 14 sessions (3 weeks) 100% RMT 0,2 Hz 1 train of 16,7 min (200 pulses) 100% SO
duration of effect: 4 weeks Sham Propo-fol 9 Stroke (5), spinal cord lesion (4) VAS 3 pts: 26-75% reduction of pain (both spontaneous and allodynia) for 1/4/16 hours 1 pt: 100% abatement of allodynia for 30 min (not spontaneous pain) 2 pts: worsening of pain 3 pts: no effect pain relief strongly correlated to propofol-induced pain relief (p=0.002) 4% reduction of pain (sham: 2%) P > 0,05 vs angled coil 6 r (50%) duration of effect: 6 days Pain reductions: 2461% / 48-88%
F8 and double cone
Rollnik et al. (2002)
20 Hz 20 trains of 2 s (800 pulses)
12
Circular and double-cone coils
80% RMT
Spinal cord lesion (2), osteomyelitis (1), peripheal nerve lesion (6), CRPS (2), phantom limb (1) Root avulsion
VAS
Toepper et al. (2003)
Posterior parietal Cortex
10 Hz / 1 Hz
No sham
VAS
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
F8
Brighina et al. 2004
Left dorsolateral prefrontal cortex F8
20 trains of 2 s (400 pulses) / 1 train of 12 min (720 pulses) 15 sessions (3 weeks) 110% RMT 20 Hz, 400 pulses, 90% MT, 12 sessions every other day
duration of effect: 215 min
Paral-lel
11
Chronic migraine
Number of migraine attacks, number of ingested pills
Pleger et al. 2004
M1 corresponding to pain distribution (hand)
10 Hz 1 session, 10 trains of 1,2 s (120 pulses) 110% RMT
Crossover
10
F8
Complex regional pain syndrome (minor trauma, Radial fracture, luxation of 2nd and 3rd finger, fracture of navicular bone) Trigeminal neuralgia (24), poststroke (24)
VAS
Significant long-term reduction in both number of headaches as well as amount of medicine ingested duration of effect: at least 2 months after treatment 21% reduction of pain (sham: pain increase) P = 0,02 vs angled coil 7 responders (70%) duration of effect: 4590 min and more 45% pain reduction (sham: 5%) P < 0,001 vs angled coil
Khedr et al. 2005
20 Hz 10 trains of 10 s (2000 pulses), 5 consecutive sessions (1 week)
Paral-lel
48
VAS and other scales
F8
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
80% RMT
RTMS: 21 responders oout of 28 Sham: 5 responders of 20 Duration of effect: 2 weeks after the last session 20 Hz rTMS-left S2 / 1 Hz rTMS-right S2: pain increase/ 62% pain reduction P = 0,037 vs sham coil (1 Hz-rTMS) Duration of effect: transient BCP (12 pts): 4 pts no response, 1 pt made worse, 7 pts: 37.5% reduction of pain (range 12.5% - 66.7%) CCP (2 pts): 52.6% and 20% reductions of pain Cauda lesion: 80%
Fregni et al. 2005
Left/Right secondary Somatosen-sory cortex F8
20 Hz / 1 Hz 1 session, 1600 pulses 90% RMT
Crossover
Chronic pancreatitis (visceral pain)
VAS
Hirayama et al. 2006,
M1 corresponding to pain distribution, plus somatosen-sory, premotor and supplemen-tary area F8 Neuro--navigated
5 Hz 10 trains of 10 s (500 pulses), one session
Crossover
20
Central pain-CP 14 (putaminal 4, thalamic 7, brainstem stroke 1, spinal cord lesion 2); peripheral neuropathic pain - PNP 6
90% RMT
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
(trigeminal nerve 3, brachial plexus 1, cauda equina lesion 1, peripheral nerve lesion 1)
Outcome Comparison with placebo Effect duration reduction of pain
Other PNP: 12.5% to 57.1% reductions of pain Duration of effect: 2-3 days in 1 pt only, in all others relief for max 3 hours CPSP has low response to rTMS
Saitoh et al. (2007)
1/5/10 Hz 500 pulses BCP: 5/10 (but not 1)Hz afforded >30% pain relief. 10Hz rTMS>5Hz rTMS brain central pain more refractory to rTMS than spinal cord pain or peripheral neuropathic pains
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
Sampson et al 2006
Right dorsolateral prefrontal cortex
1 Hz 2 trains of 800 s (1600 pulses), 5 weekly sessions for 4 weeks
Part of a controlled trial for depression
Fibromyalgia
VAS
Outcome Comparison with placebo Effect duration RTMS may predict response to extradural cortical stimulation 82% reduction of pain
all pts responded duration of effects: 1527 week
F8
110% RMT Sham only in 1 patient Crossover
Irlbacher et al. 2006
M1 corresponding to pain area
5 Hz/1 Hz 500 pulses, 5 consecutive days in 1 week
27
F8
Thalamic (3) or brainstem (7) stroke, spinal cord lesion (3), phantom limb pain (14)
Visual analog scale
5 Hz / 1 Hz / sham rTMS: 5% / 6% / 10% reductions of pain (=negative study) P = 0,06 (5 Hz rTMS) / 0,08 (1 Hz rTMS) vs sham coil 2 r (7%) whatever the type of rTMS Immediate pain abatement in all patients, whatever the frequency
95% RMT
Andre-Obadia et al. 2006
M1 corresponding to pain area
20 Hz/ 1 Hz 20 trains 0f 4 s/ 1 train of 26
Crossover
14
Thalamic/thala mocapsular (8) or brainstem (2: 1
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
F8
min (1600 pulses), one session each
Neuro-navigated
90% RMT
dropout) stroke, trigeminal nerve (1: dropout), brachial plexus (1), nerve trunk (1) or cervical spinal cord (1) lesion
20 Hz / 1 Hz / sham rTMS: 11% reduction of pain / 2 % increase of pain in several subjects up to +27% / 8% reduction of pain. No significant difference between 20 Hz and sham stimulation. Relief after 20 Hz = relief after sham Superiority of 20 Hz vs sham stimulation ONLY for prediction of response to ECS. level of significance set at 10% reduction of pain: this would make this study a negative one according to criteria adopted in other studies! Duration of effect: ~1 week
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
Johnson et al. (2006)
Left M1/S1
20 Hz 500 pulses
Crossover
17
F8
95% RMT
Fregni et al. (2006)
Left M1 (corresponding to C3)
2 mA anodal five 20 min sessions on five consecutive days
Parallel
17 (11 active, 7 sham)
Back pain (trauma, eg, heavy lifting, vehicle accident, fall; spinal disc degeneration, arthritis), skull base fracture and Crohn's disease Spinal cord injury (17)
Brief Pain Inventory
Outcome Comparison with placebo Effect duration Patients submitted to extradural cortical stimulation. 6 of 11 pts (55%) benefited. 5 of 11 responders to 20 Hz TMS benefited, vs 1 out of 5 nonresponders. 1 out of 6 responders to 1 Hz TMS benefited from MCS 28% pain reduction (sham: 1%)
P < 0,001 vs angled coil

13 out of 17 pts responded, 3 r to sham (30%)
tDCS
VAS and other scales plus Medication use
58% relief in 63% of patients, but only 37% at long term follow-up 8statistically insignificant) Marginal significance of active vs sham stimulation
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
Duration of effect: 2 weeks Unilateral stimulation afforded bilateral benefit 58% relief P = 0,01 vs sham tDCS (DLPFC: no significant effect) duration of effect: 3 weeks
Fregni et al. (2006)
Left M1 (C3), left dorsolateral prefrontal cortex (F3)
2 mA , anodal M1 (11), DLPFC (11), or sham M1 (11) five 20 minute sessions on five consecutive days 5 Hz 10 trains of 10 s (500 pulses), ten 15-30 minlong sessions (2 weeks) 115% RMT
Parallel
22
Fibromyalgia
VAS, and other scales plus medication use
tDCS
Defrin et al. (2007)
11
F8
Cord central pain (11) (T4/12 levels, 2 complete, 9 incomplete)
VAS / NRS / MPQ
27% pain reduction (sham: 37%) at the end of the treatment P > 0,05 vs sham coil 30% of responders (sham: 10%) in the follow-up period only (2-6 weeks after the last session), NOT acutely 30% pain reduction
Passard et al.
M1 corresponding
10 Hz
30
Fibromyalgia
VAS
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
(2007)
to pain distribution 25 trains of 8 s (2000 pulses), 10 sessions (2 weeks) 80% RMT
Outcome Comparison with placebo Effect duration (sham 3%)
F8
P < 0,01 vs sham coil Duration of effect: 2 weeks after the last session 10 Hz rTMS: 31% (sham: 11% pain increase) P = 0,01 vs sham coil 8 r esponders (57%), ~1 week. Significant but transient reduction in pain
Lefaucher et al. 2001a,b, 2004, 2006a,b, 2008a,b
M1 corresponding to pain distribution F8
10 Hz 20 trains of 5 s (1000 pulses) 80% RMT
Crossover
14
Trigeminal neuropathy (7), thalamic stroke (7)
VAS
NB: all groups above belong to the same database of patients duplicat-ed through studies
-------------10 Hz / 0,5 Hz 18 10 Hz / 0,5 Hz rTMS: 20 trains 0f 5 s / 1 train of 20 min (1000 / 600 pulses) 80% RMT -------------Thalamic stroke (6), brainstem lesion (6), brachial plexus lesion (6) 10 Hz / 0,5 Hz / sham rTMS: 20% / 4% / 7% P = 0,001 vs sham coil (10 Hz-rTMS) 7 r (39%) duration of effect: transient
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
10 Hz 1 session , 20 trains of 5 s (1000 pulses) 60 80% RMT
Thalamic stroke (12), brainstem lesion (12), brachial plexus lesion (12), spinal cord lesion (12), trigeminal nerve lesion (12)
23% reduction of pain (sham: 8%) P = 0,0002 vs sham coil 22 r (37%) duration of effect: transient
-------------10 Hz 20 trains of 5 s (1000 pulses), 16 sessions for 16 months 80% RMT -------------10 Hz 20 trains of 10 s (2000 pulses) 90% RMT 1
Brachial plexus Lesion
Thalamic (5) or brainstem (4) stroke, trigeminal nerve (14), brachial plexus (4), nerve trunk (4) or spinal cord
40% pain reduction (sham: 15%) P < 0,05 vs sham coil Duration of analgesia: ~1 week
10 Hz rTMS (painful zone area / adjacent zone, no sham): 15% / 32%
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
36
(5) injury
Outcome Comparison with placebo Effect duration 7 r esponders(19%) / 20 r esponders(56%) duration of effect: ~1 week
-------------10 Hz / 1 Hz 10 Hz / 1 Hz rTMS: 20 trains of 6 s / 1 train of 20 min (1200 pulses), one session 90% RMT -------------10 Hz / 1 Hz 20 trains of 6 s / 1 train of 20 min (1200 pulses) 90% RMT
Thalamic (8) or brainstem (2) stroke, brachial plexus (4), nerve trunk (4), or spinal cord (4) lesion
22
Thalamic stroke (13), trigeminal nerve (13), brachial plexus (10) or spinal cord (10) lesion
10 Hz / 1 Hz / sham rTMS: 32% / 11% / 10% reductions of pain P = 0,002 vs sham coil (10 Hz rTMS) 12 r (55%) / 3 r (14%) / 3r (14%) Significant reduction in pain only after 10 Hz rTMS
-------------10 Hz
Thalamic orcortical stroke (10), brachial plexus (10), nerve trunk (6),
10 Hz / 1 Hz / sham rTMS: 24% / 5% / 10% reductions of pain P = 0,002 vs sham coil (10 Hz rTMS)
Authors/ date
Study Design
Number of patients
Pain etiology
Outcome measures
20 trains of 10 s (2000 pulses) 90% RMT
46
or spinal cord (6) lesion
Outcome Comparison with placebo Effect duration 20 r (43%) / 4 r (9%) / 9r (20%)
26% reduction of pain (sham: 10%) P = 0,009 vs sham coil 11 r (34%)
32
De Ridder et al (2007)
SI (primary somatosensory area) F8
1-20Hz 90% RMT
Neurogenic pain
VAS
5 pts relieved (62%) 1Hz >>20Hz>sham
132
In a meta-analysis, slow rTMS produced analgesic effects in 14.3% of patients with chronic pain, whereas fast rTMS produced analgesic effects in 62.4%, a statistically significantly difference (Leo and Latif 2007); proalgesic effects could be observed in patients administered slow rTMS (Canavero et al 2002, 2003). In our studies, very low frequencies can be just as effective, or even more effective, than high-frequency rTMS (compare Canavero et al 2002, 2003 with e.g. Andr-Obadia et al 2006). The effects of 0.2-0.5Hz rTMS are known to differ from those seen after 1Hz rTMS (Seyal et al 2005 and references) and so this should not come as a surprise. Pain relief may be related to the number of pulses administered during each rTMS treatment session. For example, when 2000 pulses per session were administered, reductions in pain ratings were larger and more enduring than had been reported in other studies (Khedr et al 2005). According to a meta-analysis (Leo and Latif 2007), the proportions of patients achieving pain relief with less than 1200 pulses per session (60.1 %) were comparable to, albeit slightly less than, that of patients administered more than 1200 pulses per session (66.2%), making it impossible to definitively support the assertion that more is better. Finally, stimulations performed above motor threshold are not associated with better analgesia (e.g. Defrin et al 2007).
3-SITE OF STIMULATION
In most studies, the primary motor area (MI) has been the main target of stimulation. The target has been located either with conventional techniques (chapters 1 and 2) or neuronavigation, and confirmed by neurophysiological means (i.e. the ability to elicit motor-evoked potentials - assessed by electromyography and/or observation of twitches- in a muscle group corresponding to the affected body region). Lefaucheur et al (2004) found that, in a series of 60 patients with chronic neuropathic pain of various origins and locations, facial pain improved better than hand pain when the hand motor area was stimulated! Further, the same group reported that rTMS was more effective when the stimulation was adjacent to the cortical representation of the painful zone rather than within the painful zone itself (Lefaucheur et al 2006). Although in ECS best results are seen when the electrode strip overlies the cortical area corresponding to the painful area, this is not an absolute requirement (Brown and Pilitsis 2006). Furthermore, cases are on record in which ECS over the hand area controlled hemisoma or bilateral pains (chapters 9 and 10). However, no case has been reported with, e.g., facial pain controlled by hand area stimulation. Differences in current flow geometry or current densities may explain targeting differences between rTMS and ECS. A few studies assessed the value of cortical targets other than MI, namely the dorsolateral prefrontal cortex (Table 1) and above all the somatosensory areas. Two patients with phantom pain secondary to root avulsion benefited from rTMS over the posterior parietal cortex, regardless of the frequency of stimulation (Toepper et al 2003) and patients with chronic visceral pain secondary to pancreatitis benefited from 20 Hz rTMS delivered over the secondary somatosensory cortex (Fregni et al 2005). Hirayama et al (2006) reported ineffective high- frequency rTMS centered over the primary somatosensory cortex or the
Noninvasive Stimulation for Chronic Pain
133
premotor cortex. However, in our first report (Canavero et al 1998), we found SI, but not MI, 0.2 Hz stimulation analgesic in a patient with anesthesia dolorosa, and SI stimulation has been reported effective in another study (Johnson et al 2006). DeRidder et al (2007) relieved 5 of 8 neurogenic pain patients with rTMS centered over SI: 1Hz rTMS was far superior to 20Hz rTMS and more still to placebo stimulation.
4-ONSET AND DURATION OF ANALGESIA

The duration of analgesia produced by rTMS is brief. Specifically, after a single session, the duration of pain reduction ranged from 5 minutes to 8 days (Leo and Latif 2007). The duration of pain relief could be extended to at least 2 weeks by repeated MI rTMS application in central pain by 5 daily rTMS sessions (Khedr et al 2005- see also Lefaucheur et al 2004 and Passard et al 2007), although long-term analgesia was not achieved in 2 patients with phantom pain administered parietal cortex rTMS daily for 3 weeks (Toepper et al 2003). Whether it was differences in the sites of cortical stimulation or the types of pain conditions treated that accounted for the differences observed in these case reports is uncertain. The effect of TMS may start within minutes. ECS-induced analgesia is generally observed within 5-15 minutes of starting the stimulation, with an after-effect that may be quite long initially (Canavero et al 2002 and chapters 9-10). Some authors emphasize how, unlike single sessions, repeated rTMS sessions on consecutive days are able to produce cumulative effects lasting beyond the time of stimulation, which can be of value in clinical practice to control intractable pain for short periods in those drug-refractory/intolerant patients who are awaiting ECS (Lefaucheur 2008): we disagree. Differences in mean VAS ratings reported in all studies before and after treatment can be misleading, influenced by a large change in only a small subset of patients. Furthermore, significant differences were obtained even with small reductions in the VAS after rTMS treatment. In our opinion, noninvasive cortical stimulation cannot be recommended for chronic treatment at the present time, except in an unpredictable fashion in some highly responsive patients. TDCS does not appear to offer superior efficacy over rTMS.
5-TYPES AND LOCATION OF RESPONSIVE PAINS

Given the limited published experience, no indication is possible at this time on which chronic pain is more responsive. It has been suggested that central pain associated with brain stem stroke is the least responsive (Lefaucheur 2008), but this is moot. Likewise, pain localized to the face appeared to respond more favorably than limb pain (Lefaucheur 2008), but this is unfounded. Sample sizes were often too small to confirm response differences among subgroups of patients. The technical aspects of rTMS may have accounted for such differences. The representation of various body regions along MI influences accessibility of those regions to rTMS, with a large, easily accessible face area and a smaller, less accessible lower limb area, which may be distant from the scalp. Hence, rTMS applied over the facial region would be more apt to induce desired neural changes and, presumably, pain mitigating
134
effects, in that region than rTMS applied over the lower limb. However, in one study, analgesia in an area of the body affected by neuropathic pain was best achieved when rTMS stimulation was applied to a cortical area adjacent to that of the painful zone (Lefaucheur et al 2006). R-TMS does not appear to be more effective for any specific pain submodalities (ongoing, paroxysmal, dysesthetic) (Andr-Obadia et al 2008).
6-MECHANISMS OF ACTION
An intracortical locus of action is strongly suggested by available studies. Central pain is clearly subtended by an anomalous oscillatory loop between the somatosensory cortex and sensory thalamus (Canavero and Bonicalzi 2007). ECS and rTMS may rebalance such oscillatory activity, by restoring defective intracortical inhibition (GABAergic), in parallel with pain relief (Lefaucheur et al 2006; see also Canavero and Bonicalzi 1998). Propofol, a pure GABA agonist, renormalizes cortical and thalamic activity (Canavero and Bonicalzi 1998, 2007) and its effect correlates with TMS analgesia (Canavero et al 2002, 2003). Renormalization of the corticothalamic generator reverberates on all resting state networks altered by the pain generator, and these include structures that play a role in the motivationalaffective aspect of pain, such as the cingulate, prefrontal and orbitofrontal cortical areas (Canavero and Bonicalzi 2007). Dorsolateral prefrontal rTMS may be analgesic in nociceptive pain, unrelated to mood changes, in a way that replicates analgesia seen after cingulotomy for cancer pain (Canavero and Bonicalzi 2007). A study of neuropathic pain failed to demonstrate prevention of rTMS-associated analgesia after naloxone use (Toepper et al 2003). Pain activated by C fibers, but not A, may respond to slow rTMS (see in Leo and Latif 2007) and this may explain differential results in different patients and chronic pain syndromes. TMS has the ability to influence neurotransmitters (monoamines, dopamine, noradrenaline, serotonin), receptors (beta-adrenergic, NMDA) and associated second messengers (cAMP) systems, which are important in pain regulation in animal studies. It may also increase gene activity in neural and supportive elements (see Pridmore and Oberoi 2000).
7-CAN TMS HELP SELECT PATIENTS FOR ECS?

The first to suggest that a trial of TMS could predict analgesia following ECS were Migita et al (1995). We soon found this to be the case (Canavero et al 1995, Canavero et al 1998, 2002, 2002b, 2003). Others confirmed these findings in single case reports or small series (Lefaucheur et al 2004, Andr-Obadia et al 2006, Saitoh et al 2007). While there is a consensus on the relation between a positive effect after TMS and subsequent analgesia with ECS, a negative TMS trial should not rule out ECS, as trunk and leg pain may be less targetable with TMS (Lefaucheur 2008). We found a relation between TMS and propofol, and suggested that the propofol test may predict analgesia with ECS (Canavero et al 1995,
135
Canavero et al 1998, 2002, 2002b, 2003, Canavero and Bonicalzi 2005, 2007). More studies are needed to confirm these preliminary data.
CONCLUSIONS
The problems encountered in the literature include small sample sizes, variations in the rTMS protocols used, the heterogeneity of pain conditions enrolled, and inadequate follow-up of patients after treatment. In addition, studies in which post-rTMS treatment VAS ratings and patients are inadequately assessed over time may fail to detect treatment effects, particularly if analgesia is delayed a few days. rTMS protocols used across studies makes it difficult to decipher which elements are more likely to yield desired analgesic effects. Taken together, such factors limit the ability to make conclusive statements regarding the effectiveness of rTMS as a primary treatment modality for pain. Some pains of mostly noncentral nature seem to respond to MI rTMS applied focally (figure-of-eight coil) and posteroanteriorly (Andr-Obadia et al 2008) at a high rate (5-20 Hz) for at least 1000 pulses, although meaningful, short-lasting analgesia has been obtained in central pain with a circular coil at 0.2-1Hz (Migita et al 1995, Canavero et al 1998, 2002, 2003). The main benefit of noninvasive assessment is that a positive response to TMS (and possibly DCS) can predict a positive outcome of subsequent ECS. In sum, a short duration of analgesia (days to less than two months after intensive treatment) limits application of TMS and dCS in routine therapeutic practice, although some authors suggest a role for less severe cases or as add-on therapy (Lima and Fregni 2008). Drugs may synergize with concurrent noninvasive stimulation: pergolide for instanceincreased the efficacy of cathodal DCS in reducing the amplitude of laser-evoked potentials in healthy humans (Terney et al 2008). More work is needed in this area.
REFERENCES
Andr-Obadia N, Peyron R, Mertens P, Mauguiere F, Laurent B, Garcia-Larrea L. Transcranial magnetic stimulation for pain control. Double-blind study of different frequencies against placebo, and correlation with motor cortex stimulation efficacy. Clin. Neurophysiol. 2006; 117, 1536-1544. Andr-Obadia N, Mertens P, Gueguen A, et al. Pain relief by rTMS. Neurology 2008; 71: 832-840. Borckardt JJ, Weinstein M, Reeves ST et al. Postoperative left prefrontal repetitive transcranial magnetic stimulation reduces patient-controlled analgesia use. Anesthesiology 2006; 105, 557-562. Canavero S, Bonicalzi V, Pagni CA, et al. Propofol analgesia in central pain: preliminary clinical observations. J Neurol 1995; 242: 561-567. Canavero S, Bonicalzi V. Review article. The neurochemistry of central pain: evidence from clinical studies, hypothesis and therapeutic implications. Pain 1998; 74: 109-114.
136
Canavero S, Bonicalzi V, Paolotti R, Cerutti A. Extradural cortical stimulation for neurogenic pain and Parkinsons disease: the Turin experience. Proceedings of the 4th International Congress of the International Neuromodulation Society, Luzern, August 20-26, 1998. Canavero S, Bonicalzi V, Dotta M, Vighetti S, Asteggiano G, Cocito D: Transcranial magnetic cortical stimulation relieves central pain. Stereotact Funct Neurosurg 2002; 78:192-196. Canavero S, Bonicalzi V.Therapeutic extradural cortical stimulation for central and neuropathic pain: a review. Clin J Pain 2002; 18: 48-55. Canavero S, Bonicalzi V, Dotta M, Vighetti S, Asteggiano G. Low-rate repetitive TMS allays central pain. Neurol. Res. 2003; 25, 151-152. Canavero S, Bonicalzi V. Transcranial magnetic stimulation for central pain. Curr Pain Headache Rep 2005; 9: 87-89. Canavero S, Bonicalzi V. Central pain syndrome. Pathophysiology, diagnosis and management. New York: Cambridge University press, 2007. Defrin R, Grunhaus L, Zamir D, Zeilig G. The effect of a series of repetitive transcranial magnetic stimulations of the motor cortex on central pain after spinal cord injury. Arch. Phys. Med. Rehabil. 2007; 88, 1574-1580. De Ridder D, De Mulder G, Verstraeten E, Sunaert S, Moller A. Somatosensory cortex stimulation for deafferentation pain. Acta Neurochir. Suppl. 2007; 97(2): 67-74. Fregni F, Potvin K, Cohen D, Pascual-Leone A, Freedman SID: Treatment of chronic visceral pain with brain stimulation. Ann Neurol 58:971-972, 2005. Fregni F, Boggio, PS, Lima MC et al. A sham-controlled, Phase II trial of transcranial direct current stimulation for the treatment of central pain in traumatic spinal cord injury. Pain 2006; 122, 197-209. Fregni F, Gimenes R, Valle AC et al. A randomized sham-controlled proof-of-principle study of transcranial direct current stimulation for the treatment of pain in fibromyalgia. Arthritis Rheum. 2006; 54, 3988-3998. Hirayama A, Saitoh Y, Kishima H, et al. Reduction of intractable deafferentation pain by navigation-guided repetitive transcranial magnetic stimulation of the primary motor cortex. Pain 2006; 122:22-27, 2006. Irlbacher K, Kuhnert J, Roericht S, Meyer BU, Brandt SA. Zentrale und periphere Deafferenzierungsschmerzen: Therapie mit der repetitiven transkraniellen Magnetstimulation? Nervenarzt 2006; 77, 1196-1203. Johnson S, Summers J, Pridmore S. Changes to somatosensory detection and pain thresholds following high frequency repetitive TMS of the motor cortex in individuals suffering from chronic pain. Pain 2006; 123, 187-192. Khedr EM, Korb L, Kamel NF, Ahmed MA, Sadek R, Rothwell JC. Long-lasting antalgic effects of daily sessions of repetitive transcranial magnetic stimulation in central and peripheral neuropathic pain. J Neurol. Neurosurg. Psychiatr. 2005; 76, 833-838. Lefaucheur JP, Drouot X, Keravel Y, Nguyen JR. Pain relief induced by repetitive transcranial magnetic stimulation of precentral cortex. Neuroreport 2001; 12, 2963-2965. Lefaucheur JP, Drouot X, Menard-Lefaucheur L,et al. Neurogenic pain relief by repetitive transcranial magnetic cortical stimulation depends on the origin and the site of pain. J Neurol Neurosurg Psychiatry 2004; 75:612-616.
137
Lefaucheur JP, Hatem S, Nineb A, Mnard-Lefaucheur I, Wendling S, Nguyen JP. Somatotopic organization of the analgesic effects of motor cortex rTMS in neuropathic pain. Neurology 2006; 67, 1998-2004. Lefaucheur JP, Drouot X, Nguyen JP. Interventional neurophysiology for pain control: duration of pain relief following repetitive transcranial magnetic stimulation of the motor cortex. Neurophysiol. Clin. 2001; 31, 247-252. Lefaucheur JP, Drouot X, Menard-Lefaucheur I, Nguyen JP. Neuropathic pain controlled for more than a year by monthly sessions of repetitive transcranial magnetic cortical stimulation. Neurophysiol. Clin. 2004 ; 34, 91-95. Lefaucheur JP, Drouot X, Menard-Lefaucheur I, Keravel Y, Nguyen JP. Motor cortex rTMS restores defective intracortical inhibition in patients with chronic neuropathic pain: correlation with pain relief. Neurology 2006; 67, 1568-1574. Lefaucheur JP, Drouot X, Menard-LeFaucheur I, Keravel Y, Nguyen JP. Motor cortex rTMS in chronic neumpathic pain: pain relief is associated with thermal sensory perception improvement. J Neurol. Neurosurg. Psychiatr 2008; 79: 1044-1049. Lefaucheur JP. Use of repetitive transcranial magnetic stimulation in pain relief. Expert Rev. Neurotherapeutics 2008; 8, 799-808. Leo RJ, Latif T. Repetitive Transcranial Magnetic Stimulation (rTMS) in Experimentally Induced and Chronic Neuropathic Pain: A Review. J Pain 2007; 8, 453-459. Lima MC, Fregni F. Motor cortex stimulation for chronic pain. Systematic review and metaanalysis of the literature. Neurology 2008; 70: 2329-2337. Migita K, Uozumi T, Arita K, Monden S. Transcranial magnetic coil stimulation of motor cortex in patients with central pain. Neurosurgery 1995; 36, 1037-1040. Passard A, Attal N, Benadhira R et al. Effects of unilateral repetitive transcranial magnetic stimulation of the motor cortex on chronic widespread pain in fibromyalgia. Brain 2007;130, 2661-2670. Pleger B, Janssen F, Schwenkreis P, Volker B, Maier C, Tegenthoff A. Repetitive transcranial magnetic stimulation of the motor cortex attenuates pain perception in complex regional pain syndrome type 1. Neurosci. Lett. 2004; 356, 87-90. Pridmore S, Oberoi G. Transcranial magnetic stimulation applications and potential use in chronic pain: studies in waiting. Journal of the Neurological Sciences 2000; 182: 1-4. Pujol J, Pascual-Leone A, Dolz C, Delgado E, Dolz JL, Aldoma J. The effect of repetitive magnetic stimulation on localized musculoskeletal pain. NeuroReport 1998; 9:1745-1748. Reid P, Pridmore S. Improvement in chronic pain with transcranial magnetic stimulation. Aust. NZJ Psychiatry 2001; 35, 252. Rollnik JD, Wustefeld S, Dauper J, et al. Repetitive transcranial magnetic stimulation for the treatment of chronic pain: A pilot study. Eur Neurol 2002; 48:6-10. Saitoh Y, Hirayama A, Kishima H et al. Reduction of intractable cleafferentation pain due to spinal cord or peripheral lesion by high-frequency repetitive transcranial magnetic stimulation of the primary motor cortex. J Neurosurg. 2007; 107, 555-559. Sampson S, Rome JD, Rummans TA. Slow-frequency rTMS reduces fibromyalgia pain. Pain Med. 2006; 7, 115-118.
138
Seyal M, Shatzel AJ, Richardson SP. Crossed inhibition of sensory cortex by 0.3Hz transcranial magnetic stimulation of motor cortex. J Clin Neurophysiol 2005; 22: 418421. Strutton PH, Theodorou S, Catley M, McGregor A, Davey N. Corticospinal excitability in patients with chronic low back pain. J Spinal Disord Tech 2005; 18:420-424. Summers J, Johnson S, Pridmore S, Oberoi G. Changes to cold detection and pain thresholds following low and high frequency transcranial magnetic stimulation of the motor cortex. Neurosci Lett 2004; 368:197-200. Terney D, Bergmann I, Poreisz C, et al. Pergolide increases the efficacy of cathodal direct current stimulation to reduce the amplitude of laser-evoked potential in humans. J Pain Symptom Man 2008; 36: 79-91. Toepper R, Foltys H, Meister IG, Sparing R, Boroojerdi B. Repetitive transcranial magnetic stimulation of the parietal cortex transiently ameliorates phantom limb pain-like syndrome. Clin. Neurophysiol. 2003; 114, 1521-1530.
Chapter 9
EXTRADURAL CORTICAL STIMULATION FOR CENTRAL PAIN

J. E. Arle1,, J. L. Shils1 and S. Canavero2
1
Dept. of Neurosurgery, Lahey Clinic, Burlington, MA, USA; 2 Turin Advanced Neuromodulation Group, Turin, Italy.
INTRODUCTION
Since the original description by Tsubokawa et al (1991), stimulation of the MI/SI region of the cortex has been used in over 400 reported cases of neurogenic pain. Because extradural cortical stimulation (ECS) has advantages over DBS in that it can be performed without the need of general anesthesia, does not require stereotactic techniques or frame, nor the expertise of microelectrode recordings, and virtually eliminates the possibility of creating a hemorrhage, it has become the technique of first choice as a treatment of drug-refractory central pain (CP), particularly since DBS has been found poorly effective in such condition (Canavero and Bonicalzi 2007). Central pain affects millions of people worldwide, but it remains an underdiagnosed pain syndrome. It generally follows stroke, spinal cord injury or multiple sclerosis. An indepth knowledge is mandatory for optimal management (Canavero and Bonicalzi 2007).
RESULTS
As shown in Table 1, most cases have been submitted to MI stimulation, generally extradurally, although, as clarified in chapter 6, bipolar stimulation with the paddle placed perpendicularly to the central sulcus, is the same as distinct monopolar stimulation of MI and SI. In addition, Canavero and Bonicalzi (1995) reported the first case of effective SI ECS and
Correspondence concerning this article should be addressed to: Dr. J. E. Arle, MD, PhD, e-mail: Jeffrey.arle@lahey.org
140
J. E. Arle, J. L. Shils and S. Canavero
other cases are on record (e.g. Kuroda et al 2000, Saitoh et al 2000, De Ridder et al 2007). Some analgesic effects have been noted with premotor cortex ECS in a few patients by Pirotte (personal communication), but not by most others. Best results are seen when the stimulating poles overlie parts of cortex corresponding to painful body parts (generally face, arm, chest, leg singly or in combination), although data suggest that precise, millimetric, somatotopic localization of the electrode may not be required (e.g. Brown and Pilitsis 2005). Globally, 45-55% of all patients with CP of brain origin are relieved more than 40% on a VAS scale. Patients with cord central pain can be relieved too, but the number of treated cases is too small for any definite conclusions. Concurrent drugs can be reduced or even stopped in many cases. A suggestion of greater response of evoked versus spontaneous pains is not confirmed in most series and MI ECS does not relieve nonpainful paresthesias. Tsubokawa et al (1993b) reported a few patients with excellent analgesia and increasing periods of post-effect over 2 years, whose pain, after stopping the stimulator, never returned, a sign of neuroplastic phenomena induced by MI ECS in SI (1 case also in Peyron et al. 1995). What is rather puzzling is the observation that some groups report both good short- and long-term results, some have excellent initial results with a loss of benefit within weeks or months, some have scarce or no results at all. In practically all papers, there have failure rates between 12% (Ebel et al 1995) and 84% (Carroll et al 2000). The majority of the reports list as treatment failures cases improving by <30-40%, but there are a few patients who are actually satisfied with 20-40% relief, at least for some time, due to improved functional ability and quality of life; conversely, others may sometimes report high levels of pain reduction, yet fail to demonstrate any functional improvement. Certainly, changes in VAS scores alone often do not reflect the true effectiveness of a particular therapy and clearly are inadequate for determining its benefit with regard to functional capacity, quality of life and so on. Nonetheless, explaining such variable results is not easy. Localization of stimulation target can be ruled out. Groups that meticulously explored the whole target cortex with intraoperative stimulation and used sophisticated localization technology (neuro-navigation, evoked potentials) achieved either excellent or poor results in the same proportion. Inadequate (too short or too few contact combinations tried) neurostimulator programming is certainly a reason in some cases, and so are a teams lack of experience with neuromodulatory techniques and, above all, chronic pain management. Velasco et al (2008) ascribe their success to elimination of nonresponders to subacute therapeutic stimulation from the group receiving long-term stimulation, which amounts to a sizable population. Accordingly, some neurosurgeons are enthusiastic about the technique, whereas others are rather cold or downright negative. For instance, Kanpolat (Savas and Kanpolat 2005) state that some experienced authors have expressed to them dissatisfaction with MI ECS: ECS is performed at many more centers than those which publish and most of the failures go unreported, with only series with good results being published. This spurred the search for prognostic markers. Some have been proposed:
Table 1. Extradural cortical stimulation for central pain

Author(s) Tsubokawas group Tsubokawa et al. (1991) Type of pain (no. of patients) Thalamic pain (7 pts) Results (follow-up) Parameters Excellent (5: no drugs needed) or good (2: some drugs needed) pain relief (abstract: effective in 75% of cases (>7 mos) Notes First report of MCS for CP (1988) MCS vs. thalamic stimulation. MCS is more effective than thalamic stimulation. Improvement of motor function in some pts MCS improved movements of the painful limbs. Pain subsided within a few minutes. 5 10 min ON, 45 h relief 56 times daily, then 23 times daily Intermittent stimulation effective in 5/12 pts. No seizures; pain relief at stimulus intensities below movements threshold. Paresis improvement. Pain improvement in barbiturate-sensitive, morphine-resistant pts. Disappearance of the analgesic effect in 3 pts, with reappearance after revision of electrode placement Barbiturate-sensitive pts: 5 (+3?)/11; morphine-resistant pts: 10 (+1?)/11. Stimulation of area 4 ipsilateral to the inciting lesion. One week test period. Fair and poor responders not implanted. Satisfactory immediate pain relief in 8/11 pts (73%). Gradual effect reduction over several months in 3/8 pts. Long-term response in barbituratesensitive and morphine-resistant pts (also for Vc DBS). No pain-relieving effect by highfrequency postcentral stimulation (11/11 pts): in 2, worsening of pain, similar to their spontaneous ones. In 3 pts, areas rostral to MI stimulated without relief. Nonpainful paresthesias unrelieved
CPSP (12 pts; thalamic lesion: 6 pts; putaminal lesion: 3 pts; pontine hemorrhage: 1 pt; other lesions: 2 pts)
Complete pain relief in 5/12 pts (1 yr), considerable pain reduction in 3/12 pts (1 yr). Long-term benefit in 8/12 pts (>1 yr)
Tsubokawa et al. (1993)
CPSP (11 pts; thalamic stroke: 8 pts; putaminal hemorrhage (+ small lesion in the posterior limb of the internal capsule): 3 pts)
Pain relief: Immediate: Excellent (>80%): 6/11 (54%) Good (6079%): 2/11 (18%) Fair (4059%): 1/11 (9%) Poor (<40%): 2/11 (18%) Long-lasting: Excellent (>80%): 5/11 (45%) Good (6079%): 0/8 Fair (4059%): 0/8 Poor (<40%): 3/8 (37%) (2 yrs) 5-10 minutes ON at a time no stimulation at night 50 120 Hz
Author(s)
Type of pain (no. of patients)
Katayama et al. (1994)
CPSP (6 pts; lateral medullary infarct)
Results (follow-up) Parameters 100-500 s MCS in 3 pts. Pain relief: 2/3 > 60%; 1/3 > 40% (4 mos)
Notes
Yamamoto et al. (1997)
CPSP (39 pts; thalamic stroke: 25 cases; suprathalamic stroke: 14 pts)
28 MCS. Excellent/good (50100%) pain relief: Thalamic pts: 10/19 (53%) Suprathal. pts: 3/9 (33%) (difference not significant) T+ or K+ & M+: 2/4 (50%) T+ or K+ & M: 10/14 (71%) T & K & M+: 0/2 (0%) T & K & M: 1/8 (13%) Overall: 13/28 (46%) (12 mos) Early satisfactory (>60%) pain relief: 23/31 pts (74%). Long-term efficacy (2 yrs): 15/31 pts (48%)
CPSP (31 pts. Thalamic stroke: 20 pts; putaminal hemorrhage: 8 pts; lateral medullary infarction: 3 pts)
10-20 minutes stimulation ON at a time 20-50 Hz 100-500 , most at 200 s 2-8 volts
Pain relief > 40% in 1 pt previously unsuccessfully treated by Vc DBS. No satisfactory pain control by thalamic stimulation in any pts Suprathalamic stroke = infarct or hemorrhage of the posterior limb of internal capsule, or parietal lobe, sparing the thalamus. No pts with midbrain or medullary lesions. MCS test period: 1 wk. 8/39: morphine responsive 22/39: thiamylal responsive 11/23: ketamine responsive Thiamylal+ketamine sensitivity + morphine resistance may predict a positive effect of MCS Damage of the posterior limb of the internal capsule in pts with putaminal hemorrhage. Previous ineffective SCS. Pain relief >60%: 13/18 pts (73%) with no or mild motor weakness (70% of pts with inducible muscle contraction); 2/13 pts (15%) with moderate or severe motor weakness (difference statistically significant). Satisfactory pain control in 14/20 pts (70%) with inducible muscle contraction but in only 1/11 pts (9%) without inducible motor contractions (p < 0.01). No relationship between pain control and presence of hypesthesia, dysesthesia, hyperpathia, allodynia or disappearance of SSEP N20 wave plus stimulation-induced paresthesias, or motor performance improvement. 3 pts with MCS or DBS became pain-free without
Author(s)
Results (follow-up) Parameters
Notes
CPSP (45 pts)
Satisfactory pain control: SCS: 7% of pts; DBS: 25% of pts; MCS: 48% of pts
Paris group: Nguyen et al. Acta Neurochir Suppl (Wien) 1997; 68:5460
CP (12 pts; parietal lobe infarct: 1 pt; basal ganglia hematoma: 3 pts; thalamic pain: 5 pts (4 stroke, 1 abscess); deep post-traumatic brain lesion: 1 pt; SCI: 2 pts)
Nguyen et al. (1999)
CP (parietal stroke: 2 pts; deep brain hematoma: 5 pts; thalamic stroke: 4 pts; thalamic abscess: 1 pt;
Pain relief: (a) parietal lobe infarct 80100%: 1/1 pt (14 mos) (b) basal ganglia hematoma <40%: 1/3 pts (30 mos) 4060%: 2/3 pts (25+32 mos) (c) thalamic pain <40%: 2/5 pts (18+22 mos) 4060%: 1/5 pt (28 mos) 6080%: 2/5 pts (22+30 mos)* (d) post-traumatic lesion <40%: 1/1 pt (29 mos) (e) SCI <40% in a paraplegic pt (18 mos). 100% (visceral pain + substantial reduction of diffuse pain) in a tetraplegic pt (22 mos) Satisfactory/good (40100%) pain relief: parietal stroke: 2/2 pts; deep brain hematoma: 4/5 pts; thalamic stroke:
stimulation for years (all 3 getting initial excellent relief at progressively longer stimulation intervals during intermittent stimulation) 1 Infection 3 seizures during testing at intensities higher than muscle threshold (See text) Sometimes SCS and DBS produced longlasting pain-free intervals (stimulator switched off). DBS and MCS in 4 pts: better result: MCS 1/4 pts; DBS 2/4 Disappearance of allodynia in 2, but reduction only of spontaneous pain in 1 pt in the thalamic pain group. Globally, continuous pain improved in 4/10, evoked pain in 6/8 and paroxysms in 6/7. Reduction of the initial pain relief (to less than 40% in 2) in 3 cases. Previous ineffective Vc DBS (basal ganglia, deep lesion). Best results in the parietal lobe infarct, the thalamic abscess, and in one thalamic infarct pts (almost normal life, drugs markedly reduced). In sum: 58% of pats getting >40% relief
Pain relief in 13 pts described as CP by authors: 5 good (70100%), 5 satisfactory (4069%), 3 poor (<40%). SCI CP: 1: 70 100% relief, 2: <40% relief. Progressive loss
Author(s)
Type of pain (no. of patients) brainstem lesion: 1 pt; head trauma: 2 pts; SCI: 3 pts)
CP (16 pts; CPSP: 11 pts; thal. abscess: 1 pt; head trauma: 1 pt; SCI: 3 pts)
Drouot et al. (2002) Same pts as reported above
CP (13 pts; parietal stroke: 2 pts; thal. stroke: 8 pts; thal. abscess: 1 pt; SCI: 2 pts)
Results (follow-up) Parameters 4/4 pts; thalamic abscess: 1/1pts; brainstem lesion: 1/1pts; head trauma: 2/2 pts; SCI: 2/3 pts (350 mos) (Mean follow-up 27.3 mos) Pain relief: CP: Good (70100%): 5/13 brain; 1/3 SCI Satisfactory (4069%): 5/13 brain Poor (<40%): 3/13 Substantial pain relief: CP: 10/13 pts (77%) SCI: 1/3 pts (33%) (350 mos; mean follow-up 27.3 mos) Chronic stimulation in cycling mode 12 hours per day with 3 hours ON and 3 hours OFF 40 Hz (25 Hz 55 Hz) 82.5 (60 180) s 2.1 volts (1.3-4) Good pain relief (>40%): 9 (69%) Unsatisfactory pain relief (<40%): 4
Notes
of effect in 6/32 pts (19%), due to electrode malpositioning in 5 pts. After electrode repositioning, good pain relief. RCT in 5 pts; 5/5 positive response Progressive loss of effect in some pts reversed in most by correct repositioning of paddle. In sum: 66% had 30100% relief, but less with at least 50% relief
1 subcutaneous infection 3 patients had headache at high stim levels. 1 patient developed a symptomatic epidural hemotoma that cleared on its own MCS effective in pts with normal or quite normal non-nociceptive thermal threshold within the painful area or in pts with improved sensory thresholds during MCS. QST testing bilaterally; in non-responders, MCS induced significant changes on the side of stimulation.
Parameters: 3 h ON, 3 h OFF, 60 s, 40 Hz, 13 mV Nguyen et al (2004) BCP (24 pts) CCP(paraplegia : 4pts) 1993-2003 CP (27 pts) Ischemic lesions 11 pts (3 thalamic (Vc), 4 medulla, 2 cortical parietal, 1 18 (75%) improved >40% 2 (50%) improved >40%
Lyon group: Nuti et al. (2005)
Follow-up: 2-104 mos (mean 49 mos). Pain relief: BCP:
Prospective evaluation of MCS. Long-term outcome evaluated by means of: 1) %pain relief 2) VAS 3) postoperative VAS decrease 4) reduction in drugs intake 5) yes/no response
Author(s)
Type of pain (no. of patients) parietal/insula/ACC, 1 parietal/insula) Haemorragic lesions 11 pts (1 thalamic (Vc), 1 thalamus/midbrain, 5 capsulothalamic, 1 capsulolenticular/insula, 3 cortical parietal) 1 frontoparietal trauma SCI (discal hernia-associated myelopathy): 3 pts Spinal conus AVM: 1 pt 1992-2003
Results (follow-up) Parameters Excellent (>70%): 3 Good (40-69%): 8 Poor (10-39%): 8 Negligible (0-9%): 4 CCP: Excellent: 0 Good: 3 Poor: 1 Negligible: 0
Notes
for being operated again. MCS efficacy not predictable by motor status, pain characteristics, lesion type, QST, SSEP/LEPs, pain duration, BCP vs CCP, presence of evoked pain. No subjective sensations during active stimulation. Partial epileptic seizures in 3 pts in the early postoperative stage or during trials for increasing intensity. 1 speech disorders and 1 motor deficit resoved spontaneously Long term relief predictable from early pain relief. 1-2 paddles. 3 subdural MCS may have adverse cognitive effects. The risk may increase with age (>50 yrs)
Decreased analgesic intake: 52% of pts (complete withdrawal 36%); unchanged: 45% of pts, unavaible data: 3%. Decrease/withdrawal of analgesic in 10/11 poor responders (!: Contradictory results as noted by Authors) Favour re-intervention: 70% of pts Parameters: 0.5-5V (mean 1.5V), 30-80 Hz (mean 45.5 Hz), 60-330 s (mean 140 s), ON 30-120 min, OFF 15min-24 hours. 5 6 hours of stimulation each day
Turin Advanced Neuromodulation Group (TANG) Canavero and Bonicalzi. (1995)
CP (2 pts; CPSP: 1 pt; syringomyelia: 1 pt)
Pain relief: 3050% in syringomyelia pt (2 yrs); no relief in CPSP
Canavero et al. (1999)
CPSP (1 pt; thalamocapsular
Effective short-term pain relief
Syringomyelia pt: parietal somatosensory stimulation. Spreading of pain to contralateral side and vanishing of analgesia at 2 yrs. Modest propofol response CPSP pt: propofol unresponsive Propofol-responsive pt. Painful supernumerary
Author(s)
Type of pain (no. of patients) stroke)
Results (follow-up) Parameters (allodynia disappearance and 50% reduction of burning pain) (5 wks)
Notes
Canavero and Bonicalzi (2002, 2007)
CP (CPSP 5 pts; SC pain: 2 pts) + 1 (algodystonia)
Effective (30100%) pain relief with MCS/PCS in 2/7 pts. Long-term efficacy (4 yrs) in 2 pt (BCP and MS CP). Ineffective MCS in 4/7
phantom arm during MCS and lasting 6 mos after stimulator switch-off. Pain relapse after 5 wks Effective SI stimulation in 1, then resubmitted to MCS with same benefit plus 50% opioid reduction (however, patient unsatisfied and explanted). Overall efficacy: 3/7 CP pts, all propofol-responsive. Ineffective MCS in 4/7 CP pts, all propofol-unresponsive, but 1 who could not be assessed due to intermittent nature of pain. Algodystonia: temporary benefit
Saitohs group:
Hosomi et al (2008)
19932003 CPSP (thalamic) (10 pts) [+1 pt relieved without stimulation]
4: no initial relief 6: 21-90% initial relief (mean: 55.6%) [late relief at 14-75 mos: 10-80% (mean: 31.6%)]; 1 explant -all 3 relieved initially 60-75% (mean: 65.3%) [late relief at 13-88 mos: 1560% (mean: 41.6%)]; 1 explant 1: no relief 3: 25-63% initial relief (mean : 42.6%) [late relief at 33-73 mos: 1550% (mean: 35%) 30% initial relief, then 10% at 72 mos 50-89% initial pain relief (60-65% late relief at 27-75 months)
Pharmacological test with phentolamine, lidocaine, ketamine, thiopental, morphine, placebo. Ketamine-sensitive pts seem to be good candidates for MCS Some pain reduction by SI stimulation. Ineffective prefrontal stimulation. First report of bilateral cortical stimulation for SCI pain. 4 mos interval between implants. 2 serious ICH, with 1 vegetative. 2 infections subdural approach, 1-2 plate electrodes within central sulcus, in 9 pts 1 plate in interhemispheric fissure, extradural ECS in only 2 pts.
CPSP (putaminal) (3 pts)
CP (stroke 3, injury 1) (brainstem)
CPSP (temporoparietal) (1 pt)
Author(s)
Results (follow-up) Parameters Globally: initial VAS score reduction in BCP: 42%; late relief: 26% at 1+ year
Notes
CCP (SCI)
Area 4 within central sulcus seems to be more effective than on MI surface.
19962005
Pain control with 34 periods (30 min each) of stimulation a day, followed by 56 h benefit without stimulation. 25-50 Hz 200 s 0.9 to 5 volts
Other groups: Taskers group (see in Canavero and Bonicalzi 2007)
CPSP (1 pt, large suprathalamic infarct)
Substantial pain relief with ipsilateral to pain MCS with gradual abatement over 6 yrs; relief with subdural stimulation over a few months. 1 relief, 1 failure
Contralateral MCS due to a lack of sufficient MI on the affected side. Stimulation-induced ipsilateral paresthesias
Hosobuchi (1993)
CPSP (2 pts) (1 pt with AVM) CPSP (5 pts; post-removal of parietal cortical AVM: 1 pt; brainstem infarction: 1 pt; thalamic lesion: 3 pts)
Meyerson et al. (1993)
CPSP (3 pts; thalamic hemorrhage: 2 pts; brainstem infarction: 1 pt)
Pain relief: Initial: 5/5 excellent. At 23 mos: 4/5 excellent (>50%); 1/6 fair. At 930 mos: 3/5 excellent (thalamic, parietal, brainstem) 1 hour ON and 6 hours OFF 20 to 30 Hz 180 to 260 s 3 to 5 volts Pain relief: CPSP: none: 3/3
Efficacy dramatically reduced in 2 thalamic pain pts, to 0% in 1 pt and 30% in 1 pt 26 mos after implantation
In spite of multipolar electrode grid in 1 and relocation of paddle in another Most patients had one or two seizures during
Author(s)
Dario et al. (see in Canavero and Bonicalzi 2007)
CPSP (thal. stroke: 2 pts; brainstem stroke: 1 pt)
Franzini et al. (2003)
CPSP (3 pts, A, B, C)
Results (follow-up) Parameters 20-30 min stimulation 5 times a day 50 Hz 300 s Amplitude was 20 to 30% below motor threshold 70% pain relief in 1 thal. pt. (3 yrs) Gradual abatement of pain relief over 2 yrs. 6090% relief. Then 5070%, then 2030% at 341 mos (ave: 27 mos) Satisfactory (3050%) pain relief: Pts A (>4 yrs) and B (>2yrs). Short-term pain relief (6 mos): pt C
Notes
test stimulation Painful sensations at the electrode site in 2 pts. 1 epidural clot leading to aphasia
All pts propofol-responsive. 22.5 V, 5075 Hz, 120210 s, continuous mode.
2 responders propofol-sensitive. Pain abolition after a second stroke in pt B. Unsatisfactory pain relief (30%) by further stimulation in pt C. Complete abolition of thalamic hand
Herregodts et al. (1995)
Thalamic pain (2 pts)
Migita et al. (1995)
CPSP (2 pts; putaminal hemorrhage: pt A; post-20 mos stereotactic thalamotomy; pt B) CPSP (thal. infarction: 2 pts; thal. hemorrhage: 5 pts)
Immediate pain relief: >50% in both pts Long-lasting: 1/2 (full relapse in one at 4 mos) Stimulation was 1 hour ON during every 6 hours Pain relief: 7080% in pt A (1 yr) No relief in pt B
Fuji et al. (1997)
Satisfactory pain relief: 6/7 pts (1 mo) Unsatisfactory pain relief: 5/7 pts (3 mos) 30 minutes stimulation time 10 100 Hz 200 s 3 to 8 volts
Pt A: morphine and barbiturate unresponsive. 30% pain relief with TMS Pt B: previous 6 month effective Vc DBS. Barbiturate responsive, morphine and TMS unresponsive Lesions included internal caspule, Vc and pulvinar (MRI confirmed, 5 pts). Early electrode removal in 1 pt after unsatisfactory test stimulation
Author(s)
Barraquer-Bordas et al. (see in Canavero and Bonicalzi 2007) Kuroda et al. (2000)
Type of pain (no. of patients) CPSP (1 pt: capsuloinsular hemorrhage)
Results (follow-up) Parameters MCS trial ineffective (motor response elicited)
Notes
Hemisoma burning pain, + evoked pains. DBS reduced CP for 5 mos and evoked pains, until glioma displaced electrode with relapse and death
CPSP (evacuated putaminal hematoma) (1 pt) CPSP (cortic. stroke 1 pt; thal. stroke: 3 pts; brainstem stroke: 2 pts). Gunshot brainstem injury (1 pt)
MCS ineffective. Later SI/SII CS effective for 4 yrs Appreciable pain relief: 1 pt, cortical (4 yrs); 2 pts (weeks to months)* No pain relief: 4/7 pts (thalamic, brainstem) *Brainstem injury: 5060% (31 mos): 1 pt -both >80% relief -60% relief -90% relief f-up: 6-14 months Relief (not broken down) (mean follow-up 6 mos) 5 to 10 sessions per day ranging from 30 minutes to 2 hours 110 Hz 210 s 2 to 8 volts Evoked pain dramatically improved. Steady burning pain moderately relieved (2 mos) 7.1 V, 5 Hz, 450 s, ON 2 h, OFF 3 h, 0/2+ Relief (no details given) The only pt where it was tried: propofolsensitive. Pain disappearance for 5 mos after stimulator switched off in the responder. Enduring benefit in 1 pt only
Nandi et al. (2002)
Roux et al. (2001)
CPSP (2 pt) SCI (1 pt) Myelopathic pain (1 pt) SCI (1 pt)
Mogilner and Rezai (2001)
Rodriguez and Contreras (2002)
SCI CP (1 pt)
SCI pain following cervical ependymoma removal. Third party analysis of results. Tremor improvement. No reduction of analgesic intake after MCS. Previous ineffective thalamotomy
Frighetto et al. (see in Canavero and Bonicalzi 2007)
CPSP (1 pt)
Author(s)
Henderson et al. (2004) Brown and Pilitsis (2005)
Type of pain (no. of patients) CPSP (1 pt)
CPSP, Wallenberg (1 pt) CPSP, thalamic (1 pt)
Results (follow-up) Parameters Relief, then loss, then new relief (?) after intensive reprogramming 0% VAS 10 to 8; McGill Quest. Index from 65 to 32 (both sensory and affective scores)
Notes
Continuous stimulation 40 Hz 90 to 240 s 2 to 8 Volts
Follow-up max. in whole series (PNP and CP): 10 months. Contrary to Nguyen, they conclude that precise, somatotopic localization of the electrode may not be required, because the optimal interelectrode distance determined during cortical mapping and afterwards with subjective patient evaluation of pain control was fully 3 cm. Intraoperative neuronavigation and cortical mapping for stimulation site targeting. Strength and discriminative sensation improvement from MCS in 3 pts with facial weakness and sensory loss. Dysarthria improvement in 1 pt More than 50% reduction in pain medication dose Follow-up: 4 mos. No side effects
Savas and Kanpolat (2005) Slawek et al. (2005)
CP (1 pt) CPSP, brainstem (1 pt)
Savas and Kanpolat (2005) Gharabaghi et al. (2005)
CP (1 pt) CPSP (hemorrhage) (3 pts) CP, insular (1 pt)
No relief 20% reduction on VAS; withdrawal of narcotic and decrease of non-narcotic medications, ability to introduce rehabilitation and improvement of sleep 0% relief during test stimulation 70100% relief (follow-up: 618 mos) 90% relief (follow-up: 24 mos)
Frameless neuronavigation. Single burr hole and vacuum headrest. Awake patient. No complications Third party evaluation Volumetric rendering of a 3-D MR data set with superimposed fMR imaging data plus intraoperative electrical stimulation
Author(s)
Pirotte et al. (2005)
Type of pain (no. of patients) CPSP : subcortical: 3 pts; capsular: 2 pts; brainstem: 1 pt; MS pain: 1 pt; cervical syrinx: 1 pt; SC ependymoma: 1 pt 19982003
Rasche et al.(2006)
CPSP(thalamic): 7 pts
Results (follow-up) Parameters Pain relief (%) 100%/50%/worsening 83%/failure (both plegic) 87.5% 100% 70% Failure Cyclic 1 hour ON during every 4 hours 40 Hz 100 s 1 5 volts 3 responders (-31%,-41%, -62%) 2/7 pts placebo responder. Duration of positive effect: 2, 4, 1.5 years (4.5-6.0 V, 5085 Hz 210250 s).
Notes
5075% drug dosage reduction among responders Third party evaluation Plegia not an unfavorable prognostic factor
Test trial including a double-blind test. Results evaluated by means of VAS. Mean follow-up 3.6 years (range 1-10 years). Single burr hole, neuronavigation. Paddle parallel to central sulcus No sensation evoked by stimulation. All responders on continuous stimulation.After implantation intermittent stimulation. Lasting pain reduction with minor changes of the stimulation parameters. Immediate or almost immediate (30-60 min) pain reduction after turning the MCS on. MCS effect lasting from 30 min up to several hours after cessation of the stimulation. Severe motor deficit in distal arm and leg.
1994-2005
Relief of dysesthesia, allodynia and hyperpathy in 2 CPSP pts (pts were able to touch the painful area without having painful sensations).
Son et al (2006)
BCP (traumatic): 1 pt
90-95% relief of spontaneous burning pain in arm and lower trunk, 70-80% relief of burning pain, heaviness and deep pressure-like pain in leg, 50% relief of heaviness and deep pressure allodynia in foot
Subdural electrode for arm pain; extradural paddle for leg pain parallel to the course of the superior sagittal sulcus
Author(s)
Ito et al. (2006) Sokal et al (2006) Cioni et al (2007)
CPSP (3 pts) CP (thalamic) (1 pt) CPSP (4 pts)
Results (follow-up) Parameters follow-up: 1 year 21 Hz, 210 s, 0.8-2.5V 0-/3+ , continuous stimulation (arm electrode) 30Hz, 210 s, 2-2.5V 0-/2+ continuous stimulation (leg electrode) Almost total relief in 2, improved in 1
Notes
Post-stimulation effect: 5 minutes
Paddle parallel to MI Relief dependent on motor function Extradural multipolar (1620) grid in all plus electrophysiologic mapping; several combinations assessed over 12 h CP and PNP series: results not given by type. 2 intraoperative seizures 1 post-operative programming seizure
Decrease of pain 3 initial and 1 late failures 1 >40% relief, 1 failure
Molet et al (2007) Arle et al (2008)
SCI (2 pts) CPSP (thalamic) (3 pts) CCP(paraplegia) (1 pt) CPSP (2 pts) Mixed (BCP and movement disorders) (4 pts)
Benefit in some 1 PS < 20% 1 PS >60% 1 PSM < 20% 2 mixed 20 to 60% 1 mixed >60% Continuous stimulation 60 to 130 Hz 60 to 400 s 2 to 7 volts 60% relief (allodynia disappeared; hyperalgesia decreased) at 1 year follow-up 40-130 Hz, 2-3.5V, 90 s 1 pt: failure (<40% relief) 2 pts: 80-100% relief f-up: 4.5 yrs (mean)
Velasco et al (2008)
CPSP (thalamic) (1 pt)
Double-blind randomized trial Hypesthesia unchanged Subdural implantation. Octopolar electrode One severe motor deficit was satisfactorily relieved
Delavallee et al (2008)
CPSP (3 pts)
Extradural Cortical Stimulation for Central Pain
153
1- An intact or nearly so corticospinal motor function has been touted as a favorable prognostic sign, but several exceptions are on record (Pirotte et al 2005, Canavero and Bonicalzi 2002,2007, Son et al 2006; Delavallee et al 2008). However, about 75% of patients without major motor deficits may get benefit. 2- Severe sensory changes not modified by MI ECS represent a predictor of unfavorable outcome, whereas improvement in sensory deficits that appears at the subacute therapeutic trial is followed by a favorable outcome (Drouot et al 2002, Velasco et al 2008). Exceptions to the rule exist, and at least 3 cases of CP improved by ECS in spite of almost total lack of sensory cortex are on record (Peyron et al 1995; Nguyen et al 1999). Moreover, normal or near-normal sensory thresholds do not always portend a favorable outcome. 3- Response to noninvasive stimulation (e.g. rTMS) may identify patients who will get successful stimulation, but this is not an absolute criterion (full discussion in chapter 8). 4- Drug dissection protocols may help identify responders. The first one was proposed by Tsubokawas group (Table 1): doses of 3 mg of morphine are injected every 5 minutes up to 18 mg followed by naloxone 0.2 mg bid, followed by thiopental 50 mg IV bolus every 5 minutes up to 250 mg and ketamine 5 mg IV bolus up to 25 mg. Canavero and Bonicalzi (2002, 2007,2007b) developed the subhypnotic propofol test, in which the patient is injected in a single blind fashion with 1.5 mg of Intralipid or similar white fat solution and after 20 minutes by 0.2 mg/Kg IV bolus of propofol. 5- Many functional modifications are observed on fMRI, PET, SPECT and MEG in pain syndromes and following ECS (chapter 11). However, it is unclear how they relate to ECS-induced analgesia and whether functional imaging would improve target choice. Canavero introduced parietal cortex stimulation on the basis of neuroimaging data (Canavero and Bonicalzi 1995), actually the first report of neuroimaging-directed brain stimulation 6- Successful ECS may be impossible in patients with a large surface of the cortex destroyed (e.g. stroke) and in those with a cortex distant from the dural surface (subdural hygroma, post-stroke atrophy). However, Canavero has made a strong case for ipsilateral (to pain) ECS (Canavero and Bonicalzi 2007a,b). There is no significant statistical difference between results obtained in thalamic or cortico-subcortical brain CP patients. Coverage of restricted painful territories (e.g. chest) should not be taken as contraindicating ECS, even though no sensori-motor phenomena are elicited during test stimulation. Paddles have been placed parallel to the rolandic fissure or perpendicular, with similar results. What is clear is that, despite some claims to the contrary (e.g. Delavallee et al 2008), subdural implantation does not seem to improve results and may actually be less effective than extradural implantation: for CP, benefit at 1 year was 26% only in the largest series (Hosomi et al 2008). Pain relief is not associated with age, sex, presence or absence of cerebral lesion, treated painful region or pain laterality (Hosomi et al 2008). Some authors offer ECS if the mean VAS score is 6 or more, but this is not such a strict criterion.
154
Holsheimer et al (2007)s study of 8 patients suggested that a good analgesic effect may be obtained if the anode giving the largest motor evoked potential in the pain region during intraoperative testing is programmed as a cathode for chronic bipolar stimulation (a similar relationship was not observed between cathodes in intraoperative testing and anodes in chronic stimulation), but this conclusion was based on 5 responders, 3 of whom actually reported 20-40% relief, which in many hands would be considered a failure. Therapeutic ECS does not generally induce any motor activation, even at a high voltage, or any sensory phenomena in a majority of patients. Thus blinded controlled studies are feasible. All studies performed up to now (Canavero and Bonicalzi 2002; Rasche et al 2006, Velasco et al 2008) exclude a placebo effect as the primary basis of analgesia. In some patients, a sensation of tingling or mild vibration projected to the same area of distribution as the pain could be induced by MI ECS at intensities below the threshold for muscle contraction. Similar paresthesias may be induced in some patients in whom moderate or severe weakness is present and muscle contraction is not inducible on MI ECS. Some of these patients report paresthesias even with low frequency or single pulse stimulation such as that used intraoperatively, unlike patients without pain. SI stimulation can be accompanied by both sensory and motor phenomena. MI ECS can also elicit dysesthesias that recede after stopping the stimulator. Satisfactory analgesia may be achieved without stimulation-induced paresthesias in some patients, whereas stimulation-induced paresthesias may be obtained without satisfactory analgesia in others. Absolute exclusion criteria from ECS include: major depression accompanied by suicidal thoughts or gestures, major psychosocial stressors (job dissatisfaction, marital problems), major personality disorders, alcoholism and drug addiction. Hemisoma pains do not represent a contraindication, as these can be managed with two strips electrodes combined.
PROGRAMMING
During intraoperative testing with an external constant current pulse generator, single monopolar anodal and cathodal stimuli of identical amplitude and pulse width should be applied through each electrode separately and the resulting motor evoked potentials recorded in a muscle situated in the pain region. The electrode configuration is designed to utilize the contact that is centered over the area of highest motor response at the lowest stimulation level found intraoperatively. Stimulus intensity should remain low to restrict the potentially activated region to the superficial region of MI. Improvement usually occurs within minutes (up to 3 hours) of stimulation and, although stimulation of different pairs of contacts induces analgesia, there is always a better analgesic response or the threshold is lower in a given pair of contacts. This is to be considered the optimal target for chronic stimulation. Chronic stimulation usually exploits a bipolar setting with the negative (cathode) pole over the selected area of MI (or SI) and positive (anode) pole over another area of MI or SI, or wholly on SI. Patient programming is typically begun within 24 hours of electrode implantation. At the initial programming session, all contacts are checked in a monopolar cathodal setting at 210 s and 60 Hz (a few authors set it at 130 Hz), but often bipolar settings are used when more
155
anodal contacts are added. The voltage is slowly raised to 3-4V looking for adverse motor movements and sensory changes. As amplitude is slowly raised, the patient state is evaluated. One primary concern is the possible generation of seizure activity during these tests and some authors maintain all patients at or below 2.5 volts for the first two weeks post surgery, if they experienced a seizure during intraoperative mapping. If minimal benefit is noted, more contacts are added to the monopolar configuration. If this situation fails, then one contact (either contact 0 or contact 3) is placed in the cathodal mode and the other three contacts are set as the anode. Frequency, pulse width and amplitude are varied according to preset schedules: Canaveros policy is to change parameters every 20 minutes according to a preestablished schedule. Generally, patients report analgesia within the first hour. If stimulation does not induce analgesia in a range of 2.1-10V, the pair of contacts is considered off target. Non-responders are labeled so after up to 30 days of intensive parameters search. When analgesia is obtained, patients must be submitted to blinded sham stimulation. Programming includes a wide range of stimulation parameters (each patient is different and no indications are possible) with voltage varying from 0.5 volts to 6V up to 10.5V (mean 3.8V), pulse width varying from 60s up to 500 s (mean: 251.2 s) , and frequency varying from 5 Hz to 60Hz (often 25-50Hz, mean: 51.1Hz) up to 80Hz and 130 Hz. Fujii et al (1997) found that stimulation at 10 or 100 Hz produced the same level of analgesia, although the latter was associated with fatigue in the limbs. Chronic stimulation can be cyclical (battery sparing) or continuous. When cyclical, the duration of each stimulating session varies anywhere from 5 minutes to 9 hours 1-to-10 times a day. Patients are discharged and then seen as outpatients at regular intervals during which parameters are checked and pain levels reassessed each time. After finding analgesic parameters, further adjustments may or may not be required, according to different reports. The choice of stimulation parameters also depends on the presence of the so-called posteffect. Many, but not all patients have their pain relieved or improved immediately or within 5 to 60 minutes during intra-operative stimulation for periods ranging from several minutes to hours or several days without further stimulation. This effect has a tendency to abate over time and by the second month may stabilize at several minutes to a few hours. Analgesia also can fade over time. Repositioning of the electrode or intensive reprogramming may restore benefit in some cases, although at a lower level than before. Tolerance and fatigue are proposed mechanisms of such effects. Granulations and fibrosis around the contacts have been found in some failures (e.g. Hosomi et al 2008, Canavero and Bonicalzi 2002): curettage may restore benefit.
ADVERSE AND UNUSUAL EFFECTS

The most common adverse effect reported in the literature consisted of short generalized seizures, all of which were observed during the initial testing phases. This was described in fewer than half of the studies, hardware failures and infection being the second and third most likely complications, respectively. Following extradural MI ECS, Meyerson et al (1993) and Nguyen et al (1997) reported two extradural hematomas, one of which had to be surgically removed.
156
The literature does not reveal cases of epileptic seizures during chronic treatment for CP (Canavero and Bonicalzi 2002). Anti-epileptic medications are -or are not- administered for 1 to 6 months. However, at 18 month follow-up in one post-stroke pain patient of Arle, a seizure was generated while raising the stimulator voltage from 4.0V up to 4.3V. Her voltage has since been kept below 4.0 volts and no further seizure activity has been noted. Analgesia via ipsilateral stimulation is on record. Nguyen et al. (1999) saw no major modification of somatotopic arrangement in their patients, but one obtained bilateral benefit from unilateral stimulation. Rainov et al (1997) found that, by changing the polarity of the electrodes, it was possible to induce tingling sensations and muscle activation not only contralaterally to the stimulated MI, but also in the ipsilateral part of the face. In one unique case of peripheral neuropathic pain, Son et al (2003) relieved pain in a whole hemibody by focal stimulation of the hand area.
REFERENCES
Arle JE, Shils JL. Motor cortex stimulation for pain and movement disorders. Neurotherapeutics 2008; 5: 37-49. Brown JA, Pilitsis JG. Motor cortex stimulation for central and neuropathic facial pain: a prospective study of 10 patients and observations of enhanced sensory and motor function during stimulation. Neurosurgery. 2005;56:290-297. Canavero S, Bonicalzi V. Cortical stimulation for central pain. J Neurosurg 1995; 83: 1117. Canavero S, Bonicalzi V, Castellano G, et al. Painful supernumerary phantom arm following motor cortex stimulation for central post stroke pain. J Neurosurg 1999:91:121-122. Canavero S, Bonicalzi V. Therapeutic extradural cortical stimulation for central and neuropathic pain: a review. The Clinical Journal of Pain 2002;18:48-55. Canavero S, Bonicalzi V. Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007, 27-36. Canavero S, Bonicalzi V. Central pain syndrome. Pathophysiology, diagnosis and management. New York: Cambridge University Press, 2007b. Carroll D, Joint C, Maartens N, Shlugman D, Stein J, Aziz TZ. Motor cortex stimulation for chronic neuropathic pain: a preliminary study of 10 cases. Pain 2000; 84: 431-437. Cioni B, Meglio M. Motor cortex stimulation for chronic non-malignant pain: current state and future prospects. Acta Neurochir Suppl 97(2): 45-49. Delavalle M, Abu-Serieh B, de Tourchaninoff M, Raftopoulos C. Subdural motor cortex stimulation for central and peripheral neuropathic pain: a long-term follow-up study in a series of eight patients. Neurosurgery. 2008;63:101-5. De Ridder D, Devulder G, Verstraetens E, et al.Somatosensory cortex stimulation for deafferentation pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007, 451-462. Drouot X, Nguyen JP, Peschanski M, Lefaucheur JP. The antalgic efficacy of chronic motor cortex stimulation is related to sensory changes in the painful zone. Brain 2002; 125: 1660-1664.
157
Ebel H, Rust D, Tronnier V, Spies EH, Bker D, Kunze S. Chronic precentral stimulation in trigeminal neuropathic pain. Acta Neurochir 1996;138:1300-1306. Franzini A, Ferroli P, Dones I, Marras C, Broggi G. Chronic motor cortex stimulation for movement disorders: a promising perspective Neurol Res 2003; 25:123-126. Fuiji M, Ohmoto Y, Kitahara T, et al. Motor cortex stimulation therapy in patients with thalamic pain. No Shinkei Geka 1997;25:315-319. Fukaya C, Katayama Y, Yamamoto T, Kobayashi K, Kasai M, Oshima K. Motor cortex stimulation in patients with post-stroke pain: conscious somatosensory response and pain control. Neurol Res 2003; 25: 153-156. Gharabaghi A, Hellwig D, Rosahl SK, et al. Volumetric image guidance for motor cortex stimulation: integration of three-dimensional cortical anatomy and functional imaging. Neurosurgery 2005; 57 (ONS Suppl 1): 114-120. Henderson JM, Boongird A, Rosenow JM, LaPresto E, Rezai AR. Recovery of pain control by intensive reprogramming after loss of benefit from motor cortex stimulation for neuropathic pain. Stereotact Funct Neurosurg 2004; 82: 207-213. Herregodts P, Stadnik T, De Ridder F, DHaens J. Cortical stimulation for central neuropathic pain: 3-D surface MRl for easy determination of the motor cortex. Acta Neurochir Suppl 1995;64:132-135. Hosomi K, Saitoh Y, Kishima H, et al. Electrical stimulation of primary motor cortex within central sulcus for intractable neuropathic pain. Clin Neurophysiol 2008; 119: 993-1001. Hosobuchi Y. Motor cortex stimulation for control of central deafferentation pain. In: Devinsky O et al (eds) Electrical and magnetic stimulation of the brain and spinal cord. Advances in Neurology vol.63, New York: Raven Press, 1993, 215-217. Ito M, Kuroda S, Takano K, et al. [Motor cortex stimulation for post-stroke pain using neuronavigation and evoked potentials: report of 3 cases] No Shinkei Geka 2006; 34: 919-934. Katayama Y, Tsubokawa T, Yamamoto T. Chronic motor cortex stimulation for central deafferentation pain: experience with bulbar pain secondary to Wallenberg syndrome. Stereotact Funct Neurosurg 1994;62:295-299. Katayama Y, Fukaya C, Yamamoto T. Post-stroke pain control by chronic motor cortex stimulation neurological characteristics predicting a favorable response. J Neurosurg 1998;89:585-591. Katayama Y, Yamamoto T, Kobayashi K, Kasai M, Oshima H, Fukaya C. Motor cortex stimulation for post-stroke pain: comparison of spinal cord and thalamic stimulation. Stereotact Funct Neurosurg 2001; 77: 183-186. Kuroda R, Yamada Y, Kondo S, et al. Electrical stimulation of the second somatosensory cortex for intractable pain: a case report and experimental study. Stereotact Funct Neurosurg 2000; 92: 150-155. Meyerson BA, Lindblom U, Linderoth B, Lind B, Herregodts P. Motor cortex stimulation as treatment of trigeminal neuropathic pain. Acta Neurochir Suppl. 1993;58:150-153. Migita K, Uozumi T, arita K, Monden S. Transcranial magnetic coil stimulation in patients with central pain. Technique and application. Neurosurgery 1995;36:1037-1040. Mogilner A Y, Rezai AR. Epidural motor cortex stimulation with functional imaging guidance. Neurosurg Focus 2001;11:E4.
158
Molet J, Rodriguez R, Munoz F, et al. Estimulacion cortical para el tratamiento del dolor cronico neuropatico. Abstracts, Spanish Neurosurgical Society 2007 Congress, Zaragoza, 2007, p20. Nandi D, Smith H, Owens S, Joint C, Stein J, Aziz T. Periventricular gray stimulation versus motor cortex stimulation for poststroke neuropathic pain. J Clin Neurosci 2002; 9: 557561. Nguyen JP, Lefaucheur JP, Decq P, et al. Chronic motor cortex stimulation in the treatment of central and neuropathic pain. Correlations between clinical, electrophysiological and anatomical data. Pain 1999; 82:245-251. Nguyen JP, Lefaucheur JP, Pollin B, Brugieres P, Keravel Y. Motor cortex stimulation for the treatment of deafferentation pain. Neuromodulation 2004; 7: 146. Nuti C, Peyron R, Garcia-Larrea L, et al.: Motor cortex stimulation for refractory neuropathic pain: Four year outcome and predictors of efficacy. Pain 2005;118:4352. Peyron R, Garcia-Larrea L, Deiber MP, et al. Electrical stimulation of precentral cortical area in the treatment of central pain: electrophysiological and PET study. Pain 1995; 62: 275286. Pirotte B, Voordecker P, Neugroschl C, et al. Combination of functional magnetic resonance imageing-guided neuronavigation and intraoperative cortical brain mapping improves targeting of motor cortex stimulation in neuropathic pain. Neurosurgery 2005; 56 (ONS Suppl 2): 344-359. Rainov NG, Fels C, Heidecke V, et al. Epidural electrical stimulation of the motor cortex in patients with facial neuralgia. Clin Neurol Neurosurg 1997; 99: 205-209. Rasche D, Ruppolt M, Stippich C, Unterberg A and Tronnier VM. Motor cortex stimulation for long-term relief of chronic neuropathic pain: A 10 year experience, Pain 2006;121:4352. Rainov NG, Fels C, Heidecke V, et al. Epidural electrical stimulation of the motor cortex in patients with facial neuralgia. Clin Neurol Neurosurg 1997; 99: 205-209. Rodriguez RF, Contreras N. Bilateral motor cortex stimulation for the relief of central dysesthetic pain and intentional tremor secondary to spinal cord surgery: a case report. Neuromodulation 2002;5:189-195. Roux FE, Ibarrola D, Lazorthes Y, Berry I. Chronic motor cortex stimulation for phantom limb pain: a functional magnetic resonance imaging study: technical case report. Neurosurgery. 2001; 48:681-7. Saitoh Y, Kato A, Ninomiya H, al.: Primary motor cortex stimulation within the central sulcus for treating deafferentation pain. Acta Neurochir Suppl 2003;87:149-152. Savas A, Kanpolat Y.Comment to Brown JA, Pilitsis JG (2005) Neurosurgery 2005; 56: 296. Slawek J, reclowicz D, Zielinski P, Sloniewski P, Nguyen JP. [ Motor cortex stimulation in the central pain syndrome] Neurol Neurochir Pol 2005; 39: 237-240. Smith E, Joint C, Nandi D, Stein JF, Aziz TZ. Motor cortex stimulation for neuropathic pain. Neurosurg Focus 2000;11: E2. Sokal P, Harat M, Gryz J, Ackerman D. [Motor cortex stimulation in the treatment of the central pain: a case report] Neurol Neurochir Pol 2006; 40: 253-257.
159
Son BC, Lee SW, Choi ES, Sung JH, Hong JT. Motor cortex stimulation in a patient with intractable complex regional pain sydrome Type II with hemibody involvement. J Neurosurg 2003;98:175-179. Son BC, Lee SW, Choi ES, et al.: Motor cortex stimulation for central pain following a traumatic brain injury. Pain 2006;123:210-216. Tani N, Saitoh Y, Hirata M, et al.: Bilateral cortical stimulation for deafferentation pain after spinal cord injury, J Neurosurg 2004;101:687689. Tirakotai W, Riegel T, Sure U, Rohlfs J, Gharabaghi A, Bertalanffy H, Hellwig D. Imageguided motor cortex stimulation in patients with central pain. Minim Invas Neurosurg 2004;47:273-277. Tsubokawa T, Katayama Y, Yamamoto T, et al.: Chronic motor cortex stimulation for the treatment of central pain. Acta Neurochirurgica Suppl 1991;52:137-139. Tsubokawa T, Katayama Y, Yamamoto T, et al. Chronic motor cortex stimulation in patients with thalamic pain. J Neurosurg 1993;78:393-401. Tsubokawa T., Katayama Y., Yamamoto T., Hirayama T., Maejima S., Koyama S., Fujii M., Treatment of deafferentation pain with thalamic and motor cortex stimulation: possible role of reorganization of neural circuits, VIIth World Congress on Pain. Book of Abstracts, 1993b, IASP Publications, Seattle. pp. 504505. Velasco M, Velasco F, Brito F, et al. Motor cortex stimulation in the treatment of deafferentation pain. I. Localisation of the motor cortex. Stereotact Funct Neurosurg 2002; 79: 146-167. Velasco F, Arguelles C, Carrillo-Ruiz JD, et al. Efficacy of motor cortex stimulation in the treatment of neuropathic pain: a randomized double-blind trial. J Neurosurg 2008; 108: 698-706. Yamamoto T, Katayama Y, Hirayama T, Tsubokawa T. Pharmacological classification of central post-stroke pain: comparison with the results of chronic motor cortex stimulation therapy. Pain 1997; 72: 5-12.
Chapter 10
EXTRADURAL CORTICAL STIMULATION FOR PERIPHERAL (INCLUDING TRIGEMINAL) NEUROPATHIC PAIN

D. Rasche and V.M. Tronnier
Department of Neurosurgery, University of Lbeck, Lbeck, Germany.
INTRODUCTION
Peripheral neuropathic pain (PNP) is pain initiated or caused by a lesion (or dysfunction) of the peripheral nervous system, due to trauma, inflammation, metabolic diseases, tumors and iatrogenic causes. The extracerebral part of the cranial nerves is considered peripheral. The preganglionic part of the trigeminal nerve in the parapontine cistern may be considered peripheral too, for simplicity. Extradural cortical stimulation for PNP was pioneered by Meyerson in Stockholm who reported the first treated patients with trigeminal neuropathic pain (Meyerson et al. 1993). Thereafter, indications have expanded to other PNP conditions. This chapter reviews the field.
1-REVIEW OF THE LITERATURE

The published case load is smaller than for central pain, and less than 200 patients with different PNP syndromes, including cases of mixed neuropathic and nociceptive pain, are included. The largest experience has accrued for patients with trigeminal PNP (TNP), followed by pain syndromes arising from damage to the plexus or nerve root avulsion (Table I).
Correspondence concerning this article should be addressed to: Dr. Dirk Rasche, M. D., e-mail:dirk.rasche@uksh.de.
162
D. Rasche and V.M. Tronnier Table I. Summary of select published studies
Authors
Brown 2005 Canavero 2002 Carroll 2000 Ebel 1996
No. of Study patients design 6 prospective study
Pain
4 7
Gharabaghi 1 2005 Hosomi 15 2008
Katayama 2001 Mertens 1999 Meyerson 1993 Nguyen 1999 Pirotte 2005 Rainov 2003 Rasche 2008 * Saitoh 2000 Velasco 2008
5 4 9 15
6 2
20 4
trigeminal neuropathic pain (4), postherpetic neuralgia (2) retrospective trigeminal neuropathic pain analysis (4), anesthesia dolorosa (1), stumb pain (1) retrospective phantom limb pain (3), analysis neuroma (1) retrospective trigeminal neuropathic pain analysis retrospective trigeminal neuropathic pain analysis retrospective brachial plexus avulsion (7), analysis phantom limb pain LE (4), spinal cord injury (2), trigeminal neuropathic pain (1), peripheral nerve injury (1) retrospective phantom limb pain analysis prospective plexus avulsion study retrospective trigeminal neuropathic pain analysis retrospective TNP (12), nerve root analysis extraction (2), post-zosterneuralgia (1) retrospective trigeminal neuropathic pain analysis retrospective trigeminal neuralgia (1), analysis glossopharyngeal neuralgia (1) retrospective trigeminal neuropathic pain analysis (17), plexus avulsion (3) retrospective brachial plexus avulsion (2), analysis phantom limb pain (2) prospective randomized double-blind trial postherpetic neuralgia (5), cervical root avulsion (2), brachial plexus trauma (1), hemangiectasia syndrome (1), scleroderma (1)
Operative technique epidural lead placement
Outcome
Follow-Up
6/6 implanted
3 - 24 months unknown
epidural lead placement epidural lead placement epidural lead placement epidural lead placement subdural, intrasulcal lead placement
6/6 implanted
2/3, 0/1 implanted 6/7 implanted 1/1 implanted 14/15 implanted
< 31 months 5 - 24 months 40 months 17-112 months
epidural lead placement epidural/subdura l lead placement epidural lead placement epidural lead placement epidural lead placement epidural lead placement epidural/subdura l lead placement subdural, interhemisperal lead placement epidural lead placement
1/5 implanted 2/4 implanted 8/9 implanted 7/12, 1/2, 1/1 implanted 5/6 implanted 2/2 implanted
> 24 months 23 months 8 - 40 months 27 months
6 - 60 months 18 months
10/17, 3/3 implanted 4/4 implanted
6 - 156 months 6 - 20 months 12 months
10
7/10 implanted
114
Extradural Cortical Stimulation for Peripheral Neuropathic Pain 1-Trigeminal Neuropathic Pain (TNP)
163
This is one of the main indications for ECS. Generally it refers to pain arising after medical procedures for trigeminal neuralgia control. About 70 patients have been reported in retrospective and prospective trials. More than 80% of these patients are initial responders to ECS (VAS relief >50%), but only 64% are relieved in the long term. This may be due to the very large representation of the facial area on the motor strip. Meyerson et al. (1993) reported a positive effect in 8/9 patients with a follow-up of 8-40 months. Ebel et al (1996) achieved an initial positive (50-100%) effect in 6/7 patients, but in the long term (5-24 months), only 3/6 patients were relieved >50%, whereas the other 3 lost effect over 1-6 months. Rainov et al (1997) implanted a stimulator in 1 trigeminal and 1 glossopharyngeal neuralgia patients with sufficient and relatively stable effect at 18 months. Canavero and Bonicalzi (2002) reported 100% initial relief in a patient with post-trigeminal rhizotomy anesthesia dolorosa, but this was lost within 2 months; of three iatrogenic TNPcases, 1 (with additional facial PHN) was relieved up to 80% for approximately 6 months and then lost effect (Canavero et al 2006), whereas 1 got no relief and 1 was not certain. Another TNP (iatrogenic) patient was almot completely relieved by MI ECS for 3 months only to lose effect suddenly (Canavero S, personal communication). Nguyen et al. (2004) markedly relieved 20 of 26 (75% of the cases). Brown and Pilitsis (2005) relieved 3 of 4 patients 50-80%, with 1 failure, at 3-24 months. Pirotte et al (2005) had 2 excellent and 2 good results at 6-60 months. Gharabaghi et al (2005) reported 85-95% pain relief in 2 patients after 14-40 months. Cioni and Meglio (2007) submitted 8 patients with trigeminal PNP to ECS, but only one was successful (>40% VAS relief). Hosomi et al (2008) relieved 1 patient 93% at test stimulation and at very short term follow-up. Delavallee et al (2008) submitted to subdural MI CS 3 patients: two obtained 80-100% relief and one 40-59%. One of the former two had to be explanted due to infection: MI ECS recaptured the same benefit at 6V (rather than 4V).
2-Brachial Plexus Avulsion Pain About 30 cases have been reported. Nuti et al (2005) implanted 4 cases: 3 got initial relief, but only 1 was still relieved about 50% at long term. A large series was reported by Hosomi et al (2008). In 6 out of 7 patients, a permanent neurostimulator was implanted (1 patient had no benefit). In these 6 patients, VAS reduction during test stimulation varied from 10 to 90% (mean 54.3%), but at long term follow-up only three drew more than 50% relief (50-80%), 2 lost benefit (1 had the stimulator removed at 9 months and 1 at 76 months) and 1 had only 10% relief, but was still receiving stimulation 112 months later. Nguyen et al (2004) had a good result in 2 of 6 patients. Pirotte et al (2005) failed to relieve two patients. Velasco et al (2002) had one failure. Teixeira et al (2007) submitted 35 patients to TMS and 8 went on to receive MCS. Results were considered good to excellent in 5, fair in 1 and poor or nil in 2 at 1 year follow up. The best responses were observed in cases with enlarged areas responsive to TMS, while poor responders had small or no cortical responses to TMS (not statistically significant). Delavallee et al (2008) implanted one patient subdurally on MI with 40-59% relief.
164 3-Phantom Limb Pain
18 cases are on record. Two out of three patients published by Carroll et al (2000) showed a benefit (1 pt: 70% VAS reduction of phantom pain, 0% stump pain, for 30 months; 1 pt: 75% arm pain, 5% hand pain, for 27 months). Saitohs group (Hosomi et al 2008) submitted to subdural MCS 4 patients: VAS reduction at test stimulation was 30-90% (mean 50.5%). One got 90% relief until death 4 and a half years later, one got about 40% relief at short-term follow-up, 2 were explanted 5-6 months later for loss of effect. Katayama et al (2001) published 5 cases, but only one patient improved for more than 2 years. Sol et al (2001; Roux et al 2001) implanted 3 patients: more than two years later, 2 patients experienced 100% and 70% relief at rest and 80% and 40% relief during activity; the third patient (who had both plexus avulsion and amputation), after repositioning of the electrodes, obtained 80% relief at rest and 20% during activity. Canavero and Bonicalzi (2002) did not benefit 1 patient suffering stump pain with SI ECS. Pirotte et al (2005) failed to relieve 1 patient. Koppelstaetter et al (2007) relieved 70% 1 patient with arm amputation who failed spinal cord stimulation; follow-up was 6 weeks.
4-Postherpetic Neuralgia (PHN) Velasco et al. (2002) had 2 failures and two successes (1 excellent, 1 fair reliefs) for cervicothoracic PHN. They added one patient in a prospective randomized double-blind trial (Velasco et al 2008). In both patients with PHN of the spine and the one with trigeminal pain, pain reduction of 56-80% was achieved. The positive effect was stable over the follow-up period of one year. Nguyen et al (2004) relieved 1 patient with intercostal PHN. Brown and Pilitsis (2005) relieved 2 cases 80-100% after 10 months (mean).
5-Complex Regional Pain Syndrome Son et al (2003) reported the case of a patient with both CRPS and postamputation (finger) pain, who had excellent relief at 1 year. Velasco et al (2002) reported two excellent responses, along with measurable vascular changes. Fonoff et al (2007) had excellent results in two patients (100% and 78% reliefs) at 18 and 6 months. All these published patients had sensorimotor improvement in the affected regions. Table II reports parameters of stimulation and lead position in select series.
2-PERSONAL EXPERIENCE
Since 1994, we have submitted to ECS 20 patients with trigeminal neuropathic pain (TNP) or brachial plexus avulsion (BPA). 17 patients suffered from chronic neuropathic facial pain after partial destruction or lesions of the trigeminal nerve itself or peripheral branches. This was due to operative procedures for trigeminal neuralgia, tumors surrounding
Extradural Cortical Stimulation for Peripheral Neuropathic Pain
165
the trigeminal nerve or dental manipulations. The mean postoperative follow-up was 2.25 years (SD = 3.5; range 1.5-13 years) for patients with active ECS. Three male patients suffered from posttraumatic unilateral BPA and chronic neuropathic pain of the upper limb. Neuropsychological assessment and testing was performed preoperatively by an independent neuro-psychologist to rule out major depression, psychosis, etc. Over the years, 10 patients were excluded from this treatment option due to psychiatric disorders including severe depression and drug abuse. Intravenous testing of barbiturate, propofol and morphine was performed in selected cases, with unpredictive results. Table II. Parameters of stimulation and paddle positioning in various series
Authors/year Impulse duration 2.0-8.0 90-240 2.0-7.0 450 3.5-10.5 180-350 75% MT 200 n. s. n. s. 2.0-8.0 100-500 1.5-4.5 60 70-80% MT 300 1.3-4.0 60-180 1.0-5.0 100 5.0-8.5 300-400 2.5-6.5 210-360 Amplitude Frequency Stimulation mode continuous cycling cycling n.s. on demand n.s. cycling on demand cycling cycling continuous continuous, cycling on demand cycling Stimulation lead quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar quadripolar, octad quadripolar quadripolar Lead orientation
Brown 2005 Carroll 2000 Ebel 1996 Gharabaghi 2005 Rosomi 2008 Katayama 2001 Nuti 2005 Meyerson 1993 Nguyen 2004 Piroette 2005 Rainov 2003 Rasche 2008 * Saitoh 2000 Velasco 2008
40 15-25 60-130 20-70 n. s. 25-50 25-50 50 25-55 40 100-120 25-85 25-50 40
parallel parallel parallel parallel parallel, intrasulcal parallel parallel parallel cross cross parallel parallel/cross cross cross, parallel
1.5-5.0 2.0-7.0
200 90
Our technique has been described previously (Rasche et al 2006). Briefly, all patients have been operated using local anesthesia and neuroleptoanalgesia, with the head fixed in a 3 point pin holder. A quadripolar paddle lead was placed via a single burr hole epidurally with all four contacts covering the precentral gyrus, guided by a sterile pointer of the neuronavigation system. During intraoperative stimulation, no sensory phenomena were reported. In one case, the lead was placed subdurally, because of brain atrophy. The lead extension was fixed to the dura or the border of the burr hole with a silk suture to prevent dislocation. A test trial of about 6-14 days was conducted in all patients. On the first postoperative day, plain x-rays and a CT-scan of the head were performed to document the position of the lead and to rule out any complications like epidural bleeding. As long as the connection cables were externalized, a prophylactic oral antibiotic was administered. Pain intensity was measured using a visual analogue scale (VAS) by specially trained nurses or physicians and the patients were asked to complete a pain diary. Blinding and placebo stimulation were obtained by covering and shielding the external stimulation device in the area of the lead combinations. Lack of induction of paresthesias effectively allows blinded or double blinded testing (patient, physician and the nurse staff on the ward). An independent physician is needed to program and control the stimulation device, which is covered and shielded for setting the lead combinations, according a standardized protocol
166
with defined lead combinations, voltages and also placebo stimulation phases. Using this method, in our own patient sample, we were able to identify four patients with TNP as falsepositive responders and the stimulation lead was explanted. A positive effect of ECS was defined as pain reduction of more than 30% on the VAS, subjective impression of improvement and reduction of daily medication intake. In case of response, a permanent neurostimulator was implanted. Only continuous stimulation was used during the test trial, but after implantation intermittent stimulation was programmed to avoid habituation. After permanent implantation, stimulation parameters were 2-7 mA, 30-85 Hz and 210-300 s. In the TNP patient group, 10 out of 17 patients were screened as responders to ECS and a permanent neurostimulator was implanted. Of 3 patients with posttraumatic BPA and chronic pain of the affected upper limb, with the lead placed over the motor strip of the forearm and hand, two obtained an improvement of the allodynia and tension pain in the affected limb. In all patients a > 50% pain reduction was achieved; no effect was observed during placebo stimulation. Globally, at long-term follow-up, a positive effect of ECS with lasting improvement of pain was observed. Only minor changes of the stimulation parameters were necessary. More than 50% pain reduction was achieved in 7 of the 10 patients with TNP. Of the remaining three patients one patient had only minor pain reduction of about 30%, in one case the generator was switched off because of seizures and one patient died of unrelated reasons. Two patients with BPA and long-term response to ECS reported more than 50% pain reduction and a stable effect over more than 3 years. In the third patient, a slowly diminishing effect of the ECS was detected after 18 months. Dislocation of the lead and technical failure of the implanted devices was ruled out. Even with reprogramming, no positive clinical effect was recaptured. All responders reported improvement of pain immediately or within 30-60 minutes after activation of ECS. After deactivation, the positive effect lasted for up to several hours. Active ECS improved sensory phenomena like allodynia or dysesthesia in 9 patients, including the two patients with BPA. In all cases, initially a cyclic stimulation mode was programmed (ON 3 x 0.5 hours/day initially, 1 hour ON/ 0.5 hours OFF at later stages). In 10 patients, it was necessary to change the stimulation mode to continuous stimulation due to diminishing effect of the ECS in the cycling mode. The stimulator could be switched off at night by 5 patients (thus extending battery life). In four patients, the IPG was replaced once or more due to battery exhaustion. At long term, stimulation parameters were 2.5-6.0 mA, 30-85 Hz, 210-360 s. Intraoperative seizures due to suprathreshold stimulation were seen in 9 of the 20 patients and in 8 cases IV phenytoin was administered (19/20 patients were on pregabalin, gabapentin and/or carbamazepine for pain control and a seizure protection effect was assumed). In one case, a patient with TNP, postoperative focal seizures occurred even with very low stimulation intensities and the stimulator was removed. In all operated cases, a postoperative cranial CT-scan was performed after lead insertion to exclude intracranial complications. In one patient, local wound infection led to system extirpation, with reimplantation of the stimulation device and connection cable after several months. In sum, results in our 20 patients after the test trial were excellent, with a success rate of 59% (10/17) in TNP and 100% in BPA. Long term follow-up with pain reduction of at least
Extradural Cortical Stimulation for Peripheral Neuropathic Pain
167
30% or more was achieved in 8/10 TNP and 2/3 BPA patients. In one case, ECS had to be stopped because of uncontrollable seizures.
CONCLUSION
Most of the published series represent a very inhomogeneous and mixed collective of pain patients. Also, some cases of rare pain syndromes including both neuropathic and nociceptive pain were reported (e. g. stump pain, neuroma, scleroderma, neurofibromatosis, traumatic injury, etc.) with different results. From these individual case reports, no general conclusions can be drawn. However, ECS can be considered an alternative surgical procedure for a select patient group with chronic PNP. Well-localized neuropathic pain syndromes after nerve injury, like in cases of TNP, seem to respond more favorably than pain syndromes after lesion of the central pain pathways itself. ECS for PNP is less invasive than and with similar results to DBS, according to the literature. ECS should be performed in an experienced center following a standardized protocol including double blind, placebo testing. Only future controlled studies will determine its exact place in pain management. Canavero (personal communication) is planning a specific trial targeting idiopathic trigeminal neuralgia.
REFERENCES
Arle JE, Shils JL .Motor cortex stimulation for pain and movement disorders. Neurotherapeutics 2008; 5: 37-49. Brown JA, Pilitsis JG. Motor cortex stimulation for central and neuropathic facial pain: a prospective study of 10 patients and observations of enhanced sensory and motor function during stimulation. Neurosurgery 2005; 56: 290-297. Canavero S, Bonicalzi V.Therapeutic extradural cortical stimulation for central and neuropathic pain: a review. The Clinical Journal of Pain 2002; 18: 48-55. Canavero S, Bonicalzi V, Clemente M. No neurotoxicity from long term (>5 years) intrathecal infusion of midazolam in humans. J Pain Symptom Manage 2006; 32: 1-3. Carroll D, Joint C, Maartens N, Shlugman D, Stein J, Aziz TZ. Motor cortex stimulation for chronic neuropathic pain: a preliminary study of 10 cases. Pain 2000; 84: 431-437. Cioni B, Meglio M. Motor cortex stimulation for chronic non-malignant pain: current state and future prospects. Acta Neurochir Suppl 2007; 97(2): 45-49. Delavalle M, Abu-Serieh B, de Tourchaninoff M, Raftopoulos C. Subdural motor cortex stimulation for central and peripheral neuropathic pain: a long-term follow-up study in a series of eight patients. Neurosurgery. 2008;63:101-5. Ebel H, Rust D, Tronnier V, Spies EH, Bker D, Kunze S. Chronic precentral stimulation in trigeminal neuropathic pain. Acta Neurochir 1996; 138: 1300-1306. Fonoff ET, Teixeira MJ, Lin TY, Marcolin MA. Motor cortex stimulation for complex regional pain syndrome (CRPS type 1) results in pain relief and functional recovery. Eur J Pain 2007; S183, A413.
168
Gharabaghi A, Hellwig D, Rosahl SK, et al. Volumetric image guidance for motor cortex stimulation: integration of three-dimensional cortical anatomy and functional imaging. Neurosurgery 2005; 57 (ONS Suppl 1): 114-120. Hosomi K, Saitoh Y, Kishima H, et al. Electrical stimulation of primary motor cortex within central sulcus for intractable neuropathic pain. Clin Neurophysiol 2008; 119: 993-1001. Katayama Y, Yamamoto T, Kobayashi K, Kasai M, Oshima H, Fukaya C. Motor cortex stimulation for phantom limb pain: a comprehensive therapy with spinal cord and thalamic stimulation. Sterotact Funct Neurosurg 2001; 77: 159-161. Koppelstaetter F, Siedentopf CM, Rhomberg P, et al. FMRT vor Motorkortex-stimulation beim Phantomschmerz. Nervenarzt 2007; 78: 1435-1439. Kuroda R, Yamada Y, Kondo S. Electrical stimulation of the second somatosensory cortex for intractable pain: a case report and experimental study. Stereotact Funct Neurosurg 2000; 74: 226. Meyerson BA, Lindblom U, Linderoth B, Lind G, Herregodts P. Motor cortex stimulation as treatment of trigeminal neuropathic pain. Acta Neurochir (Wien) Suppl 1993; 58: 150153. Nguyen JP, Lefaucheur JP, Pollin B, Brugieres P, Keravel Y. Motor cortex stimulation for the treatment of deafferentation pain. Neuromodulation 2004; 7: 146. Pirotte B, Voordecker P, Neugroschl C, et al. Combination of functional magnetic resonance imageing-guided neuronavigation and intraoperative cortical brain mapping improves targeting of motor cortex stimulation in neuropathic pain. Neurosurgery 2005; 56 (ONS Suppl 2): 344-359. Rainov NG, Fels C, Heidecke V, et al. Epidural electrical stimulation of the motor cortex in patients with facial neuralgia. Clin Neurol Neurosurg 1997; 99: 205-209. Rasche D, Ruppolt M, Stippich C, Unterberg A, Tronnier VM. Motor cortex stimulation for long-term relief of chronic neuropathic pain: a 10 year experience. Pain 2006; 121: 4352. Roux FE, Ibarrola D, Lazorthes Y, Berry I. Chronic motor cortex stimulation for phantom limb pain: a functional magnetic resonance imaging study: technical case report. Neurosurgery. 2001; 48:681-7. Sol JC, Casaux J, Roux FE, et al. Chronic motor cortex stimulation for phantom limb pain: correlations between pain relief and functional imaging studies. Stereotact Funct Neurosurg 2001; 77: 172-176. Son BC, et al.: Motor cortex stimulation in a patient with intractable complex regional pain sydrome Type II with hemibody involvement. J Neurosurg 2003;98:175-179. Teixeira MJ, Fonoff ET, Macri F, Picarelli H, Lin TY, Marcolin MA. Prediction of results of motor cortex stimulation in treatment of brachial plexus avulsion by transcranial magnetic stimulation. Eur J Pain 2007 (doi : 10.1016/j.epain.2007.03.449) A434 S183. Velasco M, Velasco F, Brito F, et al. Motor cortex stimulation in the treatment of deafferentation pain. I. Localisation of the motor cortex. Stereotact Funct Neurosurg 2002; 79: 146-167. Velasco F, Argelles C, Carillo-Ruiz JD, et al. Efficacy of motor cortex stimulation in the treatment of neuropathic pain: a randomized double-blind trial. J Neurosurg 2008; 108: 698-706.
Chapter 11
NEUROIMAGING AND NEUROPHYSIOLOGICAL STUDIES OF ECS FOR CHRONIC PAIN

Sergio Canavero1,, J. E. Arle2 and J. L. Shils2
2
Turin Advanced Neuromodulation Group, Turin, Italy; Dept. of Neurosurgery, Lahey Clinic, Burlington, MA 01805, USA.
INTRODUCTION
With the advent of human neuroimaging over the last 25 years, there has been a trend to use this technique with its pretty pictures of colored blobs on brain slices as a modern-day phrenology. However, these pretty pictures can easily mislead us: their interpretation needs to take into account the wealth of scientific evidence obtained with different methods from humans. There are many, sometimes quite small, populations of neurons with different responses to different types of stimulus or event in brain regions which may not all be revealed by functional neuroimaging, which rather reflects the average metabolic demands of a brain region. Further, brain imaging does not address the issue of the information that is represented by virtue of the different tuning of individual neurons, and so does not provide the evidence on which computational models of brain function must be based. It is thus very important to consider the results of human functional neuroimaging in the light of what is known from complementary studies using, for example, neurophysiology and the effects of brain damage. With this caveat in mind, this chapter will examine studies evaluating effects of surgical cortical stimulation by means of neuroimaging techniques, such as single photon emission computed tomography (SPECT), positron emission tomography (PET) and functional magnetic resonance imaging (fMRI).
170
Sergio Canavero, J. E. Arle and J. L. Shils
REVIEW OF STUDIES
1- Tsubokawa and colleagues (1991) found that cortical stimulation changes both local cortical (SI/MI) and thalamic rCBF, by comparing pre- and post-stimulation SPECT exams. 2- Canavero and Bonicalzi (1995) found that parietal cortex stimulation renormalized a locus of SPECT hypoperfusion in the parietal cortex in one patient suffering central pain. Renormalization went along analgesia. In another central pain patient, MI ECS renormalized SPECT thalamic hypoperfusion, while providing analgesia (Canavero et al 1999). 3- MI ECS has inhibiting effects on SI/MI cortex as well as contralaterally, as reported in a fMRI study of phantom pain (Sol et al 2001). 4- Saitoh et al. (2004) submitted a right-sided CPSP patient to subdural MI CS, with excellent analgesia (VAS 8 to VAS 1) after 30 minutes of stimulation. H2(15)O PET pre- and post-stimulation revealed significant rCBF increases in left frontal areas (BA9 and 11, BA32) and the left thalamus and decreases in temporo-occipital areas (right BA22 and left BA19). The efficacy of MI CS was mainly related to increased synaptic activity in the thalamus, whereas all other changes were related to emotional processes. The same authors submitted to H2(15)O PET (resolution: 4x4x5mm at FWHM) 6 patients during right-sided 25-40 Hz CS (3 with central pain and 3 with brachial plexus avulsion pain, all left-sided) (Kishima et al 2007). The PET study was performed 1-3 years after implantation. Stimulation was stopped more than 12 hours before PET. Six PET scans were performed before subdural MI CS. MI CS was run for 30 minutes and 6 PET scans were performed after onset of analgesia and then analyzed considering all patients together with the SPM software. Comparison of rCBF before and after MI CS showed significant rCBF increases after MI CS in the left posterior thalamus (pulvinar) and left posterior insula. No areas of significant rCBF decrease were identified. By comparing early post-MI CS scans with pre-MI CS scans, the authors found significant rCBF increases in the left posterior insula and the right orbitofrontal cortex (BA11) and significant decreases in the right BA9 and the right BA4. By comparing late post-MI CS scans with pre-MI CS scans, the left caudal ACC (BA24) showed significant increases, while comparison between early post-MI CS with late post-MI ECS scans brought out significant rCBF increases in the left SMA (BA6). Unlike the Lyon groups findings (see below), the ipsilateral (to MI ECS: right) thalamus was not affected. Results were not differentiated between central and peripheral neuropathic pain. 5- Garcia-Larrea et al (1997) studied 7 CPSP and 3 PNP (brachial plexus avulsion, BPA, pain) patients submitted to contralateral MI ECS (in 3 medially, i.e., subdurally). H2(15)O PET was done before, during 5 and 20 minutes and 30 minutes after a 20 minute session of stimulation. Results were not differentiated between CP and BPA. There was no significant difference in regional cerebral blood flow (rCBF) between the two controls or the two stimulation conditions. The only locus of significant CBF increase during MI ECS was observed in the motor thalamus. Sizable, but insignificant CBF increases during MI ECS were seen in the left insula, BA2432 and upper mesencephalon (plus a rCBF decrease in BA1819 bilaterally). No significant change was seen in MI (SI could not be resolved with their machine). All changes were reversible upon stopping MI ECS, although BA24 and mesencephalic changes persisted or even increased slightly after stoppage of MI ECS. They compared 3 patients with 80100% relief and 4 with less than 40% relief. Mean thalamic
Neuroimaging and Neurophysiological Studies of ECS for Chronic Pain
171
CBF was enhanced in both groups with a similar time course, albeit rCBF increase was greater in those with 80%-plus relief. In contrast, mean CBF in BA24-32 appeared to increase during MI ECS only in patients with good relief and to decrease in poor responders, even in individual analyses. The same group (Laurent et al. 1999; Garcia-Larrea et al. 1999) evaluated 10 patients with CP and BPA (likely including the above-mentioned patients, although time from implantation to PET does not correspond). MI ECS was stopped 24 hours before PET. Four consecutive scans were first recorded (A). Then PET was recorded at 5, 15, 25 and 35 minutes after switching on MI ECS (B). MI ECS was subsequently stopped and PET recorded at 15, 30, 45, 60 and 75 minutes after MCS had been turned off (C). MI ECS (B versus A) was associated with increased rCBF in rostral ACC contralateral to the electrode. During MI ECS stoppage (C versus A) there was strong activation up to 75 minutes after MI ECS discontinuation of rostral ACC, orbitofrontal cortex, basal ganglia and brainstem. MI ECS (B+C versus A) was associated with decreased blood flow immediately below the electrode. Images of CBF changes in the brainstem did not cover the localization of the PAG. They did not find MI ECS activation of SI, a possible consequence of the spatiotemporal resolution (12 mm) limits of their PET machine. The low-threshold analysis (Z-score 3.5) of the two-step procedure yielded some regions of significant CBF increase: the whole thalamus (ipsilateral to MI ECS), the ACC (mostly contralaterally to MI ECS, plus midline), orbitofrontal areas, a region comprising the insula and descending towards the inferomedial temporal lobe including amygdala (exclusively contralateral to MI ECS) and the subthalamic-upper brainstem region (ipsilateral to MI ECS). The second (high-threshold) step of the analysis (Z-score 4) restricted spatially the above results and limited the anatomical region of significant CBF increase to thalamic VL ipsilateral to MCS, with extensions to VA and subthalamic region. Vc was outside the region of increased CBF in both high- and low-threshold analyses. The sequence included condition A (CBF assessed basally, 15 minutes before MI ECS with stimulator turned off for 18 hours), conditions B and C (2 consecutive scans performed respectively after 5 and 20 minutes of continuous MI ECS) and condition D (scan after 30 minutes after MI ECS discontinuation). Pain ratings during PET were 4.8 2.6 during condition A, 4.3 2.9 and 3.69 2.8 in conditions B and C and 3.69 2.8 in condition D. In spite of a trend to pain decrease from A to D, differences were not significant. As far as rCBF changes are concerned, in all cases there was an abrupt CBF increase during the first scan under MI ECS (5 minutes after onset) which remained stable during PET 20 minutes after MI ECS onset. These effects were reversible 30 minutes after MI ECS interruption in all sites, except in ACC where rCBF had not yet reverted to prestimulation values 30 minutes after MCS discontinuation: here two spots of increased rCBF appeared in right and left ACC/orbito-frontal boundaries (despite unilateral analgesia!) and stayed almost so after switching off the stimulator. No significant change related to MI ECS was observed in SI or MI. CBF decreased in BA1819 areas and were totally reversible upon discontinuation of MI ECS. In CP and BPA patients with >80% versus <20% relief, while lateral thalamic CBF appeared to increase in all patients (albeit to a greater extent 15% versus 5% in those relieved), BA32 CBF increased in responders (+5% at 20 minutes), but decreased in non-responders (10% at 20 minutes); upon close scrutiny, this does not seem a strong finding, as in their two reported CP cases this was not the case. Garcia-Larrea et al (2006) submitted to MI ECS a patient with left facial central pain due to a left medullary
172
infarct. Although the territory with sensory loss was much wider in the right nonpainful than in the left painful side, PET showed significant rCBF reduction in the right thalamus, contralateral to the small painful area. 40 Hz MCS afforded 60% relief and PET showed renormalization of the thalamic anomaly. Peyron et al (2007) explored the post-stimulation period using an enlarged temporal window (as long as PET studies allow). 19 morphine-nave patients suffering brain central pain (13 patients), cord central pain (4 patients) or brachial plexus avulsion (2 patients) were submitted to 35 Hz (180 s/ 2.5V/cyclical) MI ECS (paddle parallel to rolandic fissure) and subsequent PET scans. Analgesic drugs were not discontinued, bar fast-acting opioids for at least 12 hours before exam. PET resolution was 7 mm. Patients were blinded to MI ECS status (ON/OFF). After acquisition of baseline scans, the next 4 scans were acquired at 5, 15, 25 and 35 min after MCS onset. MI ECS was then turned off again and 5 further scans were recorded at 15, 30, 45, 60 and 75 minutes. Data were analyzed using SPM2 software and also considered for a functional connectivity analysis (FCA), which examines the temporal correlation of neural events between distributed brain areas. Mean pain relief was 10-40% in 8 patients, 60% in 6 and >80% in 2. Results of ON versus baseline and OFF versus baseline were as follows. Only a limited activation of the pregenual ACC(anatomically connected to MI) contralateral to MI ECS was found in the ON versus baseline comparison. The large majority of activations were found in the OFF versus baseline subtraction in the ipsilateral premotor cortex, the contralateral pregenual (pg) ACC (descending pain control) and midcingulate ( noxious processing) and supplementary motor (SMA) area, pallidum, putamen and PAG. Most of the rCBF changes that correlated with long-term analgesia occurred during the 75 minutes subsequent to MI ECS stoppage (after 35 minutes of effective stimulation). There was a correlation between rCBF changes and analgesia in the ON condition in midcingulate cortex (MCC) and pregenual ACC (BA32/24) contralaterally to MCS and in prefrontal cortices (BA10) bilaterally. There was a trend for the midcingulate to be activated in the ON condition with a persisting activation in the OFF condition, while the pregenual ACC still showed increased activity in the OFF condition. Regions whose rCBF increased relative to baseline during MI ECS and correlated positively with analgesia in the OFF condition (after stoppage of MCS) included a large ACC activation, extended from the posterior MCC and anterior MCC to the pregenual ACC, bilaterally, contralateral orbitofrontal cortex and SMA, ipsilateral cerebellum and posterior cingulate cortex, prefrontal cortices and basal ganglia bilaterally, hypothalamus, upper mesencephalon (PAG) and lower pons. These activations were maximal in the OFF condition and correlated with average analgesia. Unlike Saitohs groups findings (see above), MI rCBF below the electrode was not found to change or correlate with pain scores at any time, nor was SI. In the FCA, responses that correlated with analgesia with MI ECS ON were found to correlate also with CBF changes in other subdivisions of lateral prefrontal cortices, in contralateral orbitofrontal cortex, pgACC, anterior insula, putamen and lower pons. In the OFF condition FCA, significant covariations were found between pgACC and basal ganglia, pgACC and brainstem, pgACC and posterior cingulated cortex. Basal ganglia covariated together bilaterally, but also with posterior cingulate and insular cortices. CBF changes in mesencephalon and lower pons covariated with basal ganglia and with pgACC. The authors concluded that a network comprising the ACC/OFC/medial thalamus and PAG the same as seen during ECS induced analgesia by other procedures- appears to be
173
the final common pathway of analgesia elicited by ECS (ACC and PAG being opioid-rich areas) and becomes activated ONLY after MI ECS is discontinued. MCC and pgACC activities did not correlate with current pain relief, but with the amount of analgesia obtained after several cycles of MI ECS. The perigenual and subgenual ACC are associated with mood alterations and the production of affective states: they are part of a ventral affective system involved in the identification of the emotional significance of a stimulus, production of affective states, and automatic regulation of emotional responses, and also comprise the amygdala, anterior insula and ventral striatum. The mid-posterior cingulate cortex instead is concerned with pain unpleasantness. This study failed to replicate the authors previous finding of a significant thalamic rCBF increase, except in the FCA. They concluded that MI ECS-related thalamic activation is phasic and short-lasting, likely a trigger for other activations, and may be averaged out when 35 minutes of MI ECS are lumped together and analyzed as a whole. The same group (Maarrawi et al 2007) submitted a subgroup of the above patients (central pain 7, trigeminal peripheral neuropathic pain 1) to PET with [11C] diprenorphine PET, basally and after 2 months of chronic MI ECS. The two preoperative scans performed at 2 weeks interval did not show significant differences. Medications were kept unchanged. Data were analyzed with SPM99. Voxel-wise comparison of preoperative and postoperative PET scans showed a significant decrease of opioid receptor binding postoperatively. Buprenorphine binding decrease (group level analysis) concerned the posterior part of the midbrain (PAG) (-25.6%), anterior middle cingulate cortex (-21.2%), lateral prefrontal cortex (-23.3%) and cerebellum (-18.3%). VAS scores decreases and binding decreases correlated significantly in PAG and anterior MCC (in PFC, there was only a trend). One CP patient got minimal relief from MI ECS (VAS 8 to 7 on MI ECS) and decreases were 16.3% in PAG, 10.3% in aMCC, 10.11% in cerebellum and 17.2% in PFC. The CP patient with the best relief (VAS 8 to 2 on MCS) showed decreases respectively of 37%, 30.3%,22.2% and 25.5%, which would seem to confirm that the magnitude of decreases significantly correlated with degree of analgesia. Yet, the largest decreases were seen at PAG and PFC levels in a patient who had a VAS 7-to-2 relief, at aMCC level in the patient with the best analgesia, at cerebellar level in one with a VAS 8-to-5 change. The authors suggest that binding decreases were not due to loss of opioid receptors (as seen in some studies of central pain), but to increased endogenous opioid secretion and resulting decreased receptor availability to exogenous diprenorphine and a possible reactive down-regulation and internalization of receptors. The authors conclusion was that MCS triggers endogenous opioid secretion in part of the remaining medial pain system unaffected by opioid receptor loss in CPSP. The involved circuit would include MI that projects to PAG which in turn projects to ACC. However, their conclusion is nixed by poor opioid responsiveness of central pain (Canavero and Bonicalzi 2007). They (Garcia-Larrea et al. 1999) also recorded CO2 laser-evoked potentials (LEPs) and flexion nociceptive reflex (RIII) in a subgroup of these same patients. LEPs (amplitude and latency of each component) and RIII (surface) were studied with MI ECS turned off, on and at least 30 minutes after MI ECS interruption. LEPs were obtained after stimulation of both the painful and the intact side, while RIII was obtained after stimulation of the painful side only. In one patient, after stimulation of the nonaffected side, LEP amplitudes of the vertex component decreased significantly during active stimulation. In the group as a whole, after stimulation of the non-affected side, LEP
174
amplitudes tended to decrease under MI ECS, although not statistically significantly. RIII was not modified in the three conditions. Electrophysiological responses did not correlate with VAS. There was a lack of any significant acute change in SEPs during MI ECS in any of the recorded patients with central lesions. None of the 4 patients whose nociceptive reflexes remained unmodified by MI ECS was satisfied with the attained analgesia. Although the 7 patients with CP had sizable epidural SEPs during intraoperative monitoring, only 4 retained scalp-recorded SEPs of enough amplitude to permit assessment of MI ECS effects. Parietal somatosensory responses up to 50 ms post-stimulus did not exhibit any significant change in amplitude, latency or topography in relation to MI ECS. Thus, significant modulation of spinal nociceptive reflexes was seen during MI ECS in 3/7 patients, while it was unchanged in 4. Modification thereof corresponded in every case to attenuation of the responses during MI ECS. Two of 3 patients with MI ECS-related reflex attenuation experienced good to very good relief, while the third reported >60% abatement of allodynia during MI ECS, but only 30% of spontaneous pain.
INTERPRETATION
The reviewed studies suffer from limited statistical power due to small number of patients, shortcomings of ROI measurements, inhomogeneity in patients pains (central pain versus peripheral neuropathic pain), group analyses versus single patients, type of CS (extradural versus subdural), target (MI versus SI) and neuroimaging protocol (SPECT versus PET versus fMRI). The data available is contradictory. While the Lyons group found no cortical activation whatsoever below the electrode, i.e. in MI or SI, in all their studies, Saitohs group observed MI activation, Canavero and Bonicalzi (1995) rCBF changes in SI, Tsubokawa et al (1991) in MI, and Sol et al (2001) in SI and MI bilaterally. Importantly, analogous studies conducted during MI ECS for Parkinsons disease clearly revealed cortical changes below the electrode (see chapter 13). As concerns the thalamus, Peyron et al (2007) found no thalamic rCBF changes, whereas in their previous studies (Garcia-Larrea et al 1997, 1999, 2006) they did. Thalamic metabolic changes have been reported by Tsubokawa et al (1991), Canavero et al (1999), Saitoh et al (2004) and Kishima et al (2007) and central pain relief is accompanied by thalamic renormalization (Canavero et al 1999, Pagni and Canavero 1995). SI activation was not detected in several initial studies of acute pain, but it was in later studies (see in Canavero and Bonicalzi 2007). Consequently it is important to correlate these data with what is known of neurogenic pain pathophysiology. An impressive amount of data points to an unbalanced reverberatory loop active between the sensory cortex and the sensory thalamus as the basis of central pain (reviewed in Canavero and Bonicalzi 2007). It can be hypothesized that ECS acts locally by engaging inhibitory interneurons in the MI/SI dipole and the long corticothalamic reverberating loop, with subsequent fall-out effects on other brain regions, both through indirect trans-synaptic effects or through direct anterograde or retrograde activation of white matter projections (rostral (perigenual) ACC and PAG, insula, ). The so-called ventral affect system cannot be considered central to analgesia, since cingulotomy in CP is either ineffective or has effect on pain affect only (i.e. the pain is still
175
there, but no longer bothersome). Similarly, an increase of opioids as the basis of ECS analgesia is nixed by the almost complete lack of effect of opioids for central pain (see references in canavero and Bonicalzi 2007). Tsubokawas original hypothesis that MI, but not SI, stimulation restores the inhibitory surround of hyperactive SI nociceptive cells by anti-/orthodromic activation of nonnociceptive SI neurons has been partially disproved by successful cases of both invasive and noninvasive somatosensory cortex stimulation (chapters 8 and 9). Drouot et al (2002) concluded that MI ECS reinforces the control of non-nociceptive sensory inputs on hyperactive nociceptive SI cells, at least when these sensory afferents are partially preserved (implying that lemniscal fibers inhibit STT fibers), with improvement of sensory discrimination. In poor responders, MI ECS did not modify the sensory thresholds measured within the painful area, but induced significant changes on the opposite side, i.e., ipsilateral to stimulation. Lack of effect was ascribed to more severe disturbances of the sensory systems. However, the literature reports patients with impaired lemniscal transmission who were relieved by MI ECS, implying that a normal lamniscal system (gate control) is not required to obtain good results. Importantly, propofol may relieve both central pain and restore normal sensation in human patients (Canavero and Bonicalzi 2007). In any case, an intracortical mechanism of action is central to these theories. Fields and Adams (1974) first reported analgesia in humans by means of stimulation of sub-cortical motor fibers in the internal capsule. However, given that in humans there are no or few descending fibers from MI to the superficial dorsal horn (Schoenen and Grant 2004), ECS cannot act by descending direct inhibition to the spinal cord, and certainly not in brain central pain. The tight coupling of sensation and motricity may also explain CS effects. Suppression of natural pain-related behaviors clearly engages a potent volitional motor control process, yet movements are known to increase the threshold for detection and decrease the perceived intensity of somatosensory stimuli, including those at a painful level (active movements having greater and more consistent effect than passive movements). Humans perceive forces they exert as weaker than identical forces acted upon them: in fact, a corollary discharge of the effort attenuates the subjects sensory feedback and pain interferes with mental representations of movement (see references in Canavero and Bonicalzi 2007). Hanajima et al (2002) found that, in awake pain or movement disorders patients, chronic stimulation through a quadripolar strip electrode placed directly on the arachnoid in the subdural space over MI appears to activate low threshold (short chronaxie and refractory period) neural elements probably myelinated axons- likely at cortical level (Ikeda et al 2000). These, in turn, may activate corticospinal neurons (producing detectable effects on motoneurons) and inhibitory neurons in the cortex. The facilitation was larger (lower threshold) with cathodal (-) than with anodal (+) monopolar stimulation (which excluded corticospinal neuron axons) in most, but not all, patients (3.9:1 ratio, range 0.9-11.1). Anodal stimulation was less effective, maybe because in the awake state interneurons are better activated (Katayama et al 1988). Libet et al (1964) too found that the exposed MI of awake humans had a lower threshold to cathodal than to anodal stimulation. On the other hand, Katayama et al (1988) found that, on direct stimulation of the exposed MI of an anesthetized human, the D wave threshold was generally lower with monopolar anodal stimulation than with cathodal stimulation. Hanajima et al (2002) concluded that pulses delivered through
176
contacts implanted for chronic stimulation appear to activate corticospinal neurons mainly indirectly. Hanajima et als study could not study the activation of many corticocortical and other corticofugal systems. However, an ad-hoc software (UNCuS) (Arle et al. 2008) has been employed to simulate the general biophysics of each neuron on a conductance level, connect thousands of such independent cells together each with electrotonic dendritic processing, and keep track of each synaptic event and timestep on a 0.25msec timescale. This software has been applied to the whole intrinsic 6-layer cortical anatomic connectivity by adding sufficient layer-specific detail and interconnectivity within the MI and SI regions separately (Sloper 1973, Sloper et al 1979a,b). Layer I is primarily a dendritic arbor layer, with limited numbers of small round inhibitory stellate cells, and these cells were included in layer II in the model since they are substantially similar. MI layer V contains some of the largest pyramidal cells in the cortex (Betz cells) and is also much larger than layer V in SI, with an almost non-existent Layer IV. SI layer V also contains a smaller number of large pyramidal cells. In contradistinction, SI layer IV is much larger with a smaller, but larger than expected of a sensory area, layer V. Relative cell numbers were determined from electron microscope studies of cell types and their distributions within cortical areas. Of critical importance in this model are the so called U-fibers leaving layer III and entering layer IV (Jones et al. 1978) in both the MI and SI regions. Thalamic input to these cortical regions is primarily to layer IV with some far dendritic inputs to layer V cells (Banister 2005, Stepniewska et al. 1994) that are accounted for by having the simulated cells synapse on the furthest electrotonic dendritic segment (Arle et al. 2008). Also, efferents from cortex to thalamus come primarily from layers IV and V (Bannister et al. 1978), and these were proportionately assigned in the model. It is important to note that inhibitory synapses occur closer to cell bodies in the cortex (Porter 1997, Williams et al. 2008), likely exercising a powerful veto on the overall response of postsynaptic cells, although thalamic excitatory input to SI layer IV is still relatively stronger (Bannister 2005). This model accounts for a representation of the majority, if not all, of the major known connections and ratios (Sloper et al. 1979a, Sloper 1973, Bannister 2005, Porter 1991, Zarzecki et al. 1978, Rocco et al. 2007). Subthreshold levels of stimulation in the cortex were factored in, by trying to see driving of motor units on EMG with 10Hz stimuli. A similar thresholding was used to determine the appropriate level of MI ECS in the model by using the amplitude of stimulation in the electrodes that would change the firing rate in less than 50% of the cortical cells. Although many of these cells were closer to their baseline thresholds than without stimulation, they were not altered significantly in their baseline firing rates by the stimulation alone. Surface stimulation fields from a typical paddle-type lead likely affect horizontally-oriented fibers (Manola et al. 2007), rather than vertically-oriented fibers exiting the cortex. Many of the horizontal fibers are from the predominantly inhibitory layer 1-3 stellate interneurons (Douglas et al. 1998). MI ECS stimulates both types of fibers, rather than directly hyperpolarizing them, as has been at times speculated (Hanajima et al 2002). In the model, MI ECS at subthreshold levels for motor unit driving was applied only to MI stellate interneurons and not to the excitatory pyramidal cells, likely the output cells from the cortex overall. Since only the predominantly inhibitory interneurons received the effects of MI ECS, it is likely that this region of MI would exhibit a general decrease in firing
177
rate and, by definition, would not reveal motor unit driving by EMG. Topographically arrayed synaptic input from sensory thalamus of an 80Hz regional stimulation to the receiving layer IV pyramidal cells of SI leads to decreased activity in these layer IV pyramidal cells overall, possibly related to the U-fiber excitation from MI being inhibited locally by stimulation of the horizontally-oriented axons of the associative inhibitory cells. The lack of even baseline excitation from MI layer V pyramidal cells, which send output to layer IV pyramidal cells in SI, would create a lower firing rate in those targets. The result is a convergence and focus of activity going from sensory thalamus to SI pyramidal cells. The significant change in activity is centered only within the topographically oriented region which was coding for the painful stimulus coming from sensory thalamus. In sum, the computational representation of a painful stimulus arising through sensory thalamic cells and carried on to SI is inhibited by MI ECS acting on just the inhibitory interneurons of MI. According to the model, subthreshold MI ECS only excites the horizontal inhibitory fibers coursing horizontally within the exposed parts of the MI gyrus, and these in turn lead to local inhibition, and subsequent lack of layer 5 excitation of SI (thus inhibiting SI by providing less excitation as well). This is in line with Holsheimer et al (2007)s study, in which the neural elements most likely stimulated by MCS were superficial horizontal fibers of the largest diameter. Higher amplitudes would also stimulate excitatory layer V pyramidal cells (for instance) in MI, in addition to the inhibitory interneurons: this increased activity would over-ride the inhibition and lead to motor unit driving, which indeed does occur clinically. However, as demonstrated in chapters 6 and 7, the belief that only a cathode can activate cortical neural tissue and that the anode placed over the cortex is indifferent is wrong: when MCS is applied bipolarly, neuronal activity may be evoked under both the cathode and the anode (bifocal stimulation), which implies that SI is engaged. In chronic MI ECS, the cathode overlies MI and likely activates horizontal fibers in the upper laminae of the cortex because of their orientation (Holsheimer et al 2007). An anode on MI should activate perpendicular fibers and result in effects different from cathodal stimulation (Holsheimer et al 2007). Although the stimulation of nervous tissue generally occurs in the vicinity of the cathode (Ranck 1975), this is not true in MI ECS due to the different orientations of cortical nerve fibers (parallel and perpendicular to the laminae) and to the large size of the epidural lead electrodes: thus various populations of axons can be activated under either the cathode or the anode, making bipolar stimulation bifocal (chapters 6 and 7). Despite the large center-tocenter distance of the cathode and the anode in a quadripolar strip (=/> 10 mm), some overlap of the corresponding stimulus-evoked field is still present. Due to this overlap, the electrical fields in the cortex below the cathode and the anode are different from real monopolar fields, and elicited neural activities will most likely differ from monopolar ones. A large motor response is evoked when either anodal stimulation is applied on MI or cathodal stimulation is applied on CS, according to the orientation of cortical nerve fibers in the stimulus-induced electric field. When the MEP amplitude is high in anodal stimulation, it will be lower in cathodal stimulation and viceversa. When the electrode covers the lip of CS, anodally and cathodally evoked MEPs may have similar amplitudes. Amassian et al (1987) assumed that bipolar but not monopolar- stimulation favored the activation of nerve fibers parallel to the bipole axis (esp. in layer 1) rather than perpendicular
178
fibers. When MI ECS is applied bipolarly, the neural activity evoked near the cathode may affect the activity near the anode -and vice versa- via their intracortical connections (likely an inhibitory interaction).
REFERENCES
Arle JE, Mei LZ, Shils JL. Modeling parkinsonian circuitry and the DBS electrode: I. Biophysical background and software. Stereotact Funct Neurosurg 2008;134:1-15. Bannister AP: Inter- and intra-laminar connections of pyramidal cells in the neocortex. Neurosci Res 2005;53:95103. Canavero S, Bonicalzi V. Cortical stimulation for central pain. J Neurosurg 1995; 83: 1117 Canavero S, Bonicalzi V, Castellano G, et al. Painful supernumerary phantom arm following motor cortex stimulation for central post-stroke pain. Case report. J Neurosurg 1999; 91: 121-123 Canavero S, Bonicalzi V. Central pain syndrome. Pathophysiology, diagnosis and management. New York: Cambridge University Press, 2007 Douglas R, Martin K. Neocortex. In: Sheperd F (ed). The synaptic organization of the brain. New York: Oxford University Press; 1998:459-510. Drouot X, Nguyen JP, Peschanski M, Lefaucheur JP. The antalgic efficacy of chronic motor cortex stimulation is related to sensory changes in the painful zone. Brain 2002; 125: 1660-1664 Garcia-Larrea L, Peyron R, Mertens P, et al Positron emission tomography durino motor cortex stimulation for pain control. Stereotact Funct Neurosurg 1997; 68: 141-148 Garcia-Larrea L., Peyron R., Mertens P., et al. Electrical stimulation of motor cortex for pain control: a combined PET-scan and electrophysiological study. Pain 1999; 83: 259273 Garcia-Larrea L, Maarrawi J, Peyron et al. On the relation between sensory deafferentation, pain and thalamic activity in wallenbergs syndrome: a PET-scan study before and after motor cortex stimulation. Eur J Pain 2006; 10: 677-688 Hanajima R, Ashby P, Lang AE, Lozano AM. Effects of acute stimulation through contacts placed on the motor cortex for chronic stimulation. Clin Neurophysiol 2002; 113: 635641 Holsheimer J, Lefaucheur J-P, Buitenweg JR, Goujon C, Nineb A, Nguyen J-P. The role of intra-operative motor evoked potentials in the optimization of chronic cortical stimulation for the treatment of neuropathic pain. Clin Neurophysiol 2007 ; 118 : 22872296 Ikeda A, Ohara S, Matsumoto R, et al.: Role of primary somatosensory corticies in generating inhibitory motor response in humans. Brain 2000;123:1710-1721. Jones EG, Coulter JD, and Hendry SHC. Intracortical connectivity of architectonic fields in the somatic sensory. Motor and parietal cortex of monkeys. J. Comp. Neurol 1978;181:291-348. Kishima H, Saitoh Y, Osaki Y, et al. Motor cortex stimulation in patients with deafferentation pain : activation of the posterior insula and thalamus. J Neurosurg 2007; 107: 43-48
179
Maarrawi J, Peyron R, Mertens P, et al. Motor cortex stimulation for pain control induces changes in the endogenous opioid system. Neurology 2007; 69: 827-834 Manola L, Holsheimer J, Veltink P, et al. Anodal vs cathodal stimulation of motor cortex: a modeling study. Clin Neurophysiol 2007;118:464-74 Pagni CA, Canavero S. Functional thalamic depression in a case of reversibile central pain due to a spinal intramedullary cyst. J Neurosurg 1995; 83: 163-165 Peyron R, Faillenot I, Mertens P, Laurent B, Garcia-Larrea L. Motor cortex stimulation in neuropathic pain. Correlations between analgesic effect and hemodynamic changes in the brain. A PET study. Neuroimage 2007; 34: 310-321 Porter LL: Morphological characterization of a cortico-cortical relay in the cat somatosensory cortex. Cereb Cortex. 1997;7:100-109. Porter LL: Patterns of Connectivity in the Cat Sensory-Motor Cortex: A Light and Electron Microscope Analysis of the Projection Arising From Area 3a. J Comp Neurol 1991;312:404-414. Remple MS, Reed JL, Stepniewska I, and Kass JH. Organization of frontoparietal cortex in the tree shrew (Tupaia belangeri). I. Architecture, Microelectrode maps and corticospinal connections. J Comp Neurol 2006;497:133-154. Rocco MM, Brumberg JC. The sensorimotor slice. Journal of Neuroscience methods 2007;162:139-147. Schoenen J and Grant G. Spinal cord: connections. In: Paxinos G, Mai JK (eds) The human central nervous system. San Diego: Elsevier-Academic Press, 2004, 233-250 Shils JL, Mei LZ, Arle JE. Modeling parkinsonian circuitry and the DBS electrode: II. Evaluation of a Computer Simulation Model of the Basal Ganglia with and without Subthalamic Nucleus Stimulation. Stereotact Funct Neurosurg 2008;134:15-30. Sloper JJ, Hiorns RW, Powell TPS: A Qualitative and Quantitative Electron Microscopic Study of the Neurons in the Primate Motor and Somatic Sensory Cortices. Phil Trans Royal Soc London (Series B), Biol. Sci. 1979a; 285:141-171. Sloper JJ, Powel TPS. Ultrasonic features of the sensory-motor cortex of the primate. Phil Trans Royal Soc London, (Series B), Biol. Sci. 1979b;285:123-139. Sloper JJ: An electron microscopic study of the neurons of the primate motor and somatic sensory cortices. J Neurocytol 1973;2:351-359. Sol JC, Casaux J, Roux FE, et al. Chronic motor cortex stimulation for phantom limb pain: correlations between pain relief and functional imaging studies. Stereotact Funct Neurosurg 2001; 77: 172-176. Stepniewska I, Preuss TM, and Kass JH. Thalamic connections of the primary motor cortex (MI) of owl monkeys. The journal of comparative neurology 1994;349:558-582. Tsubokawa T., Katayama Y., Yamamoto T., Hirayama T., Koyama S., Treatment of thalamic pain by chronic motor cortex stimulation PACE, Pacing Clin. Electrophysiol 1991; 14: 131134 Williams SR and Sturat GJ. Role dendritic synapse location in the control of action potential output. Trends in Neuroscience 2008; 26:147-154. Zarzecki P, Shinoda Y, Asanuma H. Projection from area 3a to the motor cortex by neurons activated from group I muscle afferents. Exp Brain Res 1978;33:269-282.
MOVEMENT DISORDERS
Chapter 12
NONINVASIVE STIMULATION FOR TREATMENT OF MOVEMENT DISORDERS

Eman M. Khedr
Department of Neurology, Assiut University Hospital, Assiut, Egypt.
INTRODUCTION
In this chapter, we discuss the pathophysiology and summarize recent trials examining the clinical efficacy of rTMS for Parkinsons disease (PD) and dystonia. The goal of rTMS is the selective modulation of symptoms and their underlying neuropathophysiology on an individual basis.
PARKINSON'S DISEASE (PD)

1-Alteration of Cortical Excitability in PD Positron emission tomography (PET) (Jahanshahi et al 1995, Playford et al 1992), singlephoton emission-computed tomography (SPECT) (Rascol et al 1992, 1994) and functional magnetic resonance imaging (fMRI) (Haslinger et al 2001, Sabatini et al 2000) showed that the supplementary motor area (SMA) and the dorsolateral prefrontal cortex (DLPFC) are hypoactive during voluntary movements in PD. Because the SMA is involved in automated or complex movements (Picard et al 1996), SMA hypoactivation has been assigned a role in akinesia. Neurophysiologic recordings of the Bereitschafts-Potential (i.e. the contingent negative variation of somatosensory evoked potentials) were also consistent with SMA hypoactivity in PD (Pulvermuller et al 1996, Cunnington et al 1997, Bostantjopoulou et al
Correspondence concerning this article should be addressed to: Dr. Eman M. Khedr, MD, e-mail: emankhedr99@yahoo.com.
184
Eman M. Khedr
2000, Rossini et al 1989). Findings are more controversial for the primary motor cortex (MI). In patients with early, untreated PD, fMRI showed MI hypoactivation (Buhmann et al 2003). Conversely, in advanced parkinsonism, MI and the lateral premotor cortex (L-PMC) were found to be hyperactive (Haslinger et al 2001). These changes represent either primary or compensatory mechanisms due to treatment or adaptive motor strategies. In particular, MI hyperactivity has been attributed to cortical reorganization resulting from drug-induced reafferentation of the deficient subcortical motor system (Rascol et al 1998). However, imaging studies are based on regional cerebral blood flow measurement and therefore do not allow differentiation between afferent and local changes or between excitatory and inhibitory synaptic activity. Transcranial magnetic stimulation (TMS) studies can address some of these questions in the motor cortex. For example, electrical and magnetic stimulation techniques have shown that the cortical motor neuron connection is normal in PD (Dick et al 1984). This means that bradykinesia is not primarily the result of any deficit in the final output pathways of the motor areas of the cortex. Most authors reported that the motor cortex of patients with PD has the same threshold for stimulation as in healthy subjects (Priori et al 1994, Ridding et al 1995). However, when the patients are tested at rest, the slope of the input-output relationship between stimulus intensity and response size is steeper than normal. Perhaps as a result of this, voluntary contraction facilitates responses less than for normal subjects (Valls-Sole et al 1994). Although this could be the result of a primary basal ganglia deficit, it seems likely that it could also be an attempt to compensate for the slow recruitment of commands to move by making it easier to recruit activity from a resting state (Berardelli et al 2001). There are also changes in the excitability of cortical inhibitory circuits. A supra-threshold stimulus given whilst the subject makes a tonic voluntary contraction evokes a muscle twitch followed by a post-excitatory silent period (SP). The disappearance of voluntary activity during the silent period is thought to be due to activation of the cortical GABAergic inhibitory systems that suppress motor cortical output for 100-200 ms (Fuhr et al 1991). The silent period is shorter in bradykinetic patients (Cantello et al 1991; Priori et al 1994) and is normalized by treatment with L-dopa (Priori et al 1994). Cortical inhibition can also be tested in subjects at rest using the double-pulse paradigm of Kujirai et al (1993). Again, in patients the amount of inhibition is smaller than normal (Ridding et al 1995). It may be that these effects on inhibition reflect compensatory measures secondary to reduced facilitatory input from the basal ganglia (Berardelli et al 2001). Although this could improve the recruitment of cortical motor output during voluntary movements and partially reduce bradykinesia, it is possible that such increased excitability could contribute to rigidity.
2-Repetitive Transcranial Stimulation (rTMS) for Treatment of PD There have been many studies into the effect of single (Siebner et al 1999, 2000, De Groot et al 2001, Sommer et al 2002 and Lefaucheur et al 2004) or repeated (Lomarev et al 2006 Khedr et al 2003, 2006) rTMS interventions on different clinical parameters in PD patients including motor performance (UPDRS) (Lomarev et al 2006 Khedr et al 2003, 2006, Siebner et al 1999, 2000, De Groot et al 2001, Sommer et al 2002 and Lefaucheur et al 2004),
Noninvasive Stimulation for Treatment of Movement Disorders
185
speech (Dias et al 2006) and depression (Dias et al 2006, Fregni et al 2004, Boggio et al 2005). In addition, observations have been made on the effect of rTMS on cortical excitability (Buhmann et al 2003, Mir et al 2005) or on the release of dopamine and other neurotransmitters in the brain (Strafella et al 2001, 2003, 2004) and serum (Khedr et al 2007). Although initial studies tended to employ non-focal coils (circular), later studies used focal (figure-of-eight coil), low or high frequency (0.2-25 Hz) stimulation of a variety of different brain areas (Primary motor area (MI), premotor cotex (PMC), supplementary motor area (SMA), dorsolateral prefrontal cortex (DLPFC) (see Tables 1 and 2). Table 1. rTMS in patients with PD using a circular coil
Authors / year Target Number of patients 7 Stimulus frequency 20, 10, 5, 1Hz Total pulse number 1000 Results
Tergau et al., 1999
Vertex
Mally and Stone, 1999a,b Shimamoto et al., 2001 Dragasevic et al., 2002 Ikeguchi et al., 2003 Okabe et al., 2003 Mally and Stone, 2004
vertex
10, 49
1Hz
30 2 per day 710 days 30 2 per week 2 months 200 per day 10 days 30 2/2days 2 weeks 50 x 2 per week x 2 months 30 2 per day 710 days, 2 per year, 3 years 150x4
No-significant effect on UPDRS-III and reaction time UPDRS-III, HY and ADL improvement UPDRS-III, HY and ADL improvement UPDRS-III and depression score improvement UPDRS-III and ADL improvement No-significant effect on UPDRS-III Reduction in HY deterioration and in levodopa dose increase due to disease dvance Increase in dopamine neurotransmission in the putamen of both hemispheres (reduced [11C]raclopride binding)
vertex DLPFC
9 real, 9 sham 10
0.2 Hz 78% MO 0.5 Hz
Frontal vertex vertex
12 85 46
0.2 Hz 70% MO 0.2 Hz 110% MT 1Hz
Strafella et al., 2005
Left and right motor cortex
10Hz
DLPF: Dorsolateral prefrontal area; UPDRS-III: unified Parkinsons disease rating scale, motor score. MO: maximum power output of stimulator; MT: resting motor threshold. Only studies performed with a round coil, based on clinical assessment (UPDRS depression score) or assessment of cortical excitability; HY: Hoehn & Yahr.
The first rTMS study in PD showed improvement of reaction and movement times during MI stimulation at high frequency (5Hz) using a round coil (Mally et al 1999). However, these data could not be replicated in a later study. A second early report came from Mally and Stone (Mally et al 1999a, Bornke et al 2004) who applied low frequency 1 Hz rTMS (60 pulses /day for 7 days) and recorded improvement in UPDRS. Other work has confirmed that the improvement can last beyond the time of stimulation (Siebner et al 1999) and reach
186
Eman M. Khedr
significant level on clinical motor scores after single sessions of rTMS (5Hz) ( Siebner et al 2000) and after repeated rTMS sessions delivered at high frequency over MI (Khedr et al 2003, 2006, Lomarev et al 2006, Bornke et al 2004, Lefaucheur et al 2004, Dias et al 2006). It was found that speech could be improved in PD patients with dysphonia using rTMS centered over the cortical representation of the mouth (Dias et al 2006). Table 2. rTMS in patients with PD (figure-of-eight coil)
Authors / year Target Number of patients 6 Stimulus frequency 5 Hz Total pulse number Results
Pascual-Leone et al., 1994 Siebner et al., 1999 Ghabra et al., 1999 Siebner et al., 2000 Siebner et al., 2000 de Groot et al., 2001 Gilio et al., 2002 Sommer et al., 2002 Khedr et al., 2003
MI MI MI MI
12 11 10
5 Hz 90% MT 5 Hz 80-85% MT 5 Hz
750 750
Movement and reaction time shortening Movement time shortening No effect on movement (time during rTMS) UPDRS-III improvement (reduced rigidity and bradykinesia contralateral to the stimulation) Prolonged cortical silent period UPDRS-III and writing improvement No MEP amplitude increase during Transient improvement in frequency of finger tapping UPDRS-III and walking speed improvement UPDRS-III improvement, especially in OFF time UPDRS-III improvement (reduced rigidity and bradykinesia contralateral to the stimulation) UPDRS-III improvement (reduced rigidity bilaterally) Speech improvement
MI MI MI MI MI
10 9 15 11 36 (19 active, 17 sham) 12 12
5 Hz 90% MT 5 Hz 5 Hz 1 Hz 120% MT 5 Hz 120% MT 10 Hz 10 Hz 80% MT 0.5 Hz 80% MT 5 Hz
2250 (10 min) 2250 40 900 2000/ day x 10 days 1000 3.3 min. 2000
Boernke et al., MI 2004 Lefaucheur et al., MI 2004 Lefaucheur et al., MI 2004 Dias et al., 2006 MI
12 20 (10 active, 10 sham) 45 (35 active, 10 sham 20 (10 active, 10 sham) 10 10 8
600 2250
Khedr et al., 2006 Khedr et al., 2006 Boylan et al., 2001 Koch et al., 2004 Koch et al., 2005
MI
25 Hz 100% MT 10 Hz 100% MT 10 Hz 5 Hz 5 Hz 90% MT
300/ day x 6 days 3000/ day x 6 days 2000, 1 session 40 min. 2500 900 1 session 15 min.
MI
UPDRS-III, walking speed, keytapping speed and self assessment scale improvement UPDRS-III mild improvement
SMA SMA SMA
Increase in reaction time and deterioration of writing No significant effect on time perception No significant effect on dyskinesia and UPDRS-III

Koch et al., 2005 SMA 8 1 Hz 90% MT 1 Hz 80% MT Mir et al., 2005 PMC 10 5 Hz 90% MT Fregni et al., 2004 DLPFC 42 (21 active, 21 sham) 15 Hz 110% MT 3000/ day x 10 days 300x5 900 1 session 15 min. 1200x2 1 week apart
187
Reduced dyskinesia without effect on UPDRS-III Changes in motor cortex excitability, as assessed by a TMS paired-pulse paradigm, depending on levodopa intake Normalization of motor evoked potential amplitude in on-drug patients, but no effect in off-drug patients Improvement of depression score similar to fluoxetine, with better results than fluoxetine on cognitive and motor scores. Small improvement in motor function, no change on UPDRS Time perception improvement Improvement in neuropsychological testing similar to fluoxetine No significant effects on UPDRS Transiently amelioration of peakdose LID Improvement of depression score without speech effects Improvement in walking time and reduced bradykinesia for the right hand No significant effect on motor performance UPDRS significantly improved when OFF, Hamiltons DRS improved
Buhmann et al., 2004
PMC
10
Koch et al., 2004 Boggio et al., 2005 Brusa et al., 2006
DLPFC DLPFC
SMA
10 25 (13 active, 12 sham) 10
5 Hz 15 Hz
2500 3000/ day x 10 days 4500 10 sessions over 10 days 3000 (10 sessions over 2 weeks) 1200/ day x 8 days
1 Hz 90% MT
Dias et al., 2006
DLPFC
Lomarev et al., 2006
Del Olmo et al., 2007 Epstein et al., 2007
MI + DLPFC (combined stimulation) DLPFC 13 (8 active, 5 sham) DLPFC 14
20 (10 active, 10 sham) 18
15 Hz 110% MT 25 Hz 100% MT 10 Hz 90% MT 10 Hz 110% MT
4500
19000 10 sessions over 2 weeks
MI: primary motor cortex stimulation; SMA: Supplementary motor area stimulation; PMC: premotor cortex stimulation; DLPF: dorsolateral prefrontal area; UPDRS-III: Unified Parkinsons Disease Rating Scale, motor score; MT: resting motor threshold. Only studies performed with a figure-of-eight coil, based on clinical assessment (UPDRS depression score) or assessment of cortical excitability; DRS: Depression Rating Scale.
When applied over SMA, rTMS worsened the motor performance of PD patients at high frequency (10Hz) (Boylan et al 2001), but improved apomorphine-induced dyskinesia at low frequency (1Hz) (Koch et al 2005). The premotor cortex (PMC) has not yet been targeted in clinical studies of PD patients, although low-frequency PMC stimulation has been shown to enhance intracortical motor inhibitory processes more efficaciously than direct MI stimulation in normals (Bumer et al 2003). However, the premotor-motor interaction is absent in off-drug PD patients and only partially restored by levodopa intake (Buhmann et al 2003, Mir et al 2005). Parenthetically,
188
Eman M. Khedr
low frequency rTMS over PMC (1800 pulses, 1 Hz, 90% MT) led to a marked improvement of postural tremor in a single patient suffering cortical tremor, with longer benefit when stimulating during two daily sessions (Houdayer et al 2007). At low frequency (0.5Hz), DLPFC stimulation was initially found to increase motor performance concomitantly with depression relief (Dragasevic et al 2002). However, subsequent studies showed that DLPFC stimulation induced antidepressant and cognitive effects, but no motor effects in PD patients (Fregni et al 2004, Boggio et al 2005, Koch et al 2004, del Olmo et al 2007). Two independent studies using very similar parameters of rTMS reported analogous clinical effects. In a carefully controlled recent study, Lomarev et al (2006) reported that repeated sessions (8 over a 4-week period) of rTMS at 25 Hz to bilateral MI and DLPFC led to cumulative benefits on motor function, including walking speed, in patients with PD and that these continued for at least 1 month after the end of treatment. Khedr et al (2006) performed a similar study with 25 Hz stimulation over MI for 5 consecutive days. Essentially, the results confirm those of Lomarev et al (2006) and emphasize the importance of repeated sessions of rTMS to produce stable after-effects. However, the data of Khedr et al (2006) also extend the findings to show that the effects occur in patients withdrawn from medication for at least 1 week prior to treatment; and that they occur in early- as well as late-stage disease. This suggests that intrinsic levels of dopamine have little influence on the response to rTMS. Indeed, Lomarev et al (2006) had applied rTMS to medicated patients at the start of their on period. This was because they had noted a better effect of tDCS on parkinsonian symptoms when it was applied at this timing in a previous study. However, the data of Khedr et al (2003, 2006) showed that it is not necessary to apply rTMS to medicated patients. In another controlled study, rTMS applied over DLPFC (5 sessions, 10 Hz) induced a trend towards improvement of L-Dopa-induced dyskinesias (Rektorova et al 2008). Khedr et al (2006) examined the effects of two different frequencies of stimulation. They showed that for all measures, 25 Hz rTMS had a better therapeutic effect than 10 Hz rTMS. However, this does not necessarily mean that 25 Hz is the best possible frequency: in a previous study (2003), 5 Hz stimulation (albeit over 10 days rather than 6 and at 120% RMT rather than 100% RMT) gave an effect size similar to that of 25 Hz. Thus, more than one combination of frequency/intensity may be useful for treatment. It may also be, as suggested by Lefaucheur et al (2004), that different symptoms respond better to different parameters of rTMS. Clearly, more work needs to be done to address these questions. Khedr et als data also showed that, although the clinical effect may begin to decline after 1 month, they can be boosted by a short three-session periods of rTMS. Although some trials report positive effects of rTMS on parkinsonian symptoms, this is not always the case. Indeed, two studies with relatively large sample sizes showed negative results (Fregni et al 2004, Okabe et al 2003). One of the reasons to explain this contradiction might be the interface of anti-PD drugs versus TMS, as these studies assessed UPDRS after the use of levodopa (on state). This medication might mask the effects of TMS due to a ceiling effect. Thus, Bornke et al (2004) compared rTMS to levodopa challenge and did not find any significant difference. It is possible that optimization of the TMS treatment protocol and patient selection could result in benefits of greater magnitude. However, it is premature to say that long-term treatment with TMS might be as efficacious as levodopa. Although a
189
retrospective study (Mally 1999a) raised the possibility that rTMS combined with drugs can slow the development of PD, a proper clinical trial with an adequate sample size, methodology and a long follow-up comparing TMS and the pharmacologic treatment would be desirable.
3-Mechanisms of Action of rTMS on PD A general conclusion from all the previous studies seems to be that low frequency rTMS reduces rigidity and restores intracortical inhibition, whereas movement preparation and execution benefit from a global increase in cortical excitability induced by high-frequency rTMS. However, the effects of cortical stimulation are probably not restricted to the cortex. The motor cortex can influence basal ganglia activities through various glutamatergic corticostriatal and corticosubthalamic projections. Since the dopaminergic deficiency in PD is localized to the subcortical basal ganglia, beneficial effects of rTMS on PD motor symptoms are necessarily somewhat indirect (Fregni et al 2005). Three mechanisms can be proposed to explain how cortically directed rTMS can improve PD symptoms: (1) Normalization of network activation: rTMS induces network changes that relate to and positively affect basal ganglia function (Mottaghy et al 2000, Paus et al 2001); (2) Normalization of excitability (e.g. increase of MI excitability along with improved hand bradykinesia in Lomarev et als study); (3) Striatal dopamine release (presynaptic effect) by relatively preserved dopaminergic neurons, as demonstrated in healthy subjects with PET after MI or DLPFC stimulation (Strafella et al 2001, 2003, 2004). Although the pathophysiologic mechanisms of rTMS effects on PD remain uncertain, it does appear reasonable to assume that the modulation of the motor cortex can modulate basal ganglia function as well. The cumulative long lasting rTMS effect is probably explained by brain mechanisms different from those responsible for single session effects. The similarity between electroconvulsive therapy (ECT) and rTMS has been often emphasized, including studies in PD (Khedr et al 2003, Mally et al 1999a). Long-lasting improvement in PD was found after ECT in patients who were receiving L-dopa/carbidopa medication (Balldin et al 1980, Fall et al 1995). Khedr et al (2007) showed that the improvement in UPDRS was paralleled by an increase in plasma levels of dopamine and that these levels correlated with clinical status before and after treatment. Although they only measured dopamine in plasma, it seems likely that the observed changes reflected changes in cerebral dopamine levels. This is because a number of previous reports have shown that a single session of rTMS over frontal and central areas of cortex can increase dopamine release in the human striatum (Strafella et al 2001, 2003, Pogarell et al 2006). 10-Hz rTMS over the frontal cortex can induce focal dopamine release in subcortical basal ganglia structures and potentially improve PD. Given the correlations between dopamine levels and UPDRS scores at the time of testing, it seems possible that some of the clinical improvement in patients following rTMS may have been due to changes in cerebral dopamine, with increases in dopamine causing improvement in motor performance (Khedr et al 2007). We can only speculate on how this might occur. It is known, for example, that a single session of rTMS can influence levels of brain derived
190
Eman M. Khedr
neurotrophic factor (BDNF) in the human brain (Angelucci et al 2004) and that BDNF can prevent or restore damage to dopamine neurons due to environmental or chemical toxins. Thus, it is possible that repeated sessions of rTMS could improve growth and enhance activity of dopaminergic neurons in the brains of PD patients via actions on e.g. growth factors (GDNF, BDNF). If so, then, this could lead to enhanced dopamine function and improve the clinical status for long periods after completion of treatment. Also, expansion of short term range of plasticity (metaplasticity) may play a role and this could be due to, for instance, upregulation of NMDA receptors (Abraham et al 2001). Another mechanism might be that repeated episodes of long-term potentiation lead to remodeling with an increase in active synapses (Leuner et al 2004, Matsuzaki et al 2004).
4-TDCS in Parkinsons Disease Fregni et al (2006) studied the motor effects of single-session tDCS of MI and DLPFC in PD patients in the OFF state. Anodal tDCS of MI significantly enhanced motor function (assessed by UPDRS motor scores) in PD compared with sham stimulation, whereas cathodal stimulation of MI and anodal stimulation of the DLPFC were poorly effective. Whereas anodal stimulation resulted in a robust increase of corticospinal excitability, cathodal stimulation slightly decreased it. Intriguingly, increase in MI excitability after anodal tDCS was marginally correlated with motor function improvement. Boggio et al (2006) found a significant improvement in working memory after active anodal tDCS of the left DLPFC with 2 mA (but not 1 mA) in PD. The other conditions of stimulation (sham tDCS, anodal tDCS of the left DLPFC with 1 mA, or anodal tDCS of MI) did not result in a significant task performance change.
DYSTONIA
1-Alteration of Cortical Excitability in Dystonia The predominant pattern in dystonia is one of context-related excess activation of primary sensory, motor, and premotor cortices. These overactivation patterns are often interpreted as loss of specificity in muscle activation, excess maladaptive plasticity in the motor cortex, or increased difficulty of task. It has been shown that sensory input from the affected hand can trigger dystonic symptoms in writer's cramp patients (Kaji et al., 1995). Frequent use of a body part or a disease in the peripheral nervous system often precedes the manifestation of dystonia (Quartarone et al., 1998). Initial studies suggested that at rest there is little difference in regional cerebral blood flow in patients with primary dystonia and normal subjects (Lang et al 1988; Karbe et al 1992; Ceballos-Baumann et al 1995). Early reports of abnormalities in resting glucose metabolism were contradictory (e.g. Chase et al 1988: increased; Karbe et al 1992: decreased; Otsuka et al 1992: unchanged), although work in patients with torticollis showed increased metabolism of the lentiform nucleus (Galardi et al., 1996; Magyar-Lehmann et al., 1997).
191
Using a principal components method, Eidelberg et al (1995) found some evidence for a relative putaminal hypermetabolism, which has recently been linked to thalamic hypometabolism. This has been interpreted in terms of increased activity in the direct striatopallidal pathway, inhibiting the internal segment of the globus pallidus. As a result, there would be increased inhibitory synaptic activity in the medial globus pallidus (and substantia nigra pars reticulata) and reduced activity in pallidal output to the thalamus. TMS studies in focal dystonia disclosed some aspects of motor cortical hyperexcitability, such as excessive MEP facilitation, short CSPs, and reduced intracortical inhibition. Sometimes, alterations emerged just while the dystonic contraction occurred, suggesting undue overflow of motor commands from the cortex. These results fit the finding of reduced GABA in sensorimotor cortex and basal ganglia structures, as emerging from magnetic resonance spectroscopy studies (Levy and Hallett, 2002). Mavroudakis et al (1995) and Ikoma et al (1996) have investigated the inputoutput relationships of the system in more detail with magnetic stimulation. They found that the size of responses increased more steeply with increasing levels of background contraction in patients with primary dystonia, or with increasing stimulus intensity, than in normal subjects. The cortical output gain was higher in patients, and this may contribute to the excessive motor output that occurs during voluntary movement in dystonia. Among writers cramp patients, silent periods are shorter, and short-interval intracortical inhibition is reduced, compared with control subjects ( Filipovic et al 1997, Ridding et al 1995, Siebner et al 1999). This disinhibition can be task-specific within the same hand muscle (Tinazzi et al 2005). Treatment with botulinum toxin may transiently normalize intracortical inhibition in association with clinical benefit (Gilio et al 2000). Consistent with altered sensorimotor integration, the normal inhibition of excitability after a peripheral electrical stimulus becomes facilitated in patients with dystonia (Kessler et al 2005). Repeated pairs of peripheral electrical stimulation synchronized with cortical TMS pulses, that normally produce robust facilitation in excitability, resulted in further exaggerated excitability in patients with dystonia (Quartarone et al 2003). Similarly, although muscle vibration normally facilitates cortical excitability to that muscle, excitability was suppressed in patients with focal dystonia (Rosenkranz et al 2000). These abnormal responses suggest that maladaptive plasticity may be a fundamental mechanism of symptomatic dystonia (Edwards 2006).
2-Repetitive Transcranial Stimulation (rTMS) for Treatment for Dystonia Downregulation of excess plastic responses in a disorder in which disinhibition predominates can be a challenge. Low-frequency rTMS, which typically reduces excitability in normal subjects, may not necessarily have the same effect in dystonic patients. Several studies, however, have shown the potential for normalization of excitability or of abnormal network patterns with low-frequency rTMS over MI or PMd. A few preliminary studies and case reports point in this direction (Table 3). Generalization from these studies, and future directions, must take into account the heterogeneity of dystonia and lack of standardized protocols for assessing outcomes. For
192
Eman M. Khedr
example, the two controlled studies focused on writers cramp and analysis of handwriting (Siebner et al 1999, Murase et al 2005), whereas the two case reports suggested different patterns of benefit (pain: Lefaucheur et al 2004; neck dystonia: Allam et al 2007) on patients with different dystonia diagnoses. Edwards et al. (2004) reported that long lasting aftereffects of rTMS were enhanced in patients with primary dystonia. Table 3. rTMS in dystonia patients (figure-of-eight coil)
Authors / year Target Number of patients/ type of dystonia 16 patients /writer's cramp and 11 control 3 patients /generaliged dystonia 9 patients/ writer's cramp Stimulus frequency Total pulse number Results
Siebner et al (1999) Lefaucheur et al ( 2004) Murase et al (2005)
MI
1Hz 90% MT 1Hz 90% MT 0.2 Hz 85% MT
PMd
1800 (1 session in 30 min) 1200 per day 5days 250/site (1 session in 21 min) 1200 per day 5 days
PMd/S MA
Allam et al (2007)
PMd
1 patient with cervical dystonia + writers cramp.
1Hz at 90% of MT
Significant improvement in handwriting and change in excitability Slight changes in dystonia rating scales+ radically reduced pain Significant improvement in handwriting measures over PMd only + changes in excitability Cervical dystonia significantly improved, no effect on writers cramp
MI: primary motor cortex; PMd: dorsal premotor cortex; SMA: supplementary motor area; MT: resting motor threshold.
Consistent with maladaptive plasticity in dystonia, MI excitability often responds inappropriately, often with disinhibition, after a train of rTMS. After several trains of suprathreshold 1-Hz rTMS over MI, cortical excitability was suppressed in control subjects, but facilitated in writers cramp patients (Siebner et al 1999). With much lower thresholds, 1Hz rTMS over MI did not alter any of several measures of excitability in focal hand dystonia patients, even though the excitability in control subjects was reduced (Stinear and Byblow 2004). Brief trains of up to 20 pulses of suprathreshold 1 Hz did not alter excitability, but at 5 Hz it resulted in an exaggerated, and longer-lived, excitability facilitation, compared with control subjects (Gilio et al 2007). The dorsal premotor cortex (PMd) has dense reciprocal connections with both MI and SMA within the abnormal dystonic cortical network. As with MI, rTMS over PMd can modulate MI excitability bidirectionally, with low frequencies suppressing and high frequencies enhancing excitability (Rizzo et al 2004). However, rTMS effects over PMd can induce lasting effects on cortical excitability, often to a greater degree than with rTMS over MI itself (Gerschlager et al 2001). With PET scanning, subthreshold 1-Hz rTMS over the left PMd was shown to reduce rCBF in the left sensorimotor cortex, left PMd, SMA, and cerebellum both in normal control subjects and in patients with focal dystonia (Siebner et al 2003). However, the decrease was significantly greater among patients in bilateral PMd, SMA, and precuneus. Findings support the use of 1-Hz rTMS over PMd to trigger widespread inhibition throughout the motor network in dystonia patients. We note that 1-Hz
193
rTMS over the PMd also normalizes spinal reflexes that are abnormal in patients with DYT1 generalized dystonia (Huang et al 2004). Therefore, effects of rTMS over the PMd extend down the neuroaxis and provide evidence that modulation at this cortical site can normalize sensorimotor integration at multiple neuroanatomic levels. Testing low-intensity 0.2-Hz rTMS over the MI, SMA, and PMd in separate sessions with writers cramp patients, Murase et al (2005) found that only stimulation over the PMd site prolonged silent periods and improved handwriting. Two open case studies reported effects of five daily sessions of 1-Hz rTMS over the left PMd. After rTMS, three patients with severe, generalized, secondary dystonia showed reduced painful axial spasms, but less consistent reductions of abnormal movements or disability (Lefaucheur et al 2004), while a patient with primary dystonia affecting neck and limb showed improvement of neck, but not limb, dystonia symptoms (Allam et al 2007). These studies provide encouraging data in support of the use of multisession rTMS over PMd to modulate dystonia. However, the topographic nature of rTMS modulation over the left PMd and the relationship between rTMS and pain or disability in these patients remain to be clarified.
3-Mechanisms of rTMS in Dystonia In a PET study, 1 Hz rTMS of premotor cortex decreased metabolic activity at both the site of stimulation and connected areas at a distance, this effect being larger in patients with focal hand dystonia than in healthy controls (Siebner et al., 2003). Nonetheless, one study has suggested some clinical benefit in dystonia patients after low-frequency rTMS over MI in association with normalization of this abnormal disinhibition (Siebner et al 1999). Among writers cramp patients, single-sessions of subthreshold 1-Hz rTMS over MI normalized intracortical inhibition and also prolonged silent periods. Patients showed a benefit in mean writing pressure, with several showing clear improvements in handwriting. The normalization of excitability by rTMS in this study are in contrast with other studies, which showed either facilitation or no change in excitability in dystonia patients after 1-Hz rTMS (Siebner et al 1999, Stinear and Byblow 2004). Variation in intensity and excitability measures and in study design may account for these differences; replication is needed.
CONCLUSION AND RECOMMENDATIONS

rTMS and tDCS may provide a future option for adjunctive therapy in PD and dystonia. Although previous studies with clinical outcomes showed the potential for benefit, many were small and without control groups, with many different variables in the selection of patients and methodological application of rTMS. PD patients can show marked and sustained improvements on objective measures with placebo treatment even in rigorously controlled and blinded trials (Goetz et al 2000). However, in clinical practice, noninvasive stimulation could become a tool to select responders for implanted MI ECS (Canavero and Bonicalzi 2007 and chapter 13). However, repeated rTMS sessions (up to 30) can also lead to
194
Eman M. Khedr
cumulative and long-lasting effects on motor performance (up to months: Khedr et al 2003, 2007; Lomarev et al 2006) that may reach a therapeutic level. This might be considered as an alternative to functional neurosurgery in case of surgical contraindications. Future studies will have to better address functional neuroimaging and cortical excitability outcome measures and correlate them with outcome.
REFERENCES
Abraham WC, Mason-Parker SE, Bear MF, Webb S Tate WP. Heterosynaptic metaplasticity in the hippocampus in vivo: a BCM-like modifiable threshold for LTP. Proc Natl Acad Sci USA 2001; 98: 10924-10929. Alexander GE, Crutcher MD. Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends Neurosci 1990;13:26671. Allam N, Brasil-Neto JP, Brandao P, Weiler F, Barros Filho J, Tomaz C. Relief of primary cervical dystonia symptoms by low frequency transcranial magnetic stimulation of the premotor cortex: case report. Arq Neuropsiquiatr 2007;65:697 699. Angelucci F, Oliviero A, Pilato F, et al. Transcranial magnetic stimulation and BDNF plasma levels in amyotrophic lateral sclerosis. euroreport. 2004;15:717-20 Balldin J, Eden S, Granerus AK, et al . Electroconvulsive therapy in Parkinson,s syndrome with on-off phenomenon. J Neural Transm. 1980;47: 11-21. Bumer T, Lange R, Liepert J, et al. Repeated premotor rTMS leads to cumulative plastic changes of motor cortex excitability in humans. Neuroimage 2003;20:550-560. Berardelli A, Rothwell JC, Thompson PD, and Hallett M. Pathophysiology of bradykinesia in Parkinson's disease. Brain 2001; 124, 2131-2146. Boggio PS, Fregni F, Bermpohl F, et al. Effect of repetitive TMS and fluoxetine on cognitive function in patients with Parkinson's disease and concurrent depression. Mov Disord 2005;20:1178-1184. Boggio P.S., Ferrucci R., Rigonatti S.P., et al. Effects of transcranial direct current stimulation on working memory in patients with Parkinsons disease. J Neurol Sci 2006; 249: 3138 Bornke Ch, Schulte T, Przuntek H, Muller T: Clinical effects of repetitive transcranial magnetic stimulation versus acute levodopa challenge in Parkinson's disease. J Neural Transm Suppl 2004;68:61-67. Bostantjopoulou S, Katsarou Z, Zafiriou D, et al. Abnormality of N30 somatosensory evoked potentials in Parkinsons disease: a multidisciplinary approach. Neurophysiol Clin 2000;30:36876. Boylan LS, Pullman SL, Lisanby SH, Spicknall KE, Sackeim HA: Repetitive transcranial magnetic stimulation to SMA worsens complex movements in Parkinson's disease. Clin Neurophysiol 2001;112:259-264. Brusa L., Versace V., Koch G., et al. Low frequency rTMS of the SMA transiently ameliorates peak-dose LID in Parkinsons disease Clin Neurophysiol. 2006; 117: 1917 1921
195
Buhmann C, Glauche V, Sturenburg HJ, Oechsner M, Weiller C, Buchel C. Pharmacologically modulated fMRIcortical responsiveness to levodopa in drug-naive hemiparkinsonian patients. Brain 2003;126:45161. Buhmann C, Gorsler A, Baumer T, et al. Abnormal excitability of premotor connections in the de novo Parkinsons disease. Brain 2004;124:2732-46. Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, 223-232 Cantello R, Gianelli M, Bettucci D, Civardi C, De Angeis MS, Mutani R. Parkinson's disease rigidity: magnetic motor evoked potentials in a small hand muscle. Neurology 1991; 41: 1449-1456. Ceballos-Baumann AO, Passingham RE, Warner T, Playford ED, Marsden CD, Brooks DJ. Overactive prefrontal and underactive motor cortical areas in idiopathic dystonia. Ann Neurol 1995; 37: 36372. Chase TN, Tamminga CA, Burrows H. Positron emission tomographic studies of regional cerebral glucose metabolism in idiopathic dystonia. In: Fahn S, Marsden CD, Calne DB, editors. Dystonia 2. Advances in neurology, Vol 50. New York: Raven Press; 1988. p. 23741. Chen R, Kumar S, Garg RR, Lang AE: Impairment of motor cortex activation and deactivation in Parkinson's disease. Clin Neurophysiol 2001;112:600-607. Cunnington R, Iansek R, Johnson KA, Bradshaw JL. Movement-relatedpotentials in Parkinsons disease. Motor imagery and movement preparation. Brain 1997;120:1339 53. De Groot M, Hermann W, Steffen J, Wagner A, Grahmann F. Contralateral and ipsilateral repetitive transcranial magnetic stimulation in Parkinson patients. Nervenarzt. Dec, 2001;72:932-938. del Olmo FM, Bello O, Cudeiro J. Transcranial magnetic stimulation over dorsolateral prefrontal cortex in Parkinsons disease. Clin Neurophysiol. 2007;118:131-9. Dias AE, Barbosa ER, Coracini K, Maia F, Marcolin MA, Fregni F: Effects of repetitive transcranial magnetic stimulation on voice and speech in Parkinson's disease. Acta Neurol Scand 2006;113:92-99. Dick JP, Rothwell JC, Day BL, et al. The Bereitschaftspotential is abnormal in Parkinson's disease. Brain 1989;112:233-244. Dragasevic N, Potrebic A, Damjanovic A, Stefanova E, Kostic VS. Therapeutic efficacy of bilateral prefrontal slow repetitive transcranial magnetic stimulation in depressed patients with Parkinsons disease: an open study. Mov Disord 2002;17: 52832. Edwards MJ, Huang YZ, Mir P, Rothwell JC, Bhatia KP. Abnormalities in motor cortical plasticity differentiate manifesting and nonmanifesting DYT1 carriers. Mov Disord 2006;21:21812186. Eidelberg D, Moeller JR, Ishikawa T, et al. The metabolic topography of idiopathic torsion dystonia. Brain 1995; 118: 147384. Fall PA, Ekman R, Granerus AK, Thorell LH, Walinder J, ECT in Parkinsons disease. Changes in motor symptoms, monoamine metabolites and neuropeptides. J Neural Transm Park Dis Dement Sect. 1995;10:129-140.
196
Eman M. Khedr
Filipovic SR, Ljubisavljevic M, Svetel M, Milanovic S, Kacar A, Kostic VS. Impairment of cortical inhibition in writers cramp as revealed by changes in electromyographic silent period after transcranial magnetic stimulation. Neurosci Lett 1997;222:167 170. Fregni F, Santos CM, Myczkowski ML, et al. Repetitive transcranial magnetic stimulation is as effective as fluoxetine in the treatment of depression in patients with Parkinson's disease. J Neurol Neurosurg Psych 2004;75:1171-1174. Fregni F, Simon DK, Wu A, Pascual-Leone A. Non-invasive brain stimulation for Parkinsons disease: a systematic review and meta-analysis of the literature. J Neurol Neurosurg Psych 2005; 76: 1614-1623 Fregni F., Boggio P.S., Santos M.C., et al. Noninvasive cortical stimulation with transcranial direct current stimulation in Parkinsons disease. Mov Disord 2006; 21:16931702 Fuhr P, Agostino R, Hallett M. Spinal motor neuron excitability during the silent period after cortical stimulation. Electroencephalogr Clin Neurophysiol 1991; 81: 257-262. Galardi G, Perani, D, Grassi F, et al. Basal ganglia and thalamo-cortical hypermetabolism in patients with spasmodic torticollis. Acta Neurol Scand 1996; 94: 17267. Gerschlager W, Siebner HR, Rothwell JC. Decreased corticospinal excitability after subthreshold 1 Hz rTMS over lateral premotor cortex. Neurology 2001;57:449455. Ghabra MB, Hallett M, Wassermann EM. Simultaneous repetitive transcranial magnetic stimulation does not speed fine movement in PD. Neurology 1999;52:76870. Gilio F, Curr A, Inghilleri M, et al. Abnormalities of motor cortex excitability preceding movement in patients with dystonia. Brain 2003;126:17451754. Gilio F, Curra A, Inghilleri M, Lorenzano C, Manfredi M, Berardelli A. Repetitive magnetic stimulation of cortical motor areas in Parkinsons disease: implications for the pathophysiology of cortical function. Mov Disord 2002; 17:46773. Gilio F, Curr A, Lorenzano C, Modugno N, Manfredi M, Berardelli A. Effects of botulinum toxin type A on intracortical inhibition in patients with dystonia. Ann Neurol 2000;48:206. Gilio F, Suppa A, Bologna M, Lorenzano C, Fabbrini G, Berardelli A. Short-term cortical plasticity in patients with dystonia: a study with repetitive transcranial magnetic stimulation. Mov Disord 2007; 22:1436-1443. Goetz CG, Leurgans S, Raman R, Stebbins GT. Objective changes in motor function during placebo treatment in PD. Neurology. 2000;54:710-4. Haslinger B, Erhard P, Kampfe N, et al. Event-related functional magnetic resonance imaging in Parkinsons disease before and after levodopa. Brain 2001;124:55870. Houdayer E, Devanne H, Tyvaert l, Defebvre L, Derambure P, Cassim F. Low frequency repetitive transcranial magnetic stimulation over premotor cortex can improve cortical tremor. Clin Neurophysiol 2007; 118: 1557-1562 Huang YZ, Edwards MJ, Bhatia KP, Rothwell JC. One-Hz repetitive transcranial magnetic stimulation of the premotor cortex alters reciprocal inhibition in DYT1 dystonia. Mov Disord 2004; 19:54 59. Ikeguchi M, Touge T, Nishiyama Y, Takeuchi H, Kuriyama S, Ohkawa M. Effects of successive repetitive transcranial magnetic stimulation on motor performances and brain perfusion in idiopathic Parkinsons disease. J Neurol Sci 2003;209:416.
197
Ikoma K, Samii A, Mercuri B, Wassermann EM, Hallett M. Abnormal cortical motor excitability in dystonia. Neurology 1996;46:13716 Jahanshahi M, Jenkins IH, Brown RG, Marsden CD, Passingham RE, Brooks DJ. Selfinitiated versus externally triggered movements. I. An investigation using measurement of regional cerebral blood flow with PET and movement-related potentials in normal and Parkinsons disease subjects. Brain 1995; 118:91333. Kaji R, Ikeda A, Ikeda T, et al. Physiological study of cervical dystonia. Task-specific abnormality in contingent negative variation. Brain 1995; 118: 51122. Karbe H, Holthoff VA, Rudolf J, Herholz K, Heiss WD. Positron emission tomography demonstrates frontal cortex and basal ganglia hypometabolism in dystonia. Neurology 1992; 42: 15404. Kessler KR, Ruge D, Ilic TV, Ziemann U. Short latency afferent inhibition and facilitation in patients with writers cramp. Mov Disord 2005;20:238 242. Khedr EM, Farweez HM, Islam H. Therapeutic effect of repetitive transcranial magnetic stimulation on motor function in Parkinson's disease patients. Eur J Neurol 2003;10:567572. Khedr EM, Rothwell JC, Shawky OA, Ahmed MA, Hamdy A. Dopamine levels after repetitive transcranial magnetic stimulation of motor cortex in patients with Parkinson's disease: preliminary results. Mov Disord 2007; 22: 10461050 Khedr EM, Rothwell JC, Shawky OA, Ahmed MA, Hamdy A. Effect of daily repetitive transcranial magnetic stimulation on motor performance in PD. Mov Disord 2006; 21: 22012205 Koch G, Brusa L, Caltagirone C, et al. rTMS of supplementary motor area modulates therapy-induced dyskinesias in Parkinson disease. Neurology 2005;65: 623-625. Koch G, Oliveri M, Brusa L, Stanzione P, Torriero S, Caltagirone C. High-frequency rTMS improves time perception in Parkinson disease. Neurology 2004;63:2405-2406. Kujirai T, Caramia MD, Rothwell JC, et al. Corticocortical inhibition in human motor cortex. J Physiol. (London) 1993; 471: 501-519. Lang AE, Garnett ES, Firnau G, Nahmias C, Talalla A. Positron tomography in dystonia. In: Fahn S, Marsden CD, Calne DB, editors. Dystonia 2. Advances in neurology, Vol. 50. New York: Raven Press; 1988. p. 24953. Lefaucheur JP, Drouot X, Von Raison F, Mnard-Lefaucheur I, Cesaro P, Nguyen JP. Improvement of motor performance and modulation of cortical excitability by repetitive transcranial magnetic stimulation of the motor cortex in Parkinsons disease. Clin Neurophysiol 2004;115:2530-2541. Lefaucheur JP, Fnelon G, Mnard-Lefaucheur I, Wendling S, Nguyen JP. Low-frequency repetitive TMS of premotor cortex can reduce painful axial spasms in generalized secondary dystonia: a pilot study of three patients. Neurophysiol Clin 2004;34: 141145. Lefaucheur JP. Motor cortex dysfunction revealed by cortical excitability studies in Parkinsons disease: influence of antiparkinsonian treatment and cortical stimulation. Clin Neurophysiol 2005;116:244-253. Leuner B, Shors TJ, New spines, new memories. Mol Neurobiol 2004; 29: 117-130. Levy LM, Hallett M. Impaired brain GABA in focal dystonia. Ann Neurol 2002;51:93101.
198
Eman M. Khedr
Lomarev MP, Kanchana S, Bara-Jimenez W, Iyer M, Wassermann EM, Hallett M: Placebocontrolled study of rTMS for the treatment of Parkinson's disease. Mov Disord 2006;21:325-331. Magyar-Lehmann S, Antonini A, Roelcke U, et al. Cerebral glucose metabolism in patients with spasmodic torticollis. Mov Disord 1997; 12: 7048. Mally J, Farkas R, Tothfalusi L, Stone TW. Long-term follow-up study with repetitive transcranial magnetic stimulation (rTMS) in Parkinsons disease. Brain Res Bull 2004;64:25963. Mally J, Stone TW. Improvement in Parkinsonian symptoms after repetitive transcranial magnetic stimulation. J Neurol Sci 1999;162:17984. Mally J, Stone TW. Therapeutic and dose-dependent effect of repetitive microelectroshock induced by transcranial magnetic stimulation in Parkinsons disease. J Neurosci Res 1999a; 57:93540. Matsuzaki M, Honkura N, Ellis-Davies GCR, Kasai H. Structural basis of long-term potentiation in single dendritic spines. Nature 2004; 429;761-766. Mavroudakis N, Caroyer JM, Brunko E, Zegers de Beyl D. Abnormal motor evoked responses to transcranial magnetic stimulation in focal dystonia. Neurology 1995;45:16717. Mir P, Matsunaga K, Gilio F, Quinn NP, Siebner HR, Rothwell JC. Dopaminergic drugs restore facilitatory premotor-motor interactions in Parkinson disease. Neurology 2005;64:1906-1912. Mottaghy FM, Krause BJ, Kemna LJ. Modulation of the neural circuity subserving working memory in healthy human subjects by repetitive transcranial magnetic stimulation. Neurosci Lett 2000; 280: 167-170. Murase N, Rothwell JC, Kaji R, et al. Subthreshold low-frequency repetitive transcranial magnetic stimulation over the premotor cortex modulates writers cramp. Brain 2005;128:104115. Neurophysiol 2000;111:800-805. Okabe S, Ugawa Y, Kanazawa I. Effectiveness of rTMS on Parkinsons Disease Study Group. 0.2-Hz repetitive transcranial magnetic stimulation has no add-on effects as compared to a realistic sham stimulation in Parkinsons disease. Mov Disord 2003;18:3828. Otsuka M, Ichiya Y, Shima F, et al.. Increased striatal 18F-dopa uptake and normal glucose metabolism in idiopathic dystonia syndrome. J Neurol Sci 1992; 111: 1959. Pascual-Leone A, Valls-Sole J, Brasil-Neto JP, Cammarota A, Grafman J, Hallett M. Akinesia in Parkinson's disease. II. Effects of subthreshold repetitive transcranial motor cortex stimulation. Neurology 1994;44:892-898. Paus T, Castro-Alamancos MA, Pertrides M. Cortico-cortical connectivity of the human mid dorsolateral frontal cortex and its modulation by repetitive transcranial magnetic stimulation. Eur J Neurosci 2001; 14: 1405-1411. Picard N, Strick PL. Motor areas of the medial wall: a review of their location and functional activation. Cereb Cortex 1996;6:34253 Playford ED, Jenkins IH, Passingham RE, Nutt J, Frackowiak RS, Brooks DJ. Impaired mesial frontal and putamen activation in Parkinsons disease: a positron emission tomography study. Ann Neurol 1992;32:15161.
199
Pogarell O, Koch W, Popperl G, et al. Striatal dopamine release after prefrontal repetitive transcranial magnetic stimulation in major depression depression: preliminary results of a dynamic [123I] IBZM SPECT study. J Psychiatr Res 2006;40: 307314. Priori A, Berardelli A, Inghilleri M, Accornero N, Manfredi M. Motor cortical inhibition and the dopaminergic system. Pharmacological changes in the silent period after transcranial brain stimulation in normal subjects, patients with Parkinson's disease and drug induced Parkinsonism. Brain 1994; 117: 317-323. Pulvermuller F, Lutzenberger W, Muller V, Mohr B, Dichgans J, Birbaumer N. P3 and contingent negative variation in Parkinsons disease. Electroencephalogr Clin Neurophysiol 1996;98: 45667. Quartarone A, Girlanda P, Risitano G, et al. Focal hand dystonia in a patient with thoracic outlet syndrome. J Neurol Neurosurg Psychiatry. 1998;65:272-4. Quartarone A, Bagnato S, Rizzo V, et al. Abnormal associative plasticity of the human motor cortex in writers cramp. Brain 2003;126:2586 2596. Quartarone A, Rizzo V, Bagnato S, et al. Homeostatic-like plasticity of the primary motor hand area is impaired in focal hand dystonia. Brain 2005;128:19431950. Rascol O, Sabatini U, Brefel C, et al. Cortical motor overactivation in parkinsonian patients with L-dopa-induced peak-dose dyskinesia. Brain 1998;121: 52733. Rascol O, Sabatini U, Chollet F, et al. Supplementary and primary sensory motor area activity in Parkinsons disease. Regional cerebral blood flow changes during finger movements and effects of apomorphine. Arch Neurol 1992;49:1448. Rascol O, Sabatini U, Chollet F, et al. Normal activation of the supplementary motor area in patients with Parkinsons disease undergoing longterm treatment with levodopa. J Neurol Neurosurg Psychiatry 1994;57:56771. Rektorova I, Sedlackova S, Telecka S, Hlubocky A, Rektor I. Dorsolateral prefrontal cortex: a possible target for modulating dyskinesias in Parkinsons disease by repetitive transcranial magnetic stimulation. Int J Biomed Imaging 2008; 3: 721-725 Ridding MC, Sheean G, Rothwell JC, Inzelberg R, Kujirai T. Changes in the balance between motor cortical excitation and inhibition in focal, task specific dystonia. J Neurol Neurosurg Psychiatry 1995;59:493 498. Ridding MC, Taylor JL, Rothwell JC. The effect of voluntary contraction on cortico-cortical inhibition in human motor cortex. J Physiol (Lond) 1995; 487: 5418. Ridding, MC, Inzelberg R and Rothwell, JC. Changes in excitability of motor cortical circuitry in patients with Parkinsons disease. Ann. Neurol. 1995; 37: 181-188. Rizzo V, Siebner HR, Modugno N, et al. Shaping the excitability of human motor cortex with premotor rTMS. J Physiol 2004;554: 483495. Rosenkranz K, Altenmller E, Siggelkow S, Dengler R. Alteration of sensorimotor integration in musicians cramp: impaired focusing of proprioception. Clin Neurophysiol 2000;111:2040 2045. Rossini PM, Babiloni F, Bernardi G, et al. Abnormalities of short-latency somatosensory evoked potentials in parkinsonian patients. Electroencephalogr Clin Neurophysiol 1989;74:27789. Sabatini U, Boulanouar K, Fabre N, et al. Cortical motor reorganization in akinetic patients with Parkinsons disease: a functional MRI study. Brain 2000;123:394403.
200
Eman M. Khedr
Shimamoto H, Takasaki K, Shigemori M, Imaizumi T, Ayabe M, Shoji H. Therapeutic effect and mechanism of repetitive transcranial magnetic stimulation in Parkinsons disease. J Neurol 2001;248:4852. Siebner HR, Auer C, Conrad B. Abnormal increase in the corticomotor output to the affected hand during repetitive transcranial magnetic stimulation of the primary motor cortex in patients with writers cramp. Neurosci Lett 1999;262:133136. Siebner HR, Mentschel C, Auer C, Conrad B. Repetitive transcranial magnetic stimulation has a beneficial effect on bradykinesia in Parkinson's disease. NeuroReport 1999;10:589594. Siebner HR, Rossmeier C, Mentschel C, Peinemann A, Conrad B. Short-term motor improvement after sub-threshold 5-Hz repetitive transcranial magnetic stimulation of the primary motor hand area in Parkinson's disease. J Neurol Sci 2000;178:91-94. Siebner HR, Tormos JM, Ceballos-Baumann AO, et al. Lowfrequency repetitive transcranial magnetic stimulation of the motor cortex in writers cramp. Neurology 1999;52:529 537. Siebner H.R., Filipovic S.R., Rowe J.B., et al. Patients with focal arm dystonia have increased sensitivity to slow-frequency repetitive TMS of the dorsal premotor cortex. Brain 2003; 126: 2710-2725 Sommer M, Kamm T, Tergau F, Ulm G, Paulus W. Repetitive paired-pulse transcranial magnetic stimulation affects corticospinal excitability and finger tapping in Parkinsons disease. Clin Neurophysiol 2002;113:944950. Stinear CM, Byblow WD. Impaired modulation of corticospinal excitability following subthreshold rTMS in focal hand dystonia. Hum Mov Sci 2004; 23:527538. Strafella AP, Ko JH, Grant J, Fraraccio M, Monchi O. Corticostriatal functional interactions in Parkinsons disease: a rTMS/[11C]raclopride PET study. Eur J Neurosci 2005;22:2946-2952. Strafella AP, Paus T, Barrett J, Dagher A. Repetitive transcranial magnetic stimulation of the human prefrontal cortex induces dopamine release in the caudate nucleus. J Neurosci 2001;21:RC157. Strafella AP, Paus T, Fraraccio M, Dagher A. Striatal dopamine release induced by repetitive transcranial magnetic stimulation of the human motor cortex. Brain 2003;126:26092615. Tergau F, Wassermann EM, Paulus W, Ziemann U. Lack of clinical improvement in patients with Parkinsons disease after low and high frequency repetitive transcranial magnetic stimulation. Electroencephalogr Clin Neurophysiol 1999; 51: 281-288 Tinazzi M, Farina S, Edwards M, et al. Task-specific impairment of motor cortical excitation and inhibition in patients with writers cramp. Neurosci Lett 2005;378:5558. Tinazzi M, Farina S, Edwards M, et al. Task-specific impairment of motor cortical excitation and inhibition in patients with writers cramp. Neurosci Lett 2005;378:5558. Valls-Sole J, Pascual-Leone A, Brasil-Neto JP, Cammarota A, McShane L, Hallett M. Abnormal facilitation of the response to transcranial magnetic stimulation in patients with Parkinson's disease. Neurology 1994; 44: 735-741. Wagner T, Valero-Cabre A, Pascual-Leone A. Noninvasive human brain stimulation. Annu Rev Biomed Eng 2007;9:527565.
Chapter 13
INVASIVE CORTICAL STIMULATION FOR PARKINSONS DISEASE AND MOVEMENT DISORDERS

B. Cioni, A. R. Bentivoglio, C. De Simone, A. Fasano, C. Piano, D. Policicchio, V. Perotti, M. Meglio
Functional Neurosurgery, Anesthesiology, and Neurology, Universit Cattolica, Roma, Italy.
INTRODUCTION
Ever since the observation that motor cortex (MI) stimulation (MI ECS) could relieve some patients suffering post-stroke movement disorders (Katayama et al 1997, 2002) and, above all, the successful introduction of MI ECS for the treatment of Parkinsons disease by Canavero in 1998 (Canavero and Paolotti 2000), extradural MI ECS has been employed for the treatment of several movement disorders. This chapter reviews this field.
1-PARKINSONS DISEASE (PD) AND PARKINSONISM

A-Review of the Literature A total of approximately 100 patients with Parkinsons disease have been treated by MI ECS. 1- The first report was that by Canavero and Paolotti (2000), who described the case of a 72 year old woman with advanced PD who showed a substantial improvement of all three
Correspondence concerning this article should be addressed to: Dr. Beatrice Cioni, MD, e-mail: bcioni@rm.unicatt.it.
202
B. Cioni, A. R. Bentivoglio, C. De Simone et al.
cardinal parkinsonian symptoms following unilateral extradural MI ECS (electrode strip parallel to Rolandic fissure). In July 1998, a quadripolar electrode was positioned in the extradural space overlying the hand representation area of MI, contralateral to her worst clinical side. Chronic stimulation was delivered subthreshold for any movement or sensation, at 3V, 180 s, 25Hz, 3+/0- setting, off during sleep. The clinical improvement was bilateral; she could stand without assistance, climb the stairs and walk for a short distance. The UPDRS III in On-Med condition decreased from 44 to 23 after 3 months. She had a moderate to severe PD-associated dementia and this aspect also improved. L-dopa was reduced by 80%. Importantly, after 3 months, the patient developed infection and was explanted. However, she lost benefit only gradually over several weeks, a sign of likely neuroplastic changes induced by stimulation. Upon reimplantation, she improved again. These results were confirmed in two further patients operated on in 2001 and 2002 (Canavero et al 2002, 2003). Canavero also reported on a patient suffering multiple system atrophy-associated parkinsonism. Bilateral subthalamic deep brain stimulation was ineffective, but moderately responded to bilateral MI stimulation with the following parameters: 25-40 Hz, 90-180s, 22.5V, bipolar stimulation (3+/0- or viceversa), continuously. Motor symptoms improved for about 9 months, while vegetative symptoms remained improved until death almost 3 years later (Canavero et al 2003). 2- Upon Canaveros urge, Pagni et al (2003, 2005) implanted 5 new cases (in their 2005 publication, patients 5 and 6 were taken from Canaveros series). All of them were submitted to unilateral extradural MI ECS (electrode strip parallel to Rolandic fissure), opposite to the worst clinical side; chronic stimulation was delivered bipolarly at 2.5-6V, 150-180 s, 2540Hz, continuously; follow-up ranged from 4 months to 2.5 years. The patients were evaluated only in the On-Med state: global UPDRS decreased by 42-62%, UPDRS III (motor axis) by 32-83%. It was possible to decrease L-dopa by 11-33% in 3 cases and by 70-73% in 2 patients. The improvement was bilateral. One case had to be reimplanted due to misplacement. Interestingly, neuropsychological assessment (MMSE, 15 Reys Words, Corsi block-tapping test, WAIS with revised digit span subtest and block design subtest, Prose Memory Test, Attention Matrices Test, Verbal Judgement Test, Arithmetic Judgement Test, SPM, Verbal Fluency, Token Test, MADRS, BPRS, QL-Index) of three of these patients at 1 year follow-up revealed stable cognitive functions, stability of mood disorders, disappearance of hallucinations (1 patient) and mildly improved quality of life in 2 of 3 patients (Munno et al 2007). 3- The Torontos group (Kleiner-Fisman et al 2003) reported on the efficacy of highfrequency subdural MI ECS (bipolar, or monopolar in 1 case) in 5 patients with Parkinsonism due to multiple systemic atrophy (MSA). Three patients reported subjective improvement, although UPDRS scores did not change at 3 and 6 months. Thereafter, they (Strafella et al 2007) submitted 5 patients (age range 70-75)with advanced PD to unilateral subdural MI ECS (hand motor area, left in 3, right in 1). All had 30% motor improvement on levodopa and none met DBS criteria. One patient developed a cortical venous infarct and was excluded from study results. Multiple settings (5-185 Hz, 60-450 s, 0.5-7V, monopolar and bipolar stimulation) were assessed immediately after surgery and after 1-2 weeks of stimulation. At 6 month follow-up, patients underwent double-blind evaluations randomized to stimulation on or off (for 2 weeks each), with and without medications (acute levodopa
Invasive Cortical Stimulation for Parkinsons Disease and Movement Disorders
203
challenge); modified CAPIT and neuropsychological testing were performed and a PDQ-39 questionnaire was administered preoperatively and at 6 months. There were no acute improvements in akinesia, rigidity or gait function with stimulation. Tremor improved acutely in the patient with prominent tremor. The off drug UPDRS motor scores off- and onstimulation (43 +/- 7.9 versus 39.5 +/-12.5) were not significantly different, nor the onstim/on-med scores and preoperative on condition (18.7 +/- 8.6 versus 14.2 +/-2.2). Two patients worsened (more off time, less on), one was unchanged, one had some improvement with stimulation. These patients were also assessed with H2O PET at 6 months from surgical implantation following optimization of parameters (3-5V, 60-90 s). Subjects were tested at rest and while performing a simple motor task with a joystick. Scanning was performed at 10 minutes intervals on stimulation off, 50 and 130 Hz stimulation. Joystick movement compared to rest while MI ECS was turned off augmented rCBF in the sensorimotor cortex (BA 4/3), lateral premotor cortex (BA6) and caudal supplementary motor cortex (SMA) contralateral to hand movement. During 50 Hz and 130 Hz stimulation, the same areas were activated when comparing joystick movement versus rest. However, these changes in rCBF were not significantly different when comparing across different stimulation settings (OFF vs 50 Hz vs 130 Hz); no significant movement-related rCBF changes were detected during MI ECS in other cortical areas such as the rostral SMA and dorsolateral prefrontal cortex. Motor performance during the joystick task did not change significantly across different stimulation frequencies, although a trend towards worsening was observed during 50 Hz stimulation. This study is ill-conceived. MI ECS has no acute clinical effects and the 10 minute intervals are simply inadequate. 4- Pagni, upon retiring in 2003, spearheaded a multicenter study to evaluate the efficacy of MI ECS in advanced PD, on behalf of the Italian Society of Neurosurgery. The preliminary results of the multicenter study have been published (Bentivoglio et al 2005). Twenty-nine patients were treated. Initially, patients were submitted to a test period (1-2 weeks) before IPG implantation, but in the following cases the whole implant was made in a single session. All symptoms of PD (tremor, rigidity, motor dexterity, bradykinesia, posture and gait, freezing) were improved and L-dopa could be reduced. Best parameters were 2.5-6V, 150-180 (but also 90-120) s, 25-40 (but also 60-80) Hz. In an intention to treat analysis, after 6 months of MI ECS, the mean UPDRS III decreased by 21% in the offmed condition and by 34% in the on-med condition. At 1 year, the effect of MI ECS was less evident, with a decrease of UPDRS III of 13% and 21% respectively in the off and on-med conditions. Notably, the stimulation parameters differed in the various centers: monopolar or bipolar stimulation, 2-8V, 60-400 s, 20-120Hz, continuously or only during day time. Some patients were unresponsive to MI ECS: many of these patients had MRI findings of leucoencephalopathy, white matter ischemic foci, cerebral atrophy, suggesting a diagnosis of Parkinsonism rather than true PD. The long term results (up to 36 months) have been reported for 41 patients (Pagni et al 2008). UPDRS evaluation was performed before implantation and after 3, 6,12, 24 and 36 months of MI ECS, both in the on-med and off-med conditions. Analysis of variance for repeated measures and Wilcoxon signed-rank paired samples test were used for statistical analysis. At 12, 24 and 36 months of follow up, significant decreases of global UPDRS off-
204
med (by 21.9%, 14.6%, and 26.4%, respectively), of UPDRS III off-med (by 17.8%, 10.5% and 25.5%) of axial symptoms (subscores 27+28+29: by 19.9%, 16.2% and 21.3%), of UPDRS IV (by 37.7%, 18.1%, and 21.6%) and of UPDRS II off-med (by 27.6%, 21.3%, and 31.8% , respectively) were apparent. Drug therapy could be decreased by a mean of 20%. No major adverse event was encountered. Thus, MI ECS moderately improves motor symptoms (UPDRS III), decreases complications of therapy (UPDRS IV) and ameliorates activities of daily living (UPDRS II) in advanced PD. The clinical effect can be expected to last at least 3 years. 5- Benvenuti et al (2006) submitted a 68 year old woman to MI ECS. Motor fluctuations uncontrolled by therapy, dyskinesias, depression and anxiety were severe. During MI ECS, UPDRS scores changed significantly (Activities of Daily Living: from 26 to 17; Motor: from 33 to 22; Complications of Therapy: from 13 to 7; Mentation: from 8 to 6). Dyskinesias and motor fluctuations were reduced and no longer disabling, rigidity was reduced bilaterally; standing, gait and motor performance were also improved. Levodopa was not reduced. Improvement was 35% on the UPDRS scale 8 weeks after surgery. 6- Verhagen et al (2006) conducted a prospective multicenter study of 9 PD patients (H&Y OFF III) receiving extradural MI ECS (hand area) for 24 weeks. Stimulation effects were systematically explored at 20-127 Hz with constant pulse width (250 s). At end of study, the UPDRS III score OFF medication was 42.13 +/- 13.78 compared to 38.78 +/- 8.08 at baseline. Three patients remained on stimulation 1 year later. Their UPDRS III scores at baseline, 24 and 52 weeks were 37, 30, 30; 29,37,20 and 29, 25 and 25. At 24 weeks On Time Without Troublesome Dyskinesias was 11.25 +/-2.10 hours, compared to 9.54 +/- 3.89 at baseline. PDQ-39 Total scores were 38.33+/-10.17 at 24 weeks compared to 43.62+/-8.33 at baseline. During programming, there were 2 fits. This study has limitations. The stimulation hardware was different from that used by all other groups, with a different waveform and current output, but, above all, only two contacts available spaced 31 mm apart corresponding to contacts 0 and 3 of the tetrapolar electrode commonly used for ECS. The 10 Hz band was not explored, and the pulse width was fixed at 250 s. Finally, at least one patient was probably not PD (B Gliner, meeting communication, 2006). 7- Cilia et al (2007) submitted to extradural MI ECS (hand area; electrode strip parallel to Rolandic fissure) 5 patients with PD who fulfilled CAPSIT criteria for DBS, with the exception of age >70 years (range 71-77). In particular, there were no vascular abnormalities or significant brain atrophy on MRI. Patients were assessed preoperatively and after 6 months of MI ECS of the left hemisphere, on and off medication, with stimulator on and 2 weeks later with stimulator off, by a blinded neurologist. Stimulation was monopolar (0-/3-/case +), 3-4V, 60 (40 in 1 patient) Hz, 180-210 s, continuous (night and day). MI ECS determined a reduction of daily OFF time (UPDRS item 39) in 3 patients, a mean L-DOPA reduction of 16% and a mean dopamine agonist reduction dosage of 49%, and a reduction of the Abnormal Involuntary Movements Scale (AIMS) of 19%. The off-med UPDRS II and UPDRS III scores did not vary, even if a trend towards improvement was shown in axial symptoms (items 14-28-29). On-Med UPDRS II and III scores values were unmodified. No behavior or mood changes were reported at 6 months. No immediate motor change was reported when stimulation was switched off, except for 1 patient who showed worsening gait. Fifteen days later, 3 patients complained of prolonged daily OFF time and freezing gait, and
205
2 patients with unchanged medication complained of an increase in dyskinesias as well. Four patients asked spontaneously to have the stimulator turned on again. In sum, 4 of 5 patients reported subjective improvement involving mainly axial symptoms (i.e. stability, posture and gait) as well as a reduction in daily OFF time and dyskinesia. Subjective clinical benefit was reported after a variable interval after switching on the stimulator (minutes to days). This study has a few limitations. Four out of 5 patients were monopolarly stimulated. Monopolar stimulation is not the same as bipolar stimulation. In fact, the best responder was the only one to receive bipolar stimulation (0-1-/2+3+): this was also the only akinetic-rigid dominant patient, whereas the remaining 4 were tremor dominant. These authors started from a frequency of 30 Hz and then slowly increased up to 60 Hz. The 10-20 and 70-80 Hz bands were not explored, and these can be effective. In a follow-up paper (Cilia et al 2008), these same authors reported on the same 5 patients, plus 1 (a 70 year old man), and studied them with Tc-99m cysteinate dimmer bicisate (ECD) SPECT off-medication with stimulator off and on. Patients were assessed preoperatively and 6 months after surgery, with both stimulator on and 2 weeks after switch-off by a blinded neurologist. All medications remained unchanged during the 2-week stim-off period. Parameters were assessed across different configurations (monopolar and bipolar, 0.5-5V, 60-210 s, 20-120 Hz). Blinded evaluation in individual patients revealed clinical improvement of UPDRS II and III scores during stimulation compared to stim-off condition up to 40 and 20% respectively. UPDRS scores did not change significantly during bipolar stimulation (0-/3+) and 120 Hz. Clinical improvement occurred after several days to 4 weeks after stimulation parameters modifications. Objective motor benefit was observed mainly in axial symptoms (gait, stooped posture and postural instability). Subjective benefit was admitted by 4 of 6 patients, mainly in mobility and walking, associated with an increase in daily on-time (as if stimulation would delay wearing-off phenomenon). In two patients, battery exhausted about 12 months after implant: both experienced slow and progressive clinical deterioration, associated with an objective worsening of UPDRS III scores of 20% and 25% respectively, that reversed after battery replacement. ECD SPECT with a triple head camera and ultra high resolution collimators was conducted with stimulator on and 2 weeks after switching the stimulator off. At a statistical threshold of p<0.001, the authors found rCBF decrements in MI (BA4) and lateral premotor cortex (BA6) bilaterally (!), left dorsolateral prefrontal cortex (BA9) and right caudate head. Perfusion decrements occurred in BA19. At p<0.01, additional clusters were seen in right thalamus (stimulator was on left MI!), left caudate nucleus and right BA6. Significant perfusion increases were found in the right cerebellar lobe, and, with a less strict threshold, left BA18, left cerebellar lobe and vermis. Thus, effects were widespread and bilateral, paralleling clinical benefits. 8- Arle et al (2008) applied extradural MI ECS (electrode strip parallel to Rolandic fissure ) in 4 patients (bilateral in 3) aged 44-72. Three patients were eligible for DBS. All patients had at least 30% improvement in their UPDRS scores when taking L-Dopa. Programming was started within 24 hours of electrode implantation. Initially all contacts were checked in a monopolar setting at 130 Hz and 210 s, with voltage slowly increased to 4V. Contacts 1 and 2 were left ON at 3V: if not tolerated, contact 3 or 0 alone was used. In all cases 2 contacts could be activated. Cathodal (not anodal) stimulation was used. If minimal benefit was noted during later sessions, more contacts were added to the monopolar
206
configuration. In case of failure, contact 0 or 3 was placed in cathodal mode and the other 3 contacts set as the anode. Other investigated changes included PW 410 s, 4V, 80Hz and 185Hz. In the end, patient 1 was stimulated at 3.2 mA, 240 s, 130 Hz, 0-/1-/2-/3off/case +; patient 2 at 3.5mA,210 s, 130 Hz, left sided 0-/1-/2+/3+/case off and right sided 0off/1-/2/3+/case off; patient 3 at 3.4 mA (left)/3.2 Amp (right), 210 s, 100 Hz, left sided 0-/1-/2/3off/case + and right sided 0-/1-/2-/3off/case +. Patient 4 was not detailed. One patient developed an infection and was explanted at 3 months: he had an improvement of about 50% over his baseline condition in dyskinesias, rigidity and tremor following unilateral implantation, but it significantly regressed upon device explantation. After reimplantation several months later, he achieved the same benefit. Two patients had a 20-30% reduction in medication requirements, but one needed 37% more. One patient had a return of dyskinesias after 3 months due to reduction of amplitude of stimulation following initial benefit: upon reincreasing the amplitude, dyskinesias were again controlled. Globally, there was a trend toward a significant difference in the UPDRS III scores between the baseline score and the scores at 1,3 and 6 months, but not at 12 months, possibly because of fewer data point observations available for analysis. At 1 month, there was a mean decrease in the UPDRS III score of 21(+/-13) points from baseline, at 3 months of 15(+/-11) points, at 6 months of 18(+/-16) points, but only of 9(+/-9) points at 1 year for the 2 patients with sufficient followup. More in detail, the UPDRS global scores during on-med/on-stim preoperatively, at 1,3,6 and 12 months in 3 patients (4th not available) were respectively: 46,52,47,51(12th month milestone not yet reached); 57,79,42,94,62; 53.5,23,32,54,37; during off-med and on-stim: 27,17,22,15 (12th month milestone not yet reached); 46,11,31,31,43; 41.5,12,10,5,25; 18,10,9,14 (12th month milestone not yet reached). The UPDRS III (motor) scores in onmed/on-stim in the same 3 patients were respectively: 19,26,25,23 (12th month milestone not yet reached); 32, 47,13,18,35; 24.5,12,18,27,5. 9- Seijo et al (2009) submitted 6 patients not eligible for DBS to extradural unilateral MI ECS (contralateral to side of clinical onset). Five of six were akinetic-rigid, 1 had left hand tremor. Parameters adjustments were carried out during 1 week after initial implant. Optimal parameters were: case 1:3+/0-, 4.5V,450 s, 30Hz; case 2:2+/1-,3V, 400 s, 10 Hz; case 3: 3+/0-, 3.7V, 450 s, 30 Hz; case 4: 3+/0-, 3.6V, 450 s, 10 Hz; case 5: 3+/0-,3V, 330 s, 25 Hz; case 6: 1-/2+, 3.6V,450 s, 25 Hz. Follow-up was 1 year. In case 1, UPDRS III onmed/on-stim was 26 (but 42 off-med/on-stim), Hoehn&Yahr (H&Y) score was 2 and Schwab &England Daily Living Activities Scale (DLA) 90%. In case 2, they were 27 (vs 38), 3 and 80%: dyskinesias and freezing while walking were significantly decreased, with freezing restricted only to off-med periods. In case 3, they were 17 (vs 55), 2, 70%; levodopa was decreased. In case 4, they were: 27 (vs 38), 2 and 90%: dyskinesias were reduced, though tremor was not. In case 5, they were 29 (vs 42), 2 and 80%: levodopa was reduced. In case 6, they were 26 (vs 42), 2 and 90%: levodopa was decreased. Globally, UPDRS III score improved in 2 patients (cases 1 and 4) comparing preoperative condition to on-med/on-stim conditions, but did not change in cases 2,3 and 5 or slightly worsened in case 6.
Invasive Cortical Stimulation for Parkinsons Disease and Movement Disorders B-Personal Experience
207
In 2003, we started a prospective study to evaluate the efficacy of MI ECS in patients with advanced PD (Cioni et al 2007, Cioni 2007). The inclusion criteria were: idiopathic Parkinson disease (PDSBB criteria); at least 5 years diseases length; disease in the advanced state (UPDRS in off >/= 40/180; Hoehn and Yahrs >/= 3; motor complications: fluctuations and disabling dyskinesias); positive response to L-Dopa; DBS not accepted by the patient or contraindicated; patients ability to give informed consent to the study. The exclusion criteria were: history of epilepsy or EEG epileptic activity; alcohol or drug abuse; mental deterioration; psychiatric symptoms; previous basal ganglia surgery; other major illness. Ten patients met the above mentioned criteria and were submitted to the implant of an epidural plate electrode over the motor cortex contralateral to the worst clinical side in 3 cases, and to a bilateral implant in the remaining 7 cases (Figures 1-2). Therapeutic stimulation during the first year was through the electrode contralateral to the worst clinical side; parameters were: 120s, 80Hz, 3-6V (subthreshold for movements, and motor or sensory phenomena), delivered continuously through contacts 0 (anode) and 3 (cathode). After 12 months, in cases of bilateral implantation, stimulation was made bilateral (same parameters for the side ipsilateral to the worst clinical side). The clinical assessment before implant and at 1, 3, 6, 12, 18, 24, and 36 months included: UPDRS (Unified Parkinson Disease Rating Scale); finger tapping; walking time; PDQL (Parkinson Disease Quality of Life Scale); neuropsychological evaluation including MMSE (Mini Mental State Evaluation), behavioral assessment of mood and anxiety, tests for verbal short term memory, spatial short term memory, episodic verbal memory, non-verbal abstract reasoning, frontal executive functions and verbal fluency; EEG; oral medications and adverse events. The clinical motor evaluation was performed both in the off and in the on medication state and the motor assessment was videotaped. The OFF condition was achieved by withdrawing antiparkinsonian medications as follows: Levodopa for at least 12 hours, pergolide, pramipexole, ropinirole for at least 48 hours, cabergoline for at least 168 hours, apomorphine for at least 3 hours. The ON condition was achieved 60 minutes after administering a supra-threshold dose of standard levodopa, according to daily schedule. Cognitive and behavioral assessment were performed preoperatively and at 6, 12 and 18 months, in the on med status. A statistically (Wilcoxons test and analysis of variance for repeated measures) significant improvement was present after 12 months of unilateral MI ECS, as of total UPDRS off-med, UPDRS II, UPDRS III off-med, subscore for axial symptoms (UPDRS III: items 27-31), UPDRS IV, PDQL. The effect of unilateral MI ECS was bilateral, with no significant difference between the two sides. It was evident after 1-2 weeks of stimulation, and in a case of accidental switching off of the stimulator the patient became aware of something going wrong after 2-3 weeks. After 1 year of unilateral stimulation, 7 patients needed bilateral MI ECS, the remaining 3 patients continued with unilateral stimulation. The statistical analysis at 24 and 36 months demonstrated a significant improvement in total UPDRS off-med, in UPDRS III off-med (by 20% compared to the preoperative score), in the subscore for axial symptoms, in UPDRS IV and in PDQL. Notably, in all the patients, the UPDRS III off med at 24 and 36 months was always lower than UPDRS III off med at preoperative evaluation. At 2 years, bilateral stimulation showed a trend towards greater
208
benefit compared to unilateral stimulation, but a statistical analysis was not possible due to the small sample. Drug treatment could be decreased by 25%. No complication occurred; no adverse events, particularly no epileptic seizures nor EEG epileptic activity, were encountered. Cognitive assessment in the overall group of patients showed a significant postoperative improvement on the MMSE and on tasks of episodic verbal memory, but this most likely reflects a practice effect (Daniele et al 2003). No significant postoperative decline was observed on any cognitive task, including those of phonological and semantic verbal fluency, whereas DBS patients show a significant decline on verbal fluency (Contarino et al 2007). Unilateral stimulation of the left hemisphere showed a statistical trend towards a postoperative improvement of phonological verbal fluency, along with an increase of depressive symptoms. On the contrary, stimulation of the right hemisphere showed a trend towards a decrease of depression, and no effect on verbal fluency. We also assessed our patients with DAT scan (SPECT) in order to evaluate binding to the dopamine transporters before and after 6, 12 and 28 months of MI ECS, and IBZM-SPECT to evaluate activity of striatal postsynaptic D2 receptors, before and after 6 and 12 months of MI ECS. DAT-scans at 6 and 12 months showed an increase in the activity in both putamina, particularly on the side ipsilateral to the stimulation. Postsynaptic receptors activity was unmodified at 6 and 12 months.
Figure 1. Top right: Central sulcus (CS) identification. Cortical median nerve SEPs (recording: E0, E1, E2, E3, referred to Fz) showing phase reversal between traces 1 and 2 (E0 and E1). Bottom traces: Motor cortex mapping, Recording from Biceps Brachii, Abductor Pollicis Brevis, Quadriceps after stimulation delivered through electrode E1, E2 and E3 (cathode) and 6 cm in front of Cz (anode). Top left: integrated anatomo-functional position of the quadripolar electrode.
209
Furthermore, we reported (Fasano et al 2008) the case of a 72-year-old female patient affected by severe PD who underwent bilateral MI ECS. In baseline med-off condition the patient was unable to rise from a chair and to stand without assistance. Stimulation at 3 and 60 Hz failed to provide any improvement, whereas, when stimulating at 130 Hz, axial akinesia and walking improved consistently: the patient, in med-off condition, was able to rise from chair and to walk without assistance. The patient underwent two brain 99mTcEthylcysteinate Dimer-SPECT studies, which showed that the regional cerebral perfusion was significantly increased in the supplementary motor area during stimulation at 130 Hz. After five months, the benefit of MI ECS gradually disappeared. We are now performing a controlled randomized study on the clinical usefulness of MI ECS in PD. The protocol has been published on www.ClinicalTrials.gov (Identifier # NCT00637260). Another controlled study of 10 patients is on file (Identifier # NCT00159172).
Figure 2. Skull X-rays and CT scan showing the position of the bilateral quadripolar strip electrode.
2-PURE AKINESIA
Tani et al (2007) presented a case of Levodopa-resistant akinesia, a DBS resistant motor disorder. After a positive trial of 10 Hz MI rTMS (500 pulses) (but negative trial of SMA rTMS), the patient underwent a bilateral implant of subdural electrodes over the motor cortex, close to the sagittal sinus. The best stimulation parameters were: 100Hz, 210s, 1.8V, 1+/2- electrode setting, 30 on and 120 off (benefit lasted 3 hours). UPDRS decreased form 70 to 41 after 1 year and to 46 after 2 years; walking time (7 m) improved from 84.7 sec to 20sec at 1 year and 15 at 2 years, and so did step size; dysphagia also improved. Blinded switch-off resulted in deterioration within 48 hours. On the fourth day, the stimulator was reactivated and benefit was restored. A PET study showed a significant increase of rCBF in the left SMA (BA6) and right dorso-lateral prefrontal cortex (BA9) after 30 of bilateral MI ECS. Activation of SMA, involved in movement preparation and initiation, was one of the mechanisms hypothesized at the basis of MI ECS effects.
210
In our series of PD patients the best results were obtained on axial symptoms in patients with severe and unpredictable off periods. We submitted a patient with pure akinesia to bilateral MI ECS. There was a marked improvement of akinesia, posture, and gait; furthermore the patient showed a striking amelioration of verbal fluency.
3-ESSENTIAL TREMOR
Lyons et al (2006) submitted to MI ECS (only two contacts available) two patients suffering essential tremor. The two burr-hole technique developed by Canavero was used (see chapter 2). They were evaluated with the Fahn-Tolosa-Marin Tremor Rating Scale at baseline and after 1 and 4 weeks of contralateral (to tremor) MI ECS. Patient 1 (75 year old man) received MI ECS at 30Hz, 3mA, 250 s. His baseline score was 61, and 1 month after surgery it was 57. Patient 2 (60 year old man)s stimulation parameters were 50Hz, 5 mA, 250 s. His score was 47 before surgery and 43 1 month after MI ECS. In sum, MI ECS proved ineffective. Arle and Shils (2008) reported 2 other patients with essential tremor treated by MI ECS with poor results.
4-DYSTONIA
Canavero et al (2003) described the acute effects of MI ECS in a case of cervical dystonia previously treated with thalamotomy and, after recurrence, cervical rhizotomy, intrathecal baclofen and botulinum toxin. She developed painful dystonic paroxysms involving her right hemibody every 4 hours lasting about 30 min. MI ECS (MI thigh, chest and arm area) at 10 Hz, 450 s and 1V controlled both pain and dystonia, except for the neck symptoms. Franzini et al (2003) reported on the efficacy of MI ECS (130 Hz) for central painassociated thalamic hand in 2 of 3 patients. Two patients suffering intentional myoclonus were also relieved. Romito et al (2007) and Albanese et al (2007) described a case of fixed dystonia unresponsive to pallidal stimulation, but improved by MI ECS. A right handed woman progressively developed a severe segmental dystonia with fixed elevation and anterorotation of the left shoulder, abduction of the upper limb, severe trunk involvement and fixed kyphoscoliosis. Gene mutations were absent. There was no evidence of a complex regional pain syndrome, no preceding trauma, and a complete psychological and psychiatric evaluation ruled out the diagnosis of psychogenic dystonia. The diagnosis was a distinct subtype of primary dystonia primary fixed dystona. Pharmacological treatments failed as well as botulinum toxin. GPi DBS also failed. Stimulation of the right motor cortex (strip parallel to MI) started the day after implant (3.8V, 60 Hz, 60 s, 0-/1+/2- setting) was followed by a gradual improvement and the dystonic postures and pain almost resolved after 6 months. The effect was still present at 22 months. A PET study was performed before and after 6 months of MI ECS showing significant hypometabolism in the cerebellum, more
211
pronounced on the right, and no increase in cortical metabolism. Interestingly, PET during Gpi DBS showed significant CBF increases in SI/MI (more on the left) and SMA and ACC bilaterally.
5-POST-STROKE MOVEMENT DISORDERS

Tsubokawa et al (1993) and Katayama et al (1997, 2002) made the original observation that MI ECS can relieve post-stroke movement disorders. Katayama et al (2002) analyzed the effects of MI ECS on post-stroke movement disorders in 50 patients: the purpose of MI ECS was to control pain in 42 cases and involuntary movements in 8 cases. Good control of involuntary movements was observed in 3 out of 3 patients with hemichoreo-athetosis, in 2 out of 2 cases with distal resting or action tremor and in 1 out of 3 patients with proximal postural tremor. High frequencies (50-125 Hz) seem to be necessary for these effects. Subjective improvement in motor performance was reported by 8 patients who had mild motor weakness; severe motor weakness did not respond. The same group (Katayama et al 2006) developed an on-demand type stimulation system which triggers MI ECS or DBS by detecting intrinsic signals of intention to move. They applied this kind of feed-forward control of involuntary movement in 6 cases in whom motor dysfunction was evident only when patients intended to move their body. On-demand stimulation provided satisfactory feed-forward control in 4 patients with postural tremor and 2 with motor weakness, when the activity of muscles involved in posturing or intention to move was fed into the system. Nguyen et al (1998) reported a case of severe upper limb action tremor and facial pain following removal of an acoustic schwannoma completely controlled by bipolar MI ECS (1.8V, 60 s, 50 Hz,cyclical mode) for 32 months. Arle and Shils (2008) reported the effect of MI ECS in 4 cases of poststroke movement disorders associated with pain. In one case there was no benefit, while the others received varying degrees of benefit. Our group submitted to MI ECS a patient with hemichorea who had a very good response to rTMS (Di Lazzaro et al 2006). MI ECS proved inferior to rTMS.
6-AMYOTHOPHIC LATERAL SCLEROSIS

Sidoti and Agrillo (2006) published the results obtained by bilateral subdural MI ECS in 4 cases of amyotrophic lateral sclerosis (ALS). The motor hand area was targeted. The stimulating plate was parallel to the central sulcus. Stimulation parameters were 30 Hz, 60 s, 4.5V, contacts 0 and 4 positive, contacts 3 and 7 negative and the rest off. Clinical evaluation included neurological examination and ALS Functional Rating Scale (ALSFRS). At two year follow up, 2 patients showed an arrest of the natural course of the illness during the first year and a slow worsening in the second year after surgery. In these 2 cases postoperative SPECT disclosed consistent complete recovery of flow disturbances. MI ECS was ineffective in the other two patients. A neuroprotective effect of MI ECS was speculated at least in a subgroup of ASL patients.
212
Di Lazzaro et al (2006) investigated whether continuous theta burst TMS could have a beneficial effect on ALS. They performed a double blind placebo-controlled trial in 20 ALS patients. The patients submitted to active stimulation received TMS for five consecutive days every month for 6 consecutive months. The primary outcome was the rate of decline evaluated with the ALSFRS. The active group of patients showed a modest, but significant slowing of the deterioration rate. On this basis we applied bilateral MI ECS in one patient with ALS. MI ECS was delivered at 3 and at 30 Hz. The rate of disease progression was not significantly affected.
CONCLUSIONS
In sum, extradural MI ECS may relieve all three main symptoms of PD (akinesia, rigidity, tremor) on both sides simultaneously, but results vary widely, ranging from positive to negative to mixed. Benefit on the various symptoms can vary too: some are improved, others are not, in the same patient. Patients population was not homogeneous in these studies, nor were parameters of stimulation or techniques of implantation (monopolar versus bipolar; low versus high frequency, extradural versus subdural, unilateral versus bilateral, continuous versus daily only, general versus local anaesthesia, burr hole versus craniotomy). Nonetheless, our personal data and published studies suggest that MI ECS can relieve all three major symptoms of PD (akinesia, rigidity, tremor) on both sides simultaneously, as well as other movement disorders. Axial symptoms, gait, akinesia and freezing are particularly improved. The clinical effect of MI ECS cannot be compared with that of STN DBS due to the different inclusion criteria. DBS is usually contraindicated in patients submitted to MI ECS, because of age limits or the presence of MRI anatomical abnormalities (cerebral atrophy, white matter ischemic foci,). However, STN DBS appears to be more effective on motor symptoms, whereas MI ECS seems to be more effective on axial symptoms (gait, posture). Most dramatically, dyskinesias and painful dystonias are reduced (up to 90%) in almost all patients, well before reducing L-DOPA (a likely direct effect of stimulation). Clinical fluctuations too may be reduced. The complication and adverse events rates are lower for MI ECS, with no risk of intracerebral hemorrhage or infection (extradural approach only); verbal fluency is not impaired by MI ECS, and may even be improved. Hallucinations can occur sporadically and resolve with change of drug therapy and/or stimulation parameters. Our DAT-Scan data also suggest that MI ECS may have a protective effect on putaminal degeneration, at least in the short term. The clinical effects of MI ECS seem to decrease with time. This may be due to a placebo effect, and/or to the progressive nature of the disease; or it may reflect a true loss of effectiveness of the stimulation. Changing the parameters of stimulation may reverse this effect: for instance, changing the frequency or the intensity of stimulation according to the impedances (some fibrosis may develop between the electrode surface and the dura) may be useful (neurostimulators delivering impulses in current do not appear superior in this regard). With frequency up to 130Hz, fibers are more likely to be depolarized and excited (Fasano et al 2008). However, if the stimulation is maintained for long periods of time, some synapses
213
may not follow the stimulus train and be blocked, with consequent inhibition. Specific frequencies may be necessary to impose specific patterns of activity, or to suppress abnormal rhythms, or time-protracted processes. Since stimulation is usually delivered continuously, the slight decline in the clinical benefit may be due to cortical habituation. If this is true, alternate stimulation (right side or left side) may be a solution. The clinical effect is long lasting, therefore a cyclic stimulation (only during daytime or even less) is feasible. Not all the patients respond to MI ECS (7% non-responders in the Italian studies): this may be due to the rather large inclusion criteria, or to the different electrode position and different stimulation parameters. Most patients submitted to MI ECS up to now are older than 70, and cortical atrophy or micro-ischemic lesions may disrupt MI locally in some way in aged people, impairing MI ECS modulation in some cases. Younger patients must be included in future series. The effect of rTMS may be predictive of the clinical outcome following MI ECS. However, the frequency and the duration of rTMS certainly differ from those used for MI ECS. As regards the electrode position and stimulation parameters, the number of patients treated with MI ECS is still to low to allow statistical correlation analysis in order to identify prognostic factors. The need for a bilateral implant has still to be demonstrated. Unilateral MI ECS improves motor performances bilaterally, but bilateral stimulation seems to add to such an improvement. Finally, there is no place for subdural MI ECS: this can be burdened with fatal brain hemorrhage. If atrophy is found on imaging, the two burr hole technique introduced by Canavero is indicated to close the durocortical gap, even by adding a further silicone layer on the stimulating strip (Canavero S, personal communication). In sum, MI ECS may be indicated in PD patients with prominent axial symptoms, gait disturbances and therapy complications. It is easy and safe and does away with the cumbersome and time-consuming procedure of DBS. A head-to-head controlled study with DBS is the only way to establish efficacy and safety of both MI ECS and DBS.
REFERENCES
Albanese A, Romito LM, Piacentini S, Perani D, Carella F, Broggi G. Reply. Neurology 2007; 69: 1063 Arle JE, Apetauerova D, Zani J, et al. Motor cortex stimulation for Parkinson disease: 12 months follow up in 4 patients. J Neurosurg 2008; 109: 133-139 Arle JE, Shils J. Motor cortex stimulation for pain and movement disorders. Neurotherapeutics 2008; 5:37-49 Bentivoglio AR, Cavallo MA, Cioni B, et al. Motor cortex stimulation for movement disorders In: Meglio M (Ed) Proceedings of the 14th Meeting of the Worlds Society for Stereotactic and Functional Neurosurgery. Bologna: Medimond, 2005, pp 89-97
214
Benvenuti E, Cecchi F, Colombini A, Gori G. Extradural motor cortex stimulation as a method to treat advanced Parkinsons disease:new perspectives in geriatric medicine. Aging Clin Exp Res 2006; 18:347-348 Canavero S, Paolotti R. Extradural motor cortex stimulation for advanced Parkinsons disease. Mov Disord 2000; 15:169-171 Canavero S, Paolotti R, Bonicalzi V, et al. Extradural motor cortex stimulation for advanced Parkinsons disease: report of two cases. J Neurosurg 2002; 97:1208-1211 Canavero S, Bonicalzi V, Paolotti R, et al. Therapeutic extradural cortical stimulation for movement disorders: a review. Neurol Res 2003; 25: 118-122 Canavero S, Bonicalzi V. Cortical stimulation for parkinsonism. Arch Neurol 2004; 61: 606 Canavero S, Bonicalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. II Neural Networks Surgery. Wien: Springer-Verlag, 2007, 223-232 Cilia R, Landi A, Vergari F, Sganzerla E, Pezzotti G, Antonini A. Extradural motor cortex stimulation in Parkinsons disease. Mov Disord 2007; 22:111-114 Cilia R, Marotta G, Landi A, et al. Cerebral activity modulation by extradural motor cortex stimulation in Parkinsons disease: a perfusion SPECT study. Eur J Neurol 2008; 15: 2228 Cioni B. Motor cortex stimulation for Parkinsons disease. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. II Neural Networks Surgery. Wien: SpringerVerlag, 2007, 233-238 Cioni B, Meglio M, Perotti V, DeBonis P, Montano N. Neurophysiological aspects of chronic motor cortex stimulation. Clin Neurophysiol 2007; 37:441-447 Contarino MF, Daniele A, Sibilia AH, et al. Cognitive outcome 5 years after bilateral chronic stimulation of the subthalamic nucleus in patients with Parkinsons disease. J Neurol Neurosurg Psychiatry 2007; 78:248-252 Daniele A, Albanese A, Contarino MF, et al. Cognitive and behavioural effects of chronic stimulation of the subthalamic nucleus in patients with Parkinsons disease J Neurol Neurosurg Psychiatry 2003; 74:175-182 Fasano A, Piano C, De Simone C, Cioni B, Meglio M, Bentivoglio AR. High frequency extradural motor cortex stimulation dramatically improves axial symptoms in a patient with Parkinsons disease. Mov Disord 2008; 23:1916-9. Franzini A, Ferroli P, Dones I, Marras C, Broggi G. Chronic motor cortex stimulation for movement disorders: a promising perspective Neurol Res 2003; 25:123-126 Katayama Y, Fukaya C, Yamamoto T. Control of poststroke voluntary and involuntary movement disorders with deep brain or epidural cortical stimulation Stereotact Funct Neurosurg 1997; 69:73-79 Katayama Y, Oshima H, Fukaya C, Kawamata T, Yamamoto T. Control of post-stroke movement disorders using chronic motor cortex stimulation. Acta Neurochir 2002; 79(Suppl):89-92 Katayama Y, Kano T, Kobayashi K, Oshima H, Fukaia C, Yamamoto T. Feed-forward control of post-stroke movement disorders by on-demand type stimulation of the thalamus and motor cortex. Acta Neurochir Suppl 2006; 99:21-23
215
Kleiner-Fisman G, Fisman DN, Kahn FI, Sime E, Lozano A, Lang AE. Motor cortical stimulation for Parkinsonism in Multiple Systemic Atrophy. Arch Neurol 2003; 60:15541558 Lyons KE, Wilkinson SB, Pahwa R. Stimulation of the motor cortex for disabling essential tremor Clin Neurol Neurosurg 2006; 108:564-567 Munno D, Caporale S, Zullo G, et al. Neuropsychological assessment of patients with advanced Parkinson Disease submitted to extradural motor cortex stimulation. Cog Behav Neurol 2007; 20: 1-6 Nguyen JP, Pollin B, Fve A, Geny C, Cesaro P. Improvement of action tremor by chronic cortical stimulation. Mov Disord 1998; 13:84-88 Pagni CA, Zeme S, Zenga F. Further experience with extradural motor cortex stimulation for treatment of advanced Parkinsons diseases. Report of 3 new cases J Neurosurg Sci 2003; 47:189-193 Pagni CA, Zeme S, Zenga F, Maina R.Extradural motor cortex stimulation in advanced Parkinsons disease: the Turin experience. Neurosurg 2005; 57 (ONS Suppl 3): ONS-402 Pagni CA, Albanese A, Bentivoglio A, Broggi G, Canavero S, Cioni B, et al. Results by motor cortex stimulation in treatment of focal dystonia, Parkinson's disease and post-ictal spasticity. The experience of the Italian Study Group of the Italian Neurosurgical Society. Acta Neurochir Suppl. 2008;101:13-21. Romito LM, Franzini A, Perani D, et al. Fixed dystonia unresponsive to pallidal stimulation improved by motor cortex stimulation. Neurology 2007; 68:875-876 Strafella AP, Lozano AM, Lang AE, Ko J-H, Poon Y-Y, Moro E. Subdural motor cortex stimulation in Parkinsons disease does not modify movement-related rCBF pattern. Mov Disorders 2007; 22: 2113-2116 Seijo FJ, Gutierrez JC, Alvarez Vega M, Fernandez Gonzalez F, Lozano Aragonese B, Blazquez M. Therapeutic extradural cortical stimulation for Parkinsons disease: report of six cases and review of the literature. (submitted) Sidoti C, Agrillo U. Chronic cortical stimulation for amyotrophic lateral sclerosis: a report of four consecutive operated cases after a 2-year follow up Neurosurg 2006; 58:E384 Tani N, Saitoh Y, Kishima H, et al. Motor cortex stimulation for levodopa-resistant akinesia: case report. Mov Disorder 2007; 22:1645-1649 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S. Chronic motor cortex stimulation in patients with thalamic pain. J Neurosurg 1993; 78:393-401 Verhagen Metman L, Pahwa R, Lyons K, et al. Motor cortex stimulation in Parkinsons disease: a pilot study. Neurology 2006 (suppl 2) 66: A47
Chapter 13B
INVASIVE CORTICAL STIMULATION FOR PARKINSONS DISEASE (PD): WHY, WHERE AND HOW
Sergio Canavero
Turin Advanced Neuromodulation Group, Turin, Italy.
INTRODUCTION
James Parkinson described transient tremor abolition after a cortical stroke in 1824, an observation replicated by Patrick and Levy in 1922 (see in Speelman and Bosch 1998). In the 1930-1940s, Paul Bucy relieved extrapyramidal symptoms such as tremor by surgical ablation of cortical areas BA4 (MI) and BA6 (premotor cortex, PMC); this was done at the expense of motor deficits (Bucy 1945). Other groups relieved Parkinsonian tremor by pyramidotomy (see Canavero and Bonicalzi 2007 and Priori and Lefaucheur 2007). In the early 1970s, Alberts reported that stimulation at 60 Hz with a 7contact Delgado cortical plate electrode of an area near the rolandic fissure between motor and sensory sites (SI) could initiate or augment parkinsonian tremor in patients. Postcentral cortical stimuli had the same effect at, above or below the sensory threshold. In that same decade, Woolsey et al (1979) temporarily alleviated parkinsonian rigidity and tremor in two patients by direct acute intraoperative stimulation in MI. They wrote: marked tremor and strong rigidityThe results suggest the possibility that subthreshold electrical stimulation through implanted electrodes might be used to control these symptoms in parkinsonian patients. One could imagine that functional neurosurgeons would have caught on to Woolseys suggestion, since the electrodes for stimulation were already available. This, however, did not happen. Neurosurgeons like to operate more than read journals - or at least a good share of them. It was only during the literature search for my book Central Pain Syndrome that I chanced upon Woolseys paper. Extradural motor cortex
218
Sergio Canavero
stimulation had already been introduced by Tsubokawa at the end of the 1980s and I had taken it up in 1993. In 1996, after reading the above paper, I proposed MI ECS for Parkinsons disease to the late Professor Bergamasco of the University of Turin Movement Disorder Group, to no avail. Initial rTMS studies by Pascual-Leone already pointed in that direction. My former chief Pagni, too, did not warm up to the idea. Luck had it that a local neurosurgeon introduced me to a neurologist who accepted to find patients for the trial. The first case was actually his neighbor. I submitted my first paper to the journal Neurosurgery a few months after implant, only to see it rejected. Accidentally, one of the referees reports was not anonymous: prof. Broggi in Milan was the non-anonymous referee and he was very much opposed to MI ECS. Luckily, he changed his mind later and in the 2008 report of the Italian Neurosurgical Society study group, my and his name sit together. After the first implants, in 2002, I came across people from a US start-up (NorthStar Neuroscience, Seattle) who were beginning to investigate MI ECS for stroke (at the time I was ready to implant my first patient for stroke rehabilitation). They accepted to run a small trial, which however turned out a less favorable result than expected. In 2002, Pagni, who was close to retirement (November 2003) finally caught up to the idea and spearheaded an ad-hoc group of the Italian Neurosurgical Society to test MI ECS for Parkinson and other movement disorders. Incredibly, only a few groups worldwide took up the technique since the publication of my first report in January 2000 in the journal Movement Disorders. Three reasons can be adduced: 1- Deep brain stimulation is FDA approved for PD and the huge marketing effort from the manufacturer certainly swayed many neurosurgeons to stick to their guns. 2- Not many neurosurgeons have experience with MI ECS, which remains niche surgery. Even Lozanos group failed. Given their prestige and impeccable credentials, this failure is reminiscent of some initial papers on MI ECS for central pain, in which a few authors reported no benefit (e.g. Meyerson in Stockholm or Aziz in Oxford), while others were by far more successful. Whereas Benabid had the good fortune of drawing many interested people from all over to learn subthalamic DBS directly from him, the same did not happen for MI ECS. In view of the supposed simplicity, some of the few bold ones simply rushed in, without prior experience with MI ECS for pain. 3- There is a philosophical reason: neurosurgeons are both enamored of their ability to be precise and the high-tech glittering technology involved. It is in a word- empowering. Contrast this with the relative low-tech Soviet style simplicity of MI ECS.
EXPLAINING EFFECTS
A- Why does Modulation of the Primary Motor Cortex (MI) Improve Motor Symptoms in PD? 1- MI belongs to the bi-directionally interconnected cortico-basal ganglia-cortical circuit and stimulation at any station in the loop can, therefore, elicit functional changes throughout the circuit(Priori and Lefaucheur 2007) which is so evident that one wonders why Turin University (but not Milan or Paris University) neurologists did not know it in
Invasive Cortical Stimulation for Parkinsons Disease (PD)
219
1996. The changes in cortical excitability induced by DBS also argue for the involvement of MI in the mechanism of action and suggest it as a therapeutic strategy (reviewed in Priori and Lefaucheur 2007). Conversely, modulation of MI can affect basal ganglia activities through various glutamatergic cortico-striatal and cortico-subthalamic (C-STN) projections (Canavero and Bonicalzi 2007 for review of studies). Strafella et al (2004) for one- demonstrated that 74.9% of neurons in the dorsolateral STN respond to TMS of the ipsilateral motor cortex in humans. This response is characterized by a short-latency short-duration excitation followed by a long lasting inhibition (more than 100ms). They concluded that these findings clearly indicate that the human motor cortex exerts a powerful modulatory influence over the STN. In PD patients, r TMS of MI induced an increase of dopamine release in the ipsilateral putamen along with a spatially enlarged area of dopamine release (Strafella et al 2005). Kim et al (2007) compared the UPDRS III (motor) and the amount of extracellular dopamine concentration using [11C] Raclopride PET before and after two sessions of rTMS (15 trains, 5Hz, 90% rTMS, MI hand area) in 9 PD patients off medication. Unilateral stimulation increased the amount of dopamine in the striatum (head of caudate and putamen) bilaterally and improved motor symptoms. Fibers projecting from the MI hand area are located mainly in the most lateral and dorsal part of the STN (Afsharpour 1985, Nambu et al 2000), while those to the lateral part of the putamen are arranged dorso (leg, then arm)-ventrally (face) (Jones et al 1977, Takada et al 1998, Tokuno et al 1999). The supplementary motor area (SMA) and the premotor cortex (PMC) are also linked to MI. The effects of MI ECS do not likely depend on pyramidal cells and axons, but rather on small inhibitory axons (probably Golgi-II cells with long axons) connected to pyramidal cells in the cortex itself and/or with the axons of afferents and efferents running parallel to the stimulating electrodes (Canavero and Bonicalzi 2007, Priori and Lefaucheur 2007). Therefore, MI ECS may orthodromically and/or antidromically activate fibers connecting MI to the basal ganglia. 2- MI ECS may act locally decreasing cortical excitability (chapter 12). 3- MI ECS could resynchronize a disrupted pattern of cortico-thalamo-ganglionar oscillation (Canavero et al 2003). The basal ganglia-thalamo-cortical (MI/PMC) pathway is comprised of dynamically coupled, nonlinear, reentrant, polysynaptic oscillators representing a wide range of frequencies (Montgomery 2007). In particular, MI shows beta (20 Hz) oscillations coherent with similar oscillations in the activity of contralateral contracting muscles. The coherence is weaker in trunk than it is in limb muscles (Salenius and Hari 2003). These oscillations engage not just motor structures, but also return from muscle to brain via feedback afferent pathways. SI too has strong beta-band oscillations, which are synchronized with those in MI, allowing oscillatory sensory reafferences to be interpreted in the context of the oscillatory motor command which produced it (Baker 2008). An abnormal synchronization between cortical and basal ganglia has been demonstrated in human PD (Brown 2003).This activity was modulated by STN stimulation, which decreased the abnormal spreading of desynchronization and increased MI activity during movement preparation and execution, with a correlated improvement in bradykinesia (Devos et al 2004). MI ECS, like DBS, may override pathological firing patterns, reduce abnormal oscillations and replace them with less deleterious patterns of activity. A simple excitation-inhibition-frequency explanation is thus refocused on firing patterns. Thus, TMS applied to MI induces transient synchronization of
220
Sergio Canavero
spontaneous activity of cortical neurons within the beta frequency range (Salenius and Hari 2003). MI rTMS (>5 Hz) suppressed activity in the human subthalamic nucleus (DoyleGaynor et al 2008). 4- The clinical course of some MI ECS effects ( i.e. the gradual loss of effect after cessation of stimulation) suggests the possibility of time consuming processes such as synaptic plasticity (Canavero and Paolotti 2000, Canavero and Bonicalzi 2007), long term potentiation or depression, expression of secondary messengers, or polarization of brain tissue (Priori and Lefaucheur 2007).
B- Why are Effects of Unilateral Stimulation Bilateral? This effect should not come as a surprise (although it was for me on my first surgery for PD!). Transcallosal pathways are responsible for the effect. The anatomical, neuroimaging (e.g. Kim et al 2007) and TMS (e.g. Plewnia et al 2003) evidence is overwhelming and will not be reviewed here. Suffice it to say that effective inter-hemispheric conduction pathways exist between the hand representations of MI (Meyer et al 1995), but weaker transcallosal connections for body parts outside hand areas (Komssi et al 2002), which explains why the hand area should be targeted for MI ECS in PD. Transcallosal fibers connect homotopic as well as heterotopic areas (e.g. Cavada and Goldman-Rakic 1989). MI has also ipsilateral projections which are important for axial muscles and muscles supplied by cranial nerves and more generally in the generation of bilateral synergistic movements (Muellbacher et al 2001). Seyal et al (2005) showed that 0.3 Hz rTMS of the right MI also inhibits contralateral SI. Thus, MI stimulation has effects that extend to both contralateral MI and SI via the corpus callosum. However, at least in principle, subcortical pathways are possible mediators of the contralateral response (interhemispheric connections at thalamic and basal ganglia levels, including a pathway connecting the subthalamic nuclei-Liu et al 2002).
IS THE PRIMARY MOTOR CORTEX THE ONLY PRIMARY MOTOR CORTEX?

Up to now, ECS for PD has focused on MI as traditionally described in textbooks. Textbooks, though, do not convey the full picture. Since the original studies by Horsley, Foerster and Ferrier in the first half of the XX century, we know that MI and SI are both partially sensory and motor, due to the coexistence within the same neurons of motor and sensory properties. Penfield and Jasper (1954) concluded that motor units are arranged in horizontal strips that cross the central fissure and other neurosurgeons confirmed that (hand) motor responses occur in both MI and SI (Woolsey et al 1979, Uematsu et al 1992, Nii et al 1996), with great variability in superior-toinferior distribution (the MI face-arm motor boundary is 1-2 cm higher than the corresponding SI sensory boundary). The same applies to sensory responses, although with a distribution more ventral than that of motor responses. Mixed motor and sensory responses tend to be limited to the middle part of the central sulcus. Arrest of simple hand repetitive
221
voluntary movements occurs widely throughout the premotor and MI/SI. Thus, MI and SI hand cortices overlap and are not divided in a simple manner by the central sulcus. In particular, hand motor responses extend 4.7 cm anterior (i.e. 3.7 cm anterior to MI) and 3.4 cm posterior to the central sulcus, while sensory responses extend 6.8 cm anterior and 2.2 cm posterior to the central sulcus. Up to 59% of sensory responses may be obtained anterior to the central sulcus. Motor arrest and complex responses can also be seen in the posterior part of the inferior frontal gyrus or lateral end of the precentral gyrus. Sensory and motor thresholds also change from day to day in relation to time of day, mood, last meal and hours of last sleep. fMRI data confirm this picture: the motor hand area may extend to (50% of cases), or be located exclusively, in SI (20% of cases), even during the simplest motor tasks (Canavero and Bonicalzi 2007). Apart from intrinsic responses, MI and SI are so tightly interconnected by short corticocortical U-fibers that arborize over a considerable rostrocaudal distance in MI to make them almost a unique structure (Canedo 1997). SI is a major source of somatosensory input to MI and MI is strongly modulated by sensory flow (and vice-versa) (Enomoto et al 2001, Reis et al 2008). Clearly, uniformly targeting MI in ECS efforts for PD may be misplaced: SI could be another potential target. Also, BA44 (found 2 cm anterior to MI tongue area) has direct fast conducting cortico-spinal projections with a role in voluntary hand movements (Uozumi et al 2004), confirming the haziness of MI borders.
IS THERE A HOMUNCULUS IN MI?

Evidence shows a rough body-centered map of MI that matches the traditional motor homunculus. This map extends to nearby premotor areas. However, the picture is more complex than this. In one out of five patients, there are variations in the organization of MI, i.e. mosaicism (overlapping of functional areas), variability (inverted disposition of MI functional areas) or both (Branco et al 2003). Moreover, the local mosaic-like topography (somatotopy) of individual distal arm representations is highly idiosyncratic, with wide variability among individuals (Penfield and Boldrey 1937). Somatotopic differences not only exist between individuals, but also between hemispheres in an individual. Within-limb somatotopy is not spatially discrete, nor sequentially ordered (Schieber 2001). Ever since Penfield, studies showed that within the face, arm or leg regions, the territories from which stimulation can elicit movement of different parts overlap quite extensively. Rather than discrete regions of MI controlling different parts of the arm, control of each part is mediated by an extensive territory that overlaps with the territories controlling other parts (Schieber 2001, Beinsteiner et al 2001). Whereas the prior view suggested that stimulation of different regions of MI should elicit movement of different body parts, it is now clear that stimulation can elicit movement of a given body part from a broad region, i.e. MI has a broadly overlapping mosaic of points where stimulation elicits movements of different body parts. Any given MI neuron may influence the motoneuron pools of several muscles (not just one). Selective stimulation of different regions in MI can produce the same movement, due to intra-MI dense bi-directional projections of up to 1 cm. Limb joints are represented in the cortex more than once, but with different contiguity (shoulder to wrist, shoulder to elbow) (Schieber 2001). Rather than simply controlling different body parts,
222
Sergio Canavero
MI directs a host of body parts to assume complex postures. The map appears to be organized not just according to muscle groups, but to the positions in space where the movements conclude (Graziano 2006). Two dissociable systems for motor control (one for the execution of small precise movements -especially distal muscles- and another for postural stabilization especially proximal muscles) coexist in MI, with the representation of distal and proximal muscles substantially intermingled within the MI arm representation. Depending on duration of stimuli applied on MI, simple or complex movements can be elicited. Neurons in the rostral portion of MI may be more related to kinematic variables such as velocity and movement direction than more caudally placed cells (Hatsopoulos et al 2007). In PD specifically, map shifts are found in the majority of the patients, both in untreated early cases and treated cases of long duration, with a correlation between inter-side differences in the severity of PD symptoms and inter-hemispheric map displacement (Thickbroom et al 2006). This makes MI ECS less straightforward than commonly thought.
WHAT IS THE MOTOR CORTEX?

This might sound a nave question, but is not. MI is far from the passive servant of higher motor structures. It performs a complex integration of multiple influences, originating in both cerebral hemispheres, in a role as the ultimate gate-keeper that is carefully and differentially tuned to generate well-defined motor behaviors (Reis et al 2008). The discharge pattern of individual MI neurons conveys a bewildering diversity of information. Thus, some neurons receive strong sensory input, whereas others do not. Some neurons respond to contralateral, ipsilateral or bilateral movements; some neurons even reflect sensory signals used to guide action (Scott 2003). Many pyramidal tract neurons respond with a wide range of peripheral inputs (visuo-audio-vestibular) (Canedo 1997). The left and right hemispheres are specialized for controlling different features of movement. In reaching movements, the non-dominant arm appears better adapted for achieving accurate final positions and the dominant arm for specifying initial trajectory features (e.g. movement direction and peak acceleration) (Schaefer et al 2007). Also, the area of hand representation is greater in the dominant (left) than in the nondominant hemisphere, with greater dispersion of elementary movement representations and more profuse horizontal connections between them, thus leading to more dexterous behavior of the dominant hand (Hammond 2002). Stronger beta rebound after right median nerve stimulation is observed in the left compared with the right hemisphere (Salenius and Hari 2003). On the whole, left MI ECS might be preferable.
CONCLUSIONS
On the basis of evidence presented up to now, the following conclusions are possible:
223
1- the three PD axes may need different parameters, and the final choice must take into account the most disabling symptom. It is important to remember that identical TMS paradigms elicit opposite physiological effects when applied to neighboring cortical regions (Speer et al 2003) or different subjects (Gangitano et al 2002, Sommer et al 2002), that these effects are state-dependent (Pascual Leone et al 1998) and may even depend on actual mood of the subject. There exist subpopulations with specific patterns of response to cortical stimulation. Kim et al (2007) reported that rTMS increased raclopide binding 58.3% in the caudate of one patient but decreased it by 42.9% in anothers ! Relatively small variations in TMS parameters may result in unintended effects locally and remotely (Seyal et al 2005). This calls for extensive parameters search (Box 1). Box I. TANG guidelines
-Paddle implanted contralateral to side of initial symptoms; also evaluate implant contralateral to less affected side in nonresponders -Bipolar stimulation: electrode setting: 0+/3-BLOCK A (4 days) 60s 0.5V 10,20,30,40,50, 80, 130 Hz; 1V 10,20,30,40,50, 80, 130 Hz; 2V 10,20,30,40,50, 80, 130 Hz; 3V 10,20,30,40,50, 80, 130 Hz each setting kept 1hour -BLOCK B (4 days) 180 s 0.5V 10,20,30,40,50, 80, 130 Hz; 1V 10,20,30,40,50, 80, 130 Hz; 2V 10,20,30,40,50, 80, 130 Hz; 3V 10,20,30,40,50, 80, 130 Hz each setting kept 1 hour -BLOCK C (4 days) 450 s 0.5V 10,20,30,40,50, 80, 130 Hz; 1V 10,20,30,40,50, 80, 130 Hz; 2V 10,20,30,40,50, 80, 130 Hz; 3V 10,20,30,40,50, 80, 130 Hz each setting kept 1 hour -12 day block repeated with the reverse setting, 0-/3+ (total: 24 days) and then for other settings (1-/2+ and 2+/1-) -For definitive stimulation: either continuous or cyclical stimulation (1h On, 1h OFF or the like), with switch-off at night
2- MI ECS effects, unlike DBS, are almost never immediate. Intervals of assessment after a change of parameters must take into account that, after about 2-4 weeks, a long after-effect sets in as a result of neuroplastic changes. Moreover, effects, particularly on akinesia, grow over time. After 1-3 months of stimulation, UPDRS scores can be evaluated, after changing parameters, only at -at least- 1 month intervals. I would also add that, at least initially, effects of ECS on motor scores should not be evaluated off medication.
224
Sergio Canavero 3- MI is not just classical Brodmanns area 4: more anterior and posterior areas must be investigated. Premotor cortex BA6 (but not SMA, based on rTMS evidence: chapter 12) may be targeted in future studies: it lies on the crown of the precentral gyrus, thus needing less energy for activation, while MI is mostly within the central sulcus. Premotor rTMS influences interneurons in MI through corticocortical connections (Muenchau et al 2002). Thus, I suggest to stimulate MI and SI or MI and PMC with two parallel paddles in nonresponders. 4- The primary target for initial stimulation is the hand area in MI/SI unilaterally. Unilateral stimulation of the hand area induces significant perfusion reductions on SPECT cortically and subcortically, bilaterally and symmetrically (chapter 13). Bilateral stimulation is warranted in failures or failing cases: continuous stimulation may lead to cortical habituation and alternate stimulation may be a solution.
REFERENCES
Afsharpour S. Topographical projection of the cerebral cortex to the subthalamic nucleus. J Comp Neurol 1985 ; 236:14-28. Alberts WW. A simple view of parkinsonian tremor. Electrical stimulation of cortex adjacent to the rolandic fissure in awake man. Brain Res 1972; 44: 357-369. Baker SN. Oscillatory interactions between sensorimotor cortex and the periphery. Curr Op Neurobiol 2008; 17: 1-7. Beinsteiner R, Windischberger C, Lanzenberger R, Edward W, Cunnington R, Erdler M. Finger somatotopy in human motor cortex. NeuroImage 2001; 13:1016-1026. Branco DM, Coelho TM, Branco BM, et al. Functional variability of the human cortical motor map: electrical stimulation findings in perirolandic epilepsy surgery. J Clin Neurophysiol 2003; 20: 17-25. Brown P. Oscillatory nature of human basal ganglia activity. Relationship to the pathophysiology of Parkinsons disease. Mov Disord 2003; 18:357-363. Bucy PC. Special article: The neural mechanisms of athetosis and tremor. Ann Surg 1945; 122:943-954. Canavero S, Bonicalzi V, Paolotti R, et al. Therapeutic extradural cortical stimulation for movement disorders: a review. Neurol Res 2003; 25: 118-122. Canavero S, Bonicalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. II Neural Networks Surgery. Wien: Springer-Verlag, 2007, 223-232. Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey: II. Evidence for segregated corticocortical networks linking sensory and limbic areas with the frontal lobe. J Comp Neurol. 1989;287:422-45. Canedo A. Primary motor cortex influences on the descending and ascending systems. Progr Neurobiol 1997; 51 287-335. Devos D, Labyt E, Derambure P, et al. Subthalamic nucleus stimulation modulates motor cortex oscillatory activity in Parkinsons disease. Brain 2004; 127:408-419.
225
Doyle-Gaynor LMF, Kuehn AA, DiLeone M, et al. Suppression of beta oscillations in the subthalamic nucleus following cortical stimulation in humans. Eur J Neurosci 2008; 28: 1686-1695. Enomoto H, Ugawa Y, Hanajima R, et al. Decreased sensory cortical excitability after 1Hz rTMS over the ipsilateral primary motor cortex. Clin Neurophysiol 2001; 112: 21542158. Gangitano M, Valero-Cabr A, Tormos JM, Mottaghy FM, Romero JR, Pascual-Leone A. Modulation of input-output curves by low and high frequency repetitive transcranial magnetic stimulation. Clin Neurophysiol 2002; 113: 1249-1257. Graziano M.The organization of behavioral repertoire in motor cortex. Annu Rev Neurosci. 2006;29:105-34. Hammond G. Correlates of human handedness in primary motor cortex: a review and hypothesis. Neurosci Biobehav Rev 2002; 26: 285-292. Hatsopoulos N, Xu Q, Amit Y. Encoding of movement fragments in the motor cortex. J Neuroscience 2007; 27: 5105-5114. Jones EG, Coulter JD, Burton H, Porter R. Cells of origin and terminal distribution of corticostriatal fibers arising in the sensory-motor cortex of monkeys J Comp Neurol 1977; 173: 53-80. Kim JY, Chung EJ, Lee WY, et al. Therapeutic effect of repetitive transcranial magnetic stimulation in Parkinsons disease: analysis of [11C] raclopride PET study. Mov Disorders 2007 23: 207-211. Komssi S, Aronen HJ, Huttunen J, et al. Ipsi- and contralateral EEG reactions to transcranial magnetic stimulation. Clin Neurophysiol 2002; 113: 175-184. Liu X, Ford-Dunn HL, Hayward GN, et al. The oscillatory activity in the parkinsonian subthalamic nucleus investigated using the macro-electrodes for deep brain stimulation. Clin Neurophysiol 2002; 113: 1667-1672. Meyer BU, Rricht S, Grfin von Einsiedel H, Kruggel F, Weindl A. Inhibitory and excitatory interhemispheric transfers between motor cortical areas in normal humans and patients with abnormalities of the corpus callosum. Brain. 1995;118 ( Pt 2):429-40. Montgomery EB Jr. Basal ganglia physiology and pathophysiology: a reappraisal. Parkinsonism Rel Disorders 2007; 13: 455-465. Muellbacher W, Boroojerdi B, Ziemann U, Hallett M. Analogous corticocortical inhibition and facilitation in ipsilateral and contralateral human motor cortex representations of the tongue. J Clin Neurophysiol 2001 18 550-558. Muenchau A, Bloem BR, Irlbacher K, Trimble MR, Rothwell JC. Functional connectivity of human premotor and motor cortex explored with repetitive transcranial magnetic stimulation. J Neuroscience 22: 554-561, 2002. Nambu A, Tokuno H, Hamada I, et al. Excitatory cortical inputs to pallidal neurons via the subthalamic nucleus in the monkey J Neurophysiol 2000; 84:289-300. Nii Y, Uematsu S, Lesser RP, Gordon B. Does the central sulcus divide motor and sensory functions? Cortical mapping of human hand areas as revealed by electrical stimulation through subdural grid electrodes. Neurology. 1996;46:360-7.
226
Sergio Canavero
Pascual-Leone A, Tormos JM, Keenan J, Tarazona F, Caete C, Catal MD. Study and modulation of human cortical excitability with transcranial magnetic stimulation. J Clin Neurophysiol. 1998;15:333-43. Penfield W , Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation Brain 1937; 60:389-443. Penfield W, Jasper H. Epilepsy and the functional anatomy of the human brain. Boston: Little-Brown, 1954. Plewnia C, Lotze M, Gerloff C. Disinhibition of the contralateral motor cortex by lowfrequency rTMS. Neuroreport 2003; 14: 609-612. Priori A, Lefaucheur JP. Chronic epidural motor cortical stimulation for movement disorders Lancet Neurol 2007; 6:279-286. Reis j, Swayne OB, Vandermeeren Y, et al. contribution of transcranial magnetic stimulation to the understanding of cortical mechanisms involved in motor control. J Physiol 2008; 586.2: 325-351. Salenius S, Hari R. Synchronous cortical oscillatory activity during motor action. Curr op Neurobiol 2003 ;13: 678-684. Schaefer SY, Haaland KY, Sainburg RL. Ipsilesional motor deficits following stroke reflect hemispheric specializations for movement control. Brain 130: 2146-2158, 2007. Schieber MH. Constraints on somatotopic organization in the primary motor cortex. J Neurophysiol 86: 2125-2143, 2001. Scott SH. The role of primary motor cortex in goal-directed movements: insights from neurophysiological studies on non-human primates. Curr Op Neurobiol 2003; 13: 671677. Seyal M, Shatzel AJ, Richardson SP. Crossed inhibition of sensory cortex by 0.3 Hz transcranial magnetic stimulation of motor cortex. J Clin Neurophysiol 2005; 22: 418421. Sommer M Sommer M, Wu T, Tergau F, Paulus W. Intra- and interindividual variability of motor responses to repetitive transcranial magnetic stimulation. Clin Neurophysiol 2002;113: 265-269. Speelman JD, Bosch DA. Resurgence of functional neurosurgery for Parkinsons disease: a historical perspective. Mov Disord 1998; 13: 582-588. Speer AM, Willis MW, Herscovitch P, et al. Intensity-dependent regional cerebral blood flow during 1-Hz repetitive transcranial magnetic stimulation (rTMS) in healthy volunteers studied with H215O positron emission tomography: I. Effects of primary motor cortex rTMS. Biol Psychiatry. 2003;54:818-25. Strafella AP, Vanderwerf Y, Sadikot AF. Transcranial magnetic stimulation of the human motor cortex influences the neural activity of subthalamic nucleus. Eur J Neurosci 2004; 20:2245-2249. Strafella AP, Ko JH, Grant J, Fracaccio M, Monchi O. Corticostriatal functional interactions in Parkinsons disease: a rTMS/11Craclopride PET study. Eur J Neurosci 2005; 22:2946-2952. Thickbroom GW, Byrnes ML, Walters S, Stell R, Mastaglia FL. Motor cortex reorganisation in Parkinson's disease. J Clin Neurosci 2006; 13:639-42.
227
Tokuno H, Irase M, Nambu A, Akazawa T, Miyachi T, Takada M. Corticostriatal projections from distal and proximal forelimb representations of the monkey primary motor cortex Neurosci Lett 1999; 269:33-36. Uematsu S, Lesser R, Fisher RS, et al. Motor and sensory cortex in humans: topography studied with chronic subdural stimulation. Neurosurgery 1992; 31: 59-72. Uozumi T, Tamagawa A, Hashimoto T, Tsuji S. Motor hand representation in cortical area 44. Neurology 2004; 62: 757-761. Woolsey CN, Erickson T, Gilson WE. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J Neurosurg 1979; 51:476-506.
STROKE REHABILITATION
Chapter 14
NONINVASIVE STIMULATION FOR STROKE REHABILITATION

Paolo Profice, Michele Dileone, Fabio Pilato, Federico Ranieri, Fioravante Capone, Lucia Florio, Pietro A. Tonali and Vincenzo Di Lazzaro
Institute of Neurology, Catholic University, and Fondazione Don C. Gnocchi, Rome, Italy.
INTRODUCTION
Stroke is the most common cause of disability and the third leading cause of death in the United States. Approximately 700,000 strokes occur annually and about 21-36% of all stroke survivors become severely and permanently disabled. Of >5 million US stroke survivors, >1 million are functionally limited. Similar figures apply to Europe and Australasia. Age and sex-standardized annual occurrence rates for people between 45 and 84 years old are in the 300-500/100000 range (Sudlow and Warlaw 1997). There is some degree of spontaneous improvement over the first five months after a stroke, but variability between patients is remarkable, and at six months patients over 65 years of age show variable degrees of hemiparesis (50%), aphasia (19%) and depressive symptoms (35%); several patients are unable to walk without assistance (30%), are dependent in daily activities (26%) and have to be institutionalized (26%) (AHA 2007), with considerable costs associated with primary care and loss of income. There are no fully restorative treatments for chronic stroke patients and seeking new therapeutic options is mandatory. In stroke lesioned brains, the surviving system reorganizes to make the most efficient use of the remaining network in order to regain control over muscles deprived of their normal corticospinal inputs; the areas involved include intact task-related areas in the affected hemisphere, homologous areas in the unaffected hemisphere, and peri-infarct
Correspondence concerning this article should be addressed to: Dr. Vincenzo Di Lazzaro, M.D. e-mail: vdilazzaro@rm.unicatt.it.
232
Paolo Profice, Michele Dileone, Fabio Pilato et al.

regions; this process may occur even beyond the resolution of acute changes (Nudo 2006). This reorganization falls into the concept of plasticity that could be defined as any enduring change in cortical properties either morphological or functional (Sanes and Donoghue 2000). Plasticity was initially considered a property of the maturing nervous system, but it is now clear that it occurs throughout life: it is continuously engaged by experience and learning and enhanced after brain lesions through gradually developing steps. The necessity or sufficiency of plasticity changes in behavioral recovery is presently still poorly understood and it has been difficult to determine whether such phenomena are adaptive or maladaptive (Nudo 2007). Some experimental evidences suggest that healthy hemisphere hyperactivation consequent to unilateral brain lesion can be maladaptive for functional outcome (Murase et al 2004, Allred and Jones 2008). Furthermore, functional neuroimaging studies have demonstrated that recovered motor function in the paretic hand of chronic stroke patients relies predominantly on reorganized activity within motor areas of the affected hemisphere (Ward et al 2003). It can be hypothesized that reorganization in the adjacent intact cortex and contralateral healthy hemisphere could facilitate functional brain recovery. rTMS can modulate cerebral cortex excitability non-invasively and seems to be a promising tool for driving plasticity in the damaged brain (Di Lazzaro et al 2008). By delivering repetitive pulses of TMS, it is possible to produce effects on cortical circuits that outlast the duration of the stimulation. Noninvasive cortical stimulation can enhance the beneficial effects of motor training, visuomotor coordination, implicit motor learning, skilled finger movements, probabilistic classification learning, working memory and sleep-dependent consolidation of declarative memories in healthy volunteers (see references in Hummel and Cohen 2005). The mechanisms of modulation of cortical excitability by rTMS are unclear, but they might be related to long term changes in synaptic transmission analogous to long-term potentiation (LTP) and long-term depression (LTD) (Ziemann 2004).
THE RIVALRY MODEL: A POSSIBLE RATIONALE FOR APPLYING RTMS IN STROKE

It has been suggested that a hemispheric stroke can interfere with the physiological coupling and balancing between the two sides of the brain, releasing the unaffected hemisphere as a consequence. Thus, the function of the affected hemisphere is disturbed both by the lesion and by the strong inhibitory influence of the unaffected hemisphere, via transcallosal pathways, that determine hypoexcitability of MI of the lesioned hemisphere and negatively influence motor recovery. Some neurophysiological and neuroimaging experimental findings are in favor of this theory (Ward et al 2003, Werhahn et al 2003, Cramer et al 1997, Chollet et al 1991, Fridman et al 2004, Johansen-Berg et al 2002, Carey et al 2002). Moreover, in normal subjects, it has been showed that inhibitory (1 Hz) rTMS of the left MI increased the excitability of the right MI and reduced inter-hemispheric inhibition from the left-to-right hemisphere (Gilio et al 2003). Similar effects could be induced by brain lesions. Liepert et al. (2000) have examined patients in the first few weeks after a stroke and found an increased excitability (decrease in MEP threshold) of the unaffected hemisphere. They suggested that the stroke had removed a tonic inter-hemispheric inhibition from the
Noninvasive Stimulation for Stroke Rehabilitation
233
lesioned hemisphere and increased the excitability of the unaffected hemisphere. A positive effect on hand motor function after inhibitory stimulation of ipsilateral MI has also been reported in normal subjects (Kobayashi et al 2004) and in stroke patients (Werhahn et al 2003). Moreover, in an elegant experiment, Murase et al (2004) studied inter-hemispheric inhibition at different timings during a reaction-time task with the paretic hand, in chronic, subcortical stroke patients. Their findings indicate that abnormal inter-hemispheric interactions operate in generating voluntary movements of the paretic hand in stroke patients. These findings support the view of an interhemispheric unbalancing in stroke, are in line with interhemispheric models in other domains such as attention, memory, language and somatosensory processing (Floel et al 2004, Naeser et al 2005, Oliveri et al 1999, Corbetta et al 2005) and justify two different approaches for testing the potential therapeutic role of rTMS: upregulating excitability within the cortex of the affected hemisphere or downregulating the excitability in the intact one.
SAFETY OF RTMS IN STROKE PATIENTS

Although guidelines for the safe combination of rTMS frequency, intensity and train length to stimulate MI of healthy subjects have been established (chapter 4), it is unclear whether the established parameters are completely safe if used to stimulate MI of chronic stroke patients who have a higher risk of seizure occurrence than healthy subjects. It has been reported that rTMS above treshold at 20 and 25 Hz is not safe for patients with chronic stroke, as seizures could be induced in such patients (Lomarev et al 2007). However, no serious side effect or seizure induction has been reported in studies exploring the effect of rTMS in stroke patients up to now (Talelli et al 2007, Di Lazzaro et al 2008) and in clinical trials applying rTMS to stroke recovery (Talelli and Rothwell 2006), either when stimulation was delivered on the healthy hemisphere or on the affected one.
RTMS
DELIVERED ON THE AFFECTED HEMISPHERE IN STROKE RECOVERY
rTMS at a frequency of 5 Hz or more increases the excitability of MI in normal subjects (Pascual-Leone et al 1994, Maeda et al 2000) and in stroke patients (Di Lazzaro et al 2008). The effect of 5Hz rTMS seems to be mediated by a selective reduction of the excitability of GABAergic networks in the human MI (Di Lazzaro et al 2002) and decreases GABA related inhibition, facilitating practice-dependent plasticity (Ziemann et al 2001). These data suggest that rTMS of the lesioned hemisphere could facilitate plasticity modulating GABA activity in MI. The first published report on the effect of rTMS for stroke recovery (and the only one on post-stroke recovery of lower limb function) was by Uy et al (2003). The authors applied rTMS in stroke patients in an observational uncontrolled study with a stimulation paradigm suggested by a previous experiment on normal subjects by Stefan et al (2000): rTMS at 0.1 Hz was delivered over the affected lower limb motor area, coupled with electric shocks (500
234
ms train of 10 Hz, 1ms) given 35 ms before TMS over the common peroneal nerve in chronic stroke patients. Peripheral and cortical stimulation were applied at an intensity that evoked a just visible motor response 30 minutes every week-day for four weeks. Improvement in some neurophysiological and functional measures was observed. This kind of paradigm is called Paired Associative Stimulation (PAS) and is thought to induce LTP-like phenomena. LTP induced by PAS is called associative, or `Hebbian', since it occurs on concomitant presentation of two different kinds of stimuli (i.e. peripheral electric shock and magnetic stimulation over MI) on the same target cells, leading to a final and synchronous postsynaptic depolarization at the level of MI (Stefan et al 2000). Khedr et al (2005) studied a population of 52 unselected acute stroke patients (within the first 2 weeks of the stroke) receiving standard in-patient rehabilitation. They were randomized to sham or effective stimulation. Affected hand MI was stimulated at 3 Hz (ten 10-second trains, 120 % AMT, of the unaffected hemisphere with 50 s intervals between each train) for 10 consecutive days. Disability scales measured before rTMS, at the end of the last rTMS session, and 10 days later showed that real rTMS improved patients clinical scores (NIH Stroke Scale and Scandinavian Stroke Scale) more than sham stimulation (net change around 35%). However, no benefit was observed in patients with massive middle cerebral artery infarct. Kim et al (2006) measured in a population of 15 hemiparetic stroke patients (ranging from 3.5 to 41 months after stroke onset) the effect of eight 10-second trains of rTMS at 10 Hz, delivered at an intensity of 80% RMT, with an inter-train distance of 68 s, on corticospinal excitability and on motor learning. Motor skill acquisition was tested immediately after each rTMS train: the patients were instructed to practice a block of sequential finger motor tasks for 40 seconds during the intertrain intervals. The movement accuracy and movement time were measured in a sham controlled cross-over design. Highfrequency rTMS resulted in a significantly larger increase in MEP amplitude than sham rTMS, and the change was positively associated with an enhanced motor performance accuracy. The effect of single session of rTMS given as Theta Burst Stimulation in stroke patients was studied by Talelli et al. (2007). They studied six chronic stroke survivors with incomplete recovery of the hand under three conditions: excitatory TBS over the stroke hemisphere, inhibitory TBS over the intact hemisphere and sham stimulation. After excitatory TBS, simple reaction times in the paretic hands were decreased and, compared to sham stimulation, remained significantly shorter throughout the testing period. Excitatory TBS increased the amplitude of the MEPs at rest and during background contraction and the area under the InputOutput curves on the stroke side. Inhibitory TBS over the healthy hemisphere suppressed the MEPs evoked in the healthy hands, but did not change the motor behavior or the electrophysiology of the paretic hands. The effect of intermittent theta burst TMS over the affected hemisphere was directly investigated recording epidural TMS-evoked activity in a stroke patient. A direct evidence of an increase in corticospinal activity recorded after single TMS of affected hemisphere was provided. rTMS protocols that suppress the excitability of MI may be useful in the treatment of hyperkinetic disorders caused by stroke. This kind of hyperkinetic movement disorder are generally associated with abnormal excitability of the affected MI, generally sustained by a
235
small lesion in the thalamus, basal ganglia and/or related circuits. Suppression of the excitability of the affected hemisphere with rTMS in a case report was useful in reducing the hyperkinesias. Negative results were reported by Malcolm et al (2007). The authors tested the potential adjuvant effect of rTMS on motor learning in a group of stroke survivors undergoing constraint-induced therapy for upper-limb paresis. Subjects received 2000 stimulations daily for ten consecutive days. Each daily treatment of 2000 stimuli was administered as 50 trains of 40 stimuli, stimulus rate of 20 Hz, stimulus train duration of 2 sec, with an intertrain interval of 28 sec. Stimulus intensity was 90% of motor threshold. They found a significant decrease in motor threshold for subjects receiving rTMS, with no significant changes noted in those receiving sham treatment. These data provide evidence that rTMS produced a change in the excitability of the motor system. This change, however, did not translate into a clinically evident effect: regardless of group assignment, participants demonstrated significant improvement in motor function of the affected limb. The authors highlighted the complex relationship between physiologic changes and improvements in motor function.
RTMS
DELIVERED TO THE HEALTHY HEMISPHERE IN STROKE RECOVERY
Most of the rTMS studies designed to facilitate stroke recovery aimed at decreasing the excitability of the unaffected hemisphere in order to reduce potentially negative transcallosal inhibitory effects on the affected hemisphere according to the rivalry model. Stimulating the healthy hemisphere with low frequency pulses, rather than the affected one with high frequencies, is thought to be safer: an early increase in cortical excitability of the lesioned side of the brain might potentially induce seizures or increase the size of the lesioned area, even though worsening of motor function has never been observed in stroke patients following rTMS. Most of the relevant studies tested if a single application of rTMS induced an immediate gain in the selected outcome measures, but clinical application of rTMS calls for enduring effects. There are no systematic studies comparing the duration of changes in cortical excitability following one session of rTMS with the duration after repeated rTMS sessions. Speculatively, in multiple-session studies, the short-lasting effects of a single session might accumulate and lead eventually to a long lasting improvement in function. For these reasons some authors perform multiple session studies with variable follow up time.
A- Single Session Studies Mansur et al (2005) focused specifically a sham-controlled study on assessing a possible clinical benefit of rTMS on the healthy hemisphere in stroke survivors. They used a single session of 1 Hz rTMS for 10 minutes at motor threshold over MI or premotor cortex of the healthy hemisphere in 10 stroke patients. Patients showed a significant decrease in simple and choice reaction time and improved performance on the Purdue Pegboard test with their affected hand after rTMS of the unaffected MI as compared with sham rTMS. A non-
236
significant trend for the same effects was found after rTMS over premotor cortex. Takeuchi et al (2005), using similar stimulation parameters (1 Hz, 90% RMT, 25 minutes shamcontrolled) in 20 well-recovered stroke patients, found that one session of rTMS immediately induced an improvement in pinch acceleration of the affected hand, but not in pinch force. This improvement in motor function after rTMS significantly correlated with a reduced duration of trans-callosal inhibition. Acute stroke patients were also treated with rTMS over the healthy hemisphere in a single session study by Liepert et al (2007). A double-blind study of real versus placebo rTMS was conducted. Twelve patients early after subcortical stroke (mean:7 days) received 1 Hz rTMS over the healthy hemisphere (1200 stimuli of real and placebo rTMS) in a crossover design. Stimulus intensity was 90% of motor threshold at rest. Compared to sham stimulation, real rTMS improved Nine Hole Peg Test results by approximately 10%, but not grip strength in the affected hand. The effect of a single session of 1 Hz rTMS was also tested by Dafotakis et al (2008). They measured the abilities of grasping, lifting and holding an object through an instrumental object that incorporated a force sensor for grip force recording and linear acceleration sensors for recording of kinematic acceleration signals in three dimensions. They found that rTMS improved the efficiency and timing of grasping and lifting with the affected hand. The combination of rTMS at 1 Hz over the unaffected hemisphere and motor training was evaluated in a single session study as possible treatment for post-stroke recovery of arm motor function with a double blind design. The authors concluded that priming by rTMS enhances the motor training effect of the affected hand function (Takeuchi et al 2008). Nowak et al (2008), in a crossover investigation, measured the effect of 1 Hz rTMS on motor function and on neuronal activation measured by fMRI in stroke patients. They found that 1-Hz rTMS over the healthy hemisphere significantly reduced neural overactivity in contralesional motor areas, focused activity in ipsilesional motor areas, and improved movement kinematics of the affected hand in subcortical stroke.
B- Multiple Sessions Studies Few studies are available evaluating the effect of multiple sessions of rTMS on motor function after stroke. Fregni et al (2006) tested the effects of 20 minutes of 1 Hz rTMS at RMT over MI of the healthy hemisphere during and after 5 consecutive daily sessions of stimulation. They found a significant improvement of motor function performance in the affected hand that lasted for about 2 weeks. Furthermore, the authors confirmed safety during treatment with repeated neuropsychological testing and EEG monitoring. In a case report, Boggio et al (2006) found a positive effect of 1Hz rTMS of the healthy hemisphere, reporting significant and lasting improvement of hand motor function in a severely impaired patient. Benefit was also observed following tDCS in a group of stroke patients (Boggio et al 2007). rTMS was also tested in ten pediatric stroke patients (median age 13.25 years) in a doubleblind randomized controlled trial. 1 Hz rTMS was delivered every day for 8 days with a follow up of 1 week post-treatment. Contralesional inhibitory rTMS proved to be safe and feasible for pediatric patients with subcortical stroke, and seemed to improve hand function in patients with hemiparesis (Kirton et al 2008).
237
RTMS AND AND
RECOVERY OF NON MOTOR FUNCTION AFTER-STROKE
Most studies aimed to evaluate the effect of rTMS on recovery after stroke focused on hand motor function. In these studies, MI of the affected hemisphere or of the healthy one was targeted. There are several reasons for this: 1) hand function is frequently impaired in stroke patients, specially when the middle cerebral artery territory is involved; 2) the distal muscles of the upper limb have a strong cortical representation and are easily mapped on the scalp; 3) cortical motor neurons of the hand have low-thresholds for activation by TMS and presumably can be modulated by lower intensities of stimulation. Yet, stroke may often cause aphasia or hemispatial heminattention that can result from left or right hemisphere damage respectively. A few studies explored the effect of rTMS on recovery of these functions after stroke.
A- Aphasia The rivalry model in recovery from post-stroke aphasia has less experimental evidence. Winhuisen et al (2005), using PET and TMS in aphasic stroke patients, found that in some post-stroke aphasics, non dominant (right) inferior frontal gyrus activation is essential for residual language function at least in the acute phase (10 days). However, its compensatory potential seems to be less effective than in patients who recover left inferior frontal gyrus function. These results suggest a hierarchy in recovery from poststroke aphasia and a limited compensatory potential of the nondominant hemisphere. In a follow-up investigation, the authors evaluated some of the previous patients 6 weeks after stroke (Winhuisen et al 2007). Taken together, their findings show that successful outcome from poststroke aphasia seems to depend more on the integration of available language-related brain regions than on recruiting new brain regions during the rehabilitation process. Restoration of the left hemisphere network seems to be more effective, although in some cases, right hemisphere areas are integrated successfully (Winhuisen et al 2007). With an approach similar to those adopted for hand function recovery, Martin et al (2004) and Naeser et al (2005a, 2005b) investigated whether rTMS could be used to improve recovery in chronic non-fluent aphasics (see also the tDCS study by Monti et al 2008). In such patients, strong right hemisphere activation is often observed, even in the absence of behavioral improvements (Naeser et al 2004) and could be considered maladaptive. In a preliminary study, the authors applied 1 Hz rTMS over the right hemisphere in four different perisylvian language homologues in six non-fluent patients. Only rTMS of the posterior gyral portion of the pars triangularis (right Brodmanns area 45) was followed by a naming performance better than before stimulation. This was then the target for a multiple session rTMS treatment (rTMS was delivered at 1 Hz for 20 minutes, 5 days weekly for two weeks, with intensity of 90% RMT) in four chronic post-stroke aphasic patients. An improvement in naming was observed in all the patients (Martin et al 2004). In the other study (Naeser et al 2005a), four aphasic patients who were 511 years poststroke were treated with the same protocol. Signicant improvement was observed in picture naming at 2 months post-rTMS, with lasting benet at 8 months in three patients. A case
238
report confirms these findings in an aphasic patient (Naeser et al 2005b). The results are nonetheless remarkable in that they demonstrate a very long lasting effect of TMS far beyond transient first-order effects. If these findings can be independently replicated and shown to rely on a controlled regimen of TMS treatments, then rTMS might become an important part of rehabilitative therapy, at least in some aphasic patients.
B- Contralesional Visuospatial Hemineglect Spatial hemineglect is a failure to detect and explore stimuli presented contralaterally to a central nervous system lesion. It is more frequent and severe after lesions of the right hemisphere and in particular of the posterior parietal lobe. A rivalry model for spatial hemineglect has been suggested, with a disruption of attentional balance between the two hemispheres and an orienting bias to the ipsilateral side of the space; a relative hyperactivity of the unaffected hemisphere, due to release from reciprocal inhibition by its twin, seems to play a key role in the pathophysiology of this disturbance (Kinsbourne 1977). Despite several TMS studies strongly support the rivalry model for spatial hemineglect in brain lesioned patients (Oliveri et al 1999, 2001) and normal subjects (Hilgetag et al 2001, Dambeck et al 2006), only two studies on rTMS-induced recovery from neglect are available. Brighina et al (2003), using a different rTMS protocol, demonstrated that multiple sessions of 900 pulses of 1 Hz rTMS at 90% of RMT applied over the left posterior parietal cortex every other day for 2 weeks induced a significant improvement of visuo-spatial performance that remained quite unchanged 15 days later in two patients with right hemispheric stroke occurred 3 and 5 months before. Another small study found that six sessions of low-frequency rTMS over the unaffected posterior parietal cortex decreased unilateral spatial neglect for at least 6 weeks in two chronic stroke patients (Shindo et al 2006).
CONCLUSIONS
A growing number of small studies (Table 1) suggest that rTMS and tDCS may promote recovery after stroke modulating cortical activity; this modulation determines behavioral changes in brain lesioned patients that can be engaged for therapeutic purposes. rTMS, by shifting the level of cortical excitability, might allow either more or less access to brain networks relevant for therapy. The clinically functional improvements, above spontaneous recovery and conventional treatments alone, has been shown in a range of stroke types, with or without cortical involvement and at varying times post-stroke, from a few days to many years, at different ages, and with different combinations of stimulus parameters. In all cases, though, motor deficits ranged from mild to moderate: severe or even complete deficits have not yet been explored. Clearly, functional deficits, as well as neurophysiological abnormalities, can differ from one stroke patient to another, depending on localization and extension of lesion. MI lesions, for example, exhibited deficient inhibitory properties. In contrast, subcortical lesions can display an enhancement of inhibition (Liepert et al 2005). It is therefore conceivable that, in different patients, different therapeutic approaches should be
239
used. Longitudinal changes in motor-related brain activation and recovery of function have been shown in stroke patients (Ward et al 2003). It is thus also possible that some protocols of stimulation could be useful in the acute phase of stroke, whilst, in chronic stroke patients, different rTMS-induced modulation strategies might be preferable. Table 1. rTMS and tDCS studies on post-stroke motor recovery
Authors Year Study design Number of patients Time interval from stroke 3-9 years Stimulated hemisphere rTMS parameters Duration and schedule of rTMS treatment 30 min every weekday for four week. Results
Uy et al. 2003
Not randomized, not placebo group
Affected
Khedr et al 2005
Randomized shamcontrolled, parallel, double blind trial 26
5-10 days
Affected
Mansur et al 2005
Randomized, cross-over, shamcontrolled trial 10
0-12 months
unaffected
Takeuch i 2005
Randomized, parallel shamcontrolled
6-60 months
unaffected
rTMS at 0.1 Hz over lower limb motor area of affected hemisphere coupled with electric shocks (500 ms train of 10 Hz, 1ms) given 35 ms before TMS over common peroneal nerve. Peripheral and cortical stimulation were applied at an intensity that evoked a just visible motor response Abductor digiti minimi (ADM) motor cortex area of affected hemisphere at 3 Hz (ten 10second trains 120 % AMT of non stroke hemisphere with 50 s interval between each train). Figureof-8 coil rTMS (1 Hz, 100% of motor threshold, 600 pulses) to the unaffected hemisphere over the primary motor (real or sham rTMS) and over the premotor cortex (real rTMS). Figure-of-8 coil 1 Hz rTMS at 90% RMT over FDI motor cortex area 1500 pulses. Figure-
Improvement in neurophysiological and functional measures of affected lower limb
10 minutes at the same time every day for 10 days
Active rTMS improved patients disability scales scores more than sham at least 10 days poststimulation.
3 sessions
Patients showed a significant decrease in simple and choice reaction time and improved performance of the Purdue Pegboard test with their affected hand Short lasting (less then 30 min) improvement in pinch acceleration of
25 minutes Single session
240
single blind study 20 Randomized, parallel shamcontrolled trial 15 Crossover shamcontrolled single blind study 15 Boggio et al 2006 Case report 1

of-8 coil. the affected hand 20 minutes, 5 daily session significant improvement of the motor function performance in the affected hand that lasted for 2 weeks significant improvement of accuracy and time of execution of sequential motor task during and immediately after rTMS significant improvement of motor function after rTMS still present after 4 months The initial motor improvement was further enhanced after the second session. no significant effects on functional tasks evaluated by Wolf Motor Function Test (WMFT) and the Motor Activity Log (MAL)Amount of Use
Fregni et al 2006
1-10 years
unaffected
20 minutes of 1 Hz rTMS at MT over the primary motor cortex. 1200 pulses.
Kim et al 2006
3-41 months
affected
eight 10 seconds trains of 10 Hz, 80% MT with 58 s interval between each train. 160 pulses. Figure-of-8 coil 20 minutes of 1 Hz rTMS at MT over the APB motor cortex
two sessions (real and sham) at a 1week interval
23 months
unaffected
two sessions with 4 months interval compared with previous sham session
Malcolm et al 2007
Talelli et al 2007
prospective randomized, double-blind, shamcontrolled, parallel group study. 19 total 9 randomized to rTMS + constraint therapy of healthy arm (CIT), 10 to sham rTMS+ CIT three arms sham controlled trials 6
about 4 years
affected
2sec train at 20 Hz, 50 trains, 28 s interval, 90%RMT Ipsilesional M1 for two consecutive weeks
2000 stimuli daily administered as 50 trains of 40 stimuli, at a rate of 20 Hz, and with stimulus train duration of 2 secs, with an intertrain interval of 28 secs. 90% MT
31 + 37.9 months (range: 12-108)
both hemispheres
iTBS (600 stimuli) over affected hemisphere 80% AMT cTBS (300 stimuli) over unaffected hemisphere 80% AMT
single session
Liepert
double blind
7.3
unaffected
1 Hz rTMS, 1200
single session
iTBS over the stroke hemisphere transiently improved motor behaviour and corticospinal output in the paretic hands. There was no difference at any time between cTBSIH and Sham. Compared to sham

et al 2007 placebocontrolled crossover trial 12 4.5 days hemisphere stimuli, 90% RMT
241
stimulation, real rTMS improved NHPT results but not grip strength in the affected hand. No change of performance was observed for the unaffected hand. Affected hand function was better at day 10 in rTMStreated patients compared with sham-treated patients evaluated by changes in grip strength and the Melbourne assessment of upper extremity function (MAUEF) between baseline and day 10. Compared with sham stimulation, rTMS induced an increase in the excitability of the affected motor cortex (p < 0.001) and an improvement in acceleration of the affected hand (p = 0.006). the effect of motor training on pinch force was enhanced by rTMS (p < 0.001). These improvement in the motor function lasted for one week after rTMS and motor training (p < 0.001). rTMS applied to the contralesional M1 improved the kinematics of finger (frequency and peak velocity of finger tapping) and grasp movements in the affected hand (peak wrist velocity, peak velocity of grasp
Kirton et al 2008
sham controlled double blind study 10
subcorti cal AIS more than 2 years previou sly (6.33 + 3.56 years)
unaffected
1 Hz rTMS (20 min, 1200 stimuli), 100% RMT
once per day for 8 days.
Takeuch i et al 2008
sham controlled double blind study 20
range: 7-121 months
unaffected
1 Hz and a stimulus intensity of 90% rMT for 25 minutes (1500 pulses).
single session
Nowak et al 2008
Crossover investigation, sham controlled
range: 1-4 months
unaffected hemisphere
1 Hz rTMS, 10 minutes (600 stimulii)
single session
15
242

aperture and time of peak grasp aperture). 1 Hz rTMS applied over the unaffected M1, but not the vertex, improved the efficiency and timing of grasping and lifting with the affected hand. Significant reduction in JebsenTaylor hand function test in active group, especially for fine finger movements; after-effect: 25 minutes. Baseline: mild impairment. Significant improvement in motor performance in both real conditions. Cathodal > anodal (nonsignificantly). Baseline: mild impairment.
Dafokati s et al 2008
Sham controlled single blind study 12
1-4 months
unaffected
1 Hz,100% resting motor threshold over 10 min using a figure-of-eight coil applied over healthy M1.
two sessions (real and sham)
Hummel et al 2005
Randomized, double-blind, crossover study of tDCS 6
1 year or more
Anodal to contralateral MI/sham
1 mA tDCS for 20 minutes
Single session
Fregni et al 2005
Randomized, double-blind, crossover study of tDCS 6
1 year or more
Anodal to contralateral MI / cathodal to ipsilateral MI / sham
1 mA tDCS for 20 minutes
Single session
Another open question regards the optimal time after stroke for rTMS treatment. Does rTMS improve motor function by rebalancing the two hemispheres and facilitating the affected MI or also by affecting motor learning? In the first case, timing of treatment will be probably less important; if the latter is true, as suggested by Kim and colleagues (2006), the effects of rTMS would be significantly augmented when stimulating sessions are used in conjunction with a motor practice paradigm and in the acute phase. It should be further explored whether different patients might benefit from excitation of the affected hemisphere or suppression of the unaffected hemisphere or both. One approach that could turn out to be useful in individualizing treatment could be the preliminary evaluation of the baseline characteristics of excitability of the two hemispheres and baseline motor performances of the affected side, and the evaluation of the impact of a single session of rTMS on these parameters. This was performed in a recent study in a small group of patients with chronic stroke (Talelli et al 2007). This preliminary study suggests that a single session of facilitatory rTMS over the stroke hemisphere transiently improves motor behavior and corticospinal output in the paretic hand. After the demonstration of consistent -and presumably useful- changes in excitability and motor function produced by a single session of rTMS, it should be evaluated whether these changes can be enhanced and transformed into longer lasting changes by means of repeated sessions of rTMS or combining rTMS with rehabilitation therapy. The relationship between changes in brain activation and motor
243
function or motor learning are still poorly understood. Is the facilitation of the lesioned hemisphere useful for motor learning? Or, alternatively, are LTP phenomena easier to induce in a cortex with low levels of activity (as observed after inhibitory rTMS of motor cortex) according to the homeostatic plasticity concept ? (Talelli and Rothwell 2006). At the present time, it is still not clear which are the best frequencies, number and intensity of rTMS pulses. Moreover, different protocols of stimulation are being developed, combining trains or pairs of stimuli (Huang et al 2005, Thickbroom et al 2006), or coupling rTMS with electrical peripheral stimulation (Hummel and Cohen 2005). Also, different cortical areas, e.g. premotor and secondary motor cortices, have to be explored as potential target of therapeutic rTMS. Novel ways to enhance the effect of non-invasive stimulation are being explored. Repetitive application of 4 monophasic magnetic pulses (quadropulse stimulation) at 1.5 ms intervals repeated every 5 seconds over MI could be more effective than paired-pulse stimulation for plasticity induction in the human MI (Hamada et al 2007). Also, preconditioning (priming) of MI with a suppressive stimulus (cathodal tDCS) leads to an increase in the facilitatory response to subsequent 1 Hz rTMS (and viceversa with anodal tDCS) (Siebner et al 2004). The successful implementation of these techniques as interventional strategies for different patients groups will rely on improved understanding of underlying neuronal correlates of functional recovery and of the physiological effects of stimulation on cortical connectivity. At the same time, the positive effect of rTMS in chronic stroke patients should be confirmed in larger clinical trials before widespread clinical application. Also, as discussed for chronic pain, rTMS and dCS may aid in the screening of patients that might benefit from a chronic implant (chapter 15). rTMS also improved both motor and sensory functions of 4 incomplete spinal cord injury patients (Belci et al 2004), and this is another avenue to be explored in the wider field of neurologic rehabilitation.
REFERENCES
Allred RP, Jones TA. Maladaptive effects of learning with the less-affected forelimb after focal cortical infarcts in rats. Exp Neurol. 2008;210:172-81 American Heart Association. Heart disease and stroke statistics-2007 update. A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Dallas, Tx: American Heart Association, 2007. Belci M, Catley M, Husain M, Frankel HL, Davey NJ. Magnetic brain stimulation can improve clinical outcome in incomplete spinal cord injured patients. Spinal Cord. 2004;42:417-9. Boggio PS, Alonso-Alonso M, Mansur CG, et al. Hand function improvement with lowfrequency repetitive transcranial magnetic stimulation of the unaffected hemisphere in a severe case of stroke. Am J Phys Med Rehabil. 2006;85:927-30.
244
Boggio PS, Nunes A, Rigonatti SP, Nitsche MA, Pascual-Leone A, Fregni F. Repeated sessions of noninvasive brain DC stimulation is associated with motor function improvement in stroke patients. Restor Neurol Neurosci 2007;25:123-9. Brighina F, Bisiach E, Oliveri M, et al. 1 Hz repetitive transcranial magnetic stimulation of the unaffected hemisphere ameliorates contralesional visuospatial neglect in humans. Neurosci Lett 2003; 336, 131-3 Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term outcome of arm function after stroke: results of a follow-up sudy. Disabil Rehabil. 1999; 21, 357-64. Carey JR, Kimberley TJ, Lewis SM, et al. Analysis of fMRI and finger tracking training in subjects with chronic stroke. Brain 2002; 125(Pt 4), 773-88 Chollet F, DiPiero V, Wise RJ, Brooks DJ, Dolan RJ, Frackowiak RS. The functional anatomy of motor recovery after stroke in humans: a study with positron emission tomography. Ann Neurol. 1991; 29, 63-71 Corbetta M, Kircade MJ, Lewis C, Snyder AZ, Sapir A. Neural basis and recovery of spatial attention deficits in spatial neglect. Nat Neurosci 2005; 8: 160310. Cramer SC, Nelles G, Benson RR, at al. A functional MRI study of subjects recovered from hemiparetic stroke. Stroke. 1997; 28, 2518-27 Dafotakis M, Grefkes C, Eickhoff SB, Karbe H, Fink GR, Nowak DA. Effects of rTMS on grip force control following subcortical stroke. Exp Neurol. 2008;211:407-12 Dambeck N, Sparing R, Meister IG, et al. Interhemispheric imbalance during visuospatial attention investigated by unilateral and bilateral TMS over human parietal cortices. Brain Res. 2006;1072:194-9. Di Lazzaro V, Oliviero A, Mazzone P, et al. Short-term reduction of intracortical inhibition in the human motor cortex induced by repetitive transcranial magnetic stimulation. Exp Brain Res. 2002; 147, 108-13 Di Lazzaro V, Dileone M, Profice P, et al. Direct demonstration that repetitive transcranial magnetic stimulation can enhance corticospinal excitability in stroke. Stroke. 2006;37: 2850-3 Di Lazzaro V, Dileone M, Pilato F, et al. Repetitive transcranial magnetic stimulation of the motor cortex for hemichorea. J Neurol Neurosurg Psychiatry. 2006;77:1095-7 Di Lazzaro V, Pilato F, Dileone M, et al. Modulating cortical excitability in acute stroke: a repetitive TMS study. Clin Neurophysiol. 2008;119: 715-23 Floel A, Nagorsen U, Werhahn KJ, et al. Influence of somatosensory input on motor function in patients with chronic stroke. Ann Neurol 2004; 56: 20612. Fregni F, Boggio PS, Mansur CG, et al. Transcranial direct current stimulation of the unaffected hemisphere in stroke patients. Neuroreport 2005; 16: 1551-1555 Fregni F, Boggio PS, Valle AC, et al. A sham-controlled trial of a 5-day course of repetitive transcranial magnetic stimulation of the unaffected hemisphere in stroke patients. Stroke 2006; 37: 21152122. Fridman EA, Hanakawa T, Chung M, Hummel F, Leiguarda RC, Cohen LG. Reorganization of the human ipsilesional premotor cortex after stroke. Brain 2004; 127(Pt 4), 747-58 Gilio F, Rizzo V, Siebner HR, Rothwell JC. Effects on the right motor hand-area excitability produced by low-frequency rTMS over human contralateral homologous cortex. J Physiol. 2003; 551(Pt 2), 563-73
245
Hamada M, Hanajima R, Terao Y, et al. Quadropulse stimulation is more effective than paired-pulse stimulation for plasticity induction in the human motor cortex. Clin Neurophysiol 2007; 118: 2672-2682 Hilgetag CC, Theoret H, Pascual-Leone A. Enhanced visual spatial attention ipsilateral to rTMS-induced 'virtual lesions' of human parietal cortex. Nat Neurosci. 2001;4, 953-7 Hummel F, Celnik P, Giraux P, et al. Effects of noninvasive cortical stimulation on skilled motor function in chronic stroke. Brain 2005; 128: 490-499 Hummel FC, Cohen LG. Drivers of brain plasticity. Curr Opin Neurol 2005;18:667-674. Johansen-Berg H, Rushworth MF, Bogdanovic MD, Kischka U, Wimalaratna S, Matthews PM. The role of ipsilateral premotor cortex in hand movement after stroke. Proc Natl Acad Sci U S A 2002; 99: 14518-23 Khedr EM, Ahmed MA, Fathy N, Rothwell JC. Therapeutic trial of repetitive transcranial magnetic stimulation after acute ischemic stroke. Neurology. 2005;65:466-8. Kim YH, You SH, Ko MH, et al. Repetitive transcranial magnetic stimulation-induced corticomotor excitability and associated motor skill acquisition in chronic stroke. Stroke. 2006;37:1471-6 Kinsbourne, M., Hemineglect and hemisphere rivalry. In: E.A. Weinstein and R.P. Friedland (Eds.), Hemi-inattention and Hemisphere Specialization. Advances in Neurology, Vol. 18, Raven Press, New York, 4149, 1977 Kirton A, Chen R, Friefeld S, Gunraj C, Pontigon AM, Deveber G. Contralesional repetitive transcranial magnetic stimulation for chronic hemiparesis in subcortical paediatric stroke: a randomised trial. Lancet Neurol. 2008;7:507-13. Kobayashi M, Hutchinson S, Theoret H, Schlaug G, Pascual-Leone A. Repetitive TMS of the motor cortex improves ipsilateral sequential simple finger movements. Neurology 2004; 62, 91-8 Liepert J, Hamzei F & Weiller C. Motor cortex disinhibition of the unaffected hemisphere after acute stroke. Muscle Nerve 2000; 23: 1761-1763 Liepert J, Restemeyer C, Kucinski T, Zittel S, Weiller C. Motor strokes: the lesion location determines motor excitability changes. Stroke. 2005; 36, 2648-53 Liepert J, Zittel S, Weiller C. Improvement of dexterity by single session low-frequency repetitive transcranial magnetic stimulation over the contralesional motor cortex in acute stroke: a double-blind placebo-controlled crossover trial. Restor Neurol Neurosci. 2007;25(5-6):461-5. Lomarev MP, Kim DY, Richardson SP, Voller B, Hallett M. Safety study of high-frequency transcranial magnetic stimulation in patients with chronic stroke. Clin Neurophysiol. 2007;118:2072-5 Malcolm MP, Triggs WJ, Light KE, et al. Repetitive transcranial magnetic stimulation as an adjunct to constraint-induced therapy: an exploratory randomized controlled trial. Am J Phys Med Rehabil. 2007;86:707-15. Mansur CG, Fregni F, Boggio PS, et al. A sham stimulation-controlled trial of rTMS of the unaffected hemisphere in stroke patients. Neurology 2005;64:1802-4. Martin PI, Naeser MA, Theoret H, et al. Transcranial magnetic stimulation as a complementary treatment for aphasia. Semin Speech Lang. 2004; 25,181-91
246
Monti A, Cogiamanian F, Marceglia S, et al. Improved naming after transcranial direct current stimulation in aphasia. J Neurol Neurosurg Psychiatry 2008; 79:451-453 Murase N, Duque J, Mazzocchio R, Cohen LG. Influence of interhemispheric interactions on motor function in chronic stroke. Ann Neurol. 2004; 55:400-9. Naeser MA, Martin PI, Nicholas M, et al. Improved picture naming in chronic aphasia after TMS to part of right Brocas area: an open protocol study. Brain Lang 2005; 93: 95-105. Naeser MA, Martin PI, Nicholas M, et al. Improved naming after TMS treatments in a chronic, global aphasia patientcase report. Neurocase. 2005b; 11: 182-93 Naeser MA, Martin PI, Baker EH, et al. Overt propositional speech in chronic nonfluent aphasia studied with the dynamic susceptibility contrast fMRI method. Neuroimage 2004;22:2941. Nowak DA, Grefkes C, Dafotakis M, et al. Effects of low-frequency repetitive transcranial magnetic stimulation of the contralesional primary motor cortex on movement kinematics and neural activity in subcortical stroke. Arch. Neurol. 2008; 65: 741-747 Nudo RJ. Mechanisms for recovery of motor function following cortical damage. Curr Opin Neurobiol. 2006;16, 638-44. Nudo RJ. Postinfarct cortical plasticity and behavioral recovery. Stroke. 2007;38 (2 Suppl):840-5 Oliveri M, Rossini PM, Traversa R, et al. Left frontal transcranial magnetic stimulation reduces contralesional extinction in patients with unilateral right brain damage. Brain 1999; 122 (pt 9): 173139. Oliveri M, Bisiach E, Brighina F, et al. rTMS of the unaffected hemisphere transiently reduces contralesional visuospatial hemineglect. Neurology. 2001; 57, 1338-40 Sanes JN, Donoghue JP Plasticity and primary motor cortex. Annu Rev Neurosci 2000; 23, 393415 Shindo K, Sugiyama K, Huabao L, Nishijima K, Kondo T, Izumi S. Long-term effect of lowfrequency repetitive transcranial magnetic stimulation over the unaffected posterior parietal cortex in patients with unilateral spatial neglect. J Rehabil Med. 2006; 38, 65-7 Siebner HR, Lang N, Rizzo V, et al. Preconditioning of low-frequency repetitive transcranial magnetic stimulation with transcranial direct current stimulation: evidence for homeostatic plasticity in the human motor cortex. J Neurosci. 2004;24:3379-85. Stefan K, Kunesch E, Cohen LG, Benecke R, Classen J. Induction of plasticity in the human motor cortex by paired associative stimulation. Brain 2000; 123: 572584. Sudlow CL, Warlow CP. Comparable studies of the incidence of stroke and its pathological types: results from an international collaboration. International Stroke Incidence Collaboration. Stroke 1997; 28:491-499, 1997 Takeuchi N, Chuma T, Matsuo Y, Watanabe I, Ikoma K. Repetitive transcranial magnetic stimulation of contralesional primary motor cortex improves hand function after stroke. Stroke 2005; 36: 26812686. Takeuchi N, Tada T, Toshima M, Chuma T, Matsuo Y, Ikoma K. Inhibition of the unaffected motor cortex by 1 Hz repetitive transcranical magnetic stimulation enhances motor performance and training effect of the paretic hand in patients with chronic stroke. J Rehabil Med. 2008;40:298-303.
247
Talelli P, Rothwell J. Does brain stimulation after stroke have a future? Curr Opin Neurol. 2006;19:543-50. Talelli P, Greenwood RJ, Rothwell JC. Exploring Theta Burst Stimulation as an intervention to improve motor recovery in chronic stroke. Clin Neurophysiol. 2007;118:333-42 Thickbroom GW, Byrnes ML, Edwards DJ, Mastaglia FL. Repetitive paired-pulse TMS at Iwave periodicity markedly increases corticospinal excitability: A new technique for modulating synaptic plasticity Clin Neurophysiol. 2006; 117: 6166 Uy J, Ridding MC, Hillier S, Thompson PD, Miles TS. Does induction of plastic change in motor cortex improve leg function after stroke? Neurology 2003; 14;61: 982-4. Ward NS, Brown MM, Thompson AJ, Frackowiak RS. Neural correlates of outcome after stroke: a cross-sectional fMRI study. Brain. 2003; 126: 1430-1448 Werhahn KJ, Conforto AB, Kadom N, Hallett M, Cohen LG. Contribution of the ipsilateral motor cortex to recovery after chronic stroke. Ann. Neurol. 2003; 54: 464-72 Winhuisen L, Thiel A, Schumacher B, et al. Role of the contralateral inferior frontal gyrus in recovery of language function in poststroke aphasia: a combined repetitive transcranial magnetic stimulation and positron emission tomography study. Stroke 2005; 36:1759-63. Winhuisen L, Thiel A, Schumacher B, et al. The right inferior frontal gyrus and poststroke aphasia: a follow-up investigation. Stroke. 2007;38:1286-92. Ziemann U. TMS induced plasticity in human cortex. Rev Neurosci. 2004; 15: 253-66 Ziemann U, Muellbacher W, Hallett M, Cohen LG. Modulation of practice-dependent plasticity in human motor cortex. Brain. 2001; 124: 1171-81
Chapter 15
EXTRADURAL CORTICAL STIMULATION FOR STROKE RECOVERY PART A: REVIEW OF CLINICAL STUDIES
Jeffrey A. Brown1, and Hyoung-Ihl Kim2
Department of Neurological Surgery, Wayne State University School of Medicine, Detroit, Michigan, USA; 2 Brain Function Modulation Center, Presbyterian Medical Center, Jeonju, Department of Rehabilitation Medicine and Neurosurgery, Wonkwang University School of Medicine, Iksan, South Korea.
1
INTRODUCTION
Following a stroke, the initial degree of motor impairment is the most important predictor of motor recovery (Hendricks et al 2002). Initial mild leg paresis is four times as likely to recover compared with severe leg paresis (Jorgensen et al. 1995) and better outcomes may be expected in patients with subcortical infarcts compared with cortical infarcts (Dominkus et al 1990). Preservation of both parts of the corticospinal tract and thalamic circuitry is a major determinant of the quality of hand-motor recovery following acute brain ischemia in adults. However, there is no significant difference between motor recovery in hemispheric and brain stem infarcts (Turney et al 1984). The degree of recovery is idiosyncratic, despite similar rehabilitation protocols. This may depend, among others, on inter-individual variability in vascularization (e.g. a different extent of collaterals or different blood supply from adjacent arteries) or intrinsic reorganizational capability (Rossini et al 2003; Bang et al 2008). After acute injury, marginally functional neurons may, or may not, die off. Later recovery processes involve relearning, with
Correspondence concerning this article should be addressed to: Dr. Jeffrey A. Brown, M.D., e-mail: drbrown@neurosurgeryli.net.
250
Jeffrey A. Brown and Hyoung-Ihl Kim

strengthening of existing neural pathways, and other neuroplastic structural changes (Pascual-Leone et al 2005). While early physiotherapic rehabilitation is known to contribute to regaining prestroke functions, actual recovery slowly subsides after the first post-stroke month and tapers off in the chronic stage (Bonita and Beaglehole 1998; Duncan et al 1992). Improvement after one month will most likely reflect compensatory behavior (Kreisel et al 2006). Six months following a stroke, patients do not respond so much to rehabilitative training and are left impaired (Duncan et al 2000). Integrity of motor-related areas is one of the determinants in promoting recovery. These include the intact perilesional area of the primary motor cortex (MI), the contralesional corticospinal tract, and the ipsilateral and contralateral supplementary and/or premotor areas (Nelles et al 1999; Rossini et al 2003). A high degree of interconnections among these structures is likely to compensate for the motor impairment induced by cortical infarcts. As discussed in chapter 14, noninvasive cortical stimulation can be used therapeutically to enhance recovery in chronic stroke. Additionally, some patients submitted to ECS for chronic pain reported an improvement of spasticity and dystonia that were also caused by the stroke (Canavero and Bonicalzi 2007a); ECS can also allay selected movement disorders (chapter 13 and Canavero and Bonicalzi 2007b). These studies led to direct testing of extradural cortical stimulation for stroke rehabilitation, with its superior ability to provide continuous unencumbered stimulation. A study in primates submitted to ECS and concurrent rehabilitation was also published (Plautz et al 2003). These authors created an infarct in squirrel monkeys using bipolar electrocoagulation over the cortex responsible for MI hand representation. After 3.5-5 months (chronic stroke model), spontaneous motor recovery stabilized, despite the persistence of significant motor impairment. They combined cortical stimulation and rehabilitation for 2-4 weeks. The monkeys achieved statistically significant gains in a pre-specified motor task, though not to pre-infarct levels. Cortical hand representation increased significantly and new hand representations developed adjacent to and at considerable distance from the infarct. Improvement was written up to, among others, enhanced synaptic function, enhanced neuronal connections and long-term cortical synaptic potentiation, mainly in cortical layer V.
CLINICAL STUDIES OF EXTRADURAL CORTICAL STIMULATION

Drawing from these foundations, direct stimulation of the cortex via implanted extradural electrodes (ECS) has been simultaneously attempted for the first time in 2002 by Canavero (Canavero et al 2006) and Brown (Brown et al. 2003), and then pursued by others (Brown 2006; Brown et al 2006; Huang et al 2008; Kim et al 2008; Levy et al 2008), as an adjunct to standard rehabilitation (Table I). While Canaveros group tested ECS in a stroke patient for the first time in 2002 on the basis of evidence presented in chapter 14, simultaneously a US industry-led trial (NorthStar Neuroscience) was initiated to test the hypothesis that subthreshold cortical stimulation of the region of MI that produces residual motion in an impaired limb after a non-hemorrhagic cortical or subcortical infarction enhances motor recovery. The first successful patient was a 65 year-old man with a subcortical ischemic infarct and right spastic hemiparesis occurring 19 months before treatment. He underwent subthreshold MI ECS delivered concurrently with
Extradural Cortical Stimulation for Stroke Recovery
251
three weeks of structured rehabilitation. Before stimulation the patients affected arm rested in a flexion posture without the ability to voluntarily flex or extend the fingers. After stimulation combined with rehabilitation, he was able to grasp a pen and to write letters. This improvement persisted for at least four weeks after conclusion of the rehabilitation therapy (Brown et al 2003). Table 1. Studies of Cortical Stimulation (CS) for Chronic Stroke Rehabilitation
Authors Year N Infarct Location Aim of CS Device -Upper arm motor recovery -Grid + EPG -Motor and language -Double Strip + IPG -Upper arm motor recovery -Paddle + IPG Target structures Duration of CS Parameters Of CS -Effect of CS -Complications
Brown et al (2003)
SCI
Ipsi MI
3 weeks
4,5mA, 50Hz, 100 sec stimulation during rehabilitative training Continuous, 3-3,7 V, 50 or 130 Hz, 210sec
Canavero et al (2006)
CI
Ipsi and Contral. MI
7 months
-UEFM improved by 10 points, Hand dexterity -none -Modest improvement of motor & aphasia -none
Brown et al. (2006)
SCI or CI
Perirolandic fMRIactivated area
3 weeks
Kim et al (2008).
SCI & CI
-Motor & language -Paddle + IPG
Ipsi MI+PMC & Brocas area
6 months
Bipolar stimulation during rehabilitative training Half-threshold or 6.5 mA, 50Hz, 250sec Unipolar & Continuous 5V, 50Hz, 200sec
-UEFM from 35.37.8 to 45.5 10,8 -2 infections
Levy et al (2008) Huang et al. (2008)
1 2
SCI or CI
-Motor recovery -Paddle + IPG
Perirolandic fMRIactivated area
6 weeks
Bipolar stimulation + rehabilitative training (2.5 hours daily); Half-threshold or 6.5 mA, 50 or 101 Hz, 250sec. 3 anodes + 3 cathodes (1.8 cm2 stimulation area)
-UEFM from 8 to 27 in one: 35 to 42 in the other) Marked language improvement -none -UEFM: improved by 5.54.4 AMAT: improved by 0.4 0.6 -1 seizure
*SCI: subcortical infarct, CI: cortical infarct, Ipsi MI: ipsilesional motor cortex, Cont MI: contralesional MI, PMC: premotor cortex, IPG: internal pulse generator, EPG: external pulse generator, PD: pulse duration, UEFM: upper extremity Fugl Meyer, N: number of patients.
Brown et al (2006) further published a prospective, randomized, multi-center safety study of sub-threshold motor cortical electrical stimulation of patients with motor deficits
252
resulting from a stroke that occurred at least 4 months prior to enrollment. They randomized patients into two groups. The treatment group had an electrode implanted, then received epidural electrical stimulation (at 50 Hz, 50% of the current needed to evoke gross motor movement during daily rehabilitation throughout their three weeks of rehabilitation, or 6.5 mA if no movement occurred). A second group, the control group, underwent the same threeweek period of rehabilitation, but did not have an electrode implanted. The mean time interval after the stroke was 18 +/- 18 months (range 9-33) in the surgical group and 38+/- 24 months (range 15-68) in the control group. The eight patients who completed surgical implantation and rehabilitation improved significantly better than controls in their upper extremity Fugl-Meyer (UEFM) score (P=0.003 overall) and in the hand function score of the Stroke Impact Scale (P=0.001 overall). In an intention to treat analysis, scores were also analyzed using all available, serially collected Fugl-Meyer data for all 10 patients entered into the study (including two who dropped out of the study due to infection). The difference between treatment groups remained significant (p =0.027). Improvements persisted throughout the 12-week follow-up assessment (study week 16). In comparison, lesser improvements in control patients occurred within the first two weeks and then seemed to decrease over time. Levy et al. (2008) completed a larger multicenter study of 24 patients randomized between stimulation and six weeks of rehabilitation or rehabilitation alone, again under the sponsorship of NorthStar Neuroscience. The mean interval from the time of stroke until treatment in this study was 2 years (range 4 months to 8 years). The mean patient age was 57 years (range 26-81 years). Rehabilitation focused on upper extremity tasks, especially for the hand and wrist. These tasks included increasing range of motion and optimizing self-care tasks. The AMAT, a measure of activities of daily living focusing on upper extremity paresis, increased by 0.4 points compared to control patients whose score increased by 0.2 points (p=0.2). In other words, 67% of patients in the treatment group had clinically meaningful improvement in the UEFM (equal to or more than 3.5- point increase); 25% of control patients improved this much. Half of treatment patients had clinical improvement in both UEFM and AMAT scores (including 0.21 point improvement on AMAT) compared to 8% of patients in the control group. Patients in the treatment arm began at a baseline UEFM of 35 points. This indicates moderate to severe impairment. These patients improved by 5.5 points (range 0-17). A two-point improvement is equal to an improvement from no activity to full activity. This implies that even a small improvement in UEFM score can be significant functionally. Control patients improved also, but only from a baseline of 32.4 to 34.3 points, a significant difference (p=0.03). Thus, improvement after ECS is observed long after further recovery is thought to be no longer possible. Interestingly, patients with left-side stroke who were stimulated showed more language improvement than controls, as measured by the Wechsler Abbreviated Scale of Intelligence vocabulary t-score (1.3 versus control score -3.8, p=0.05).
253
SUGGESTED MECHANISMS OF ACTION OF MOTOR RECOVERY

Motor functional improvement must derive from more than just direct enhancement of surrounding marginally functional cortical neurons. Likewise, improvement would not occur only as an indirect result of electrical inhibition of confounding regions of hyperactivity. Even at the time of initial surgery, during intra operative cortical mapping, observable finger contractions could be stimulated. These movements were not possible voluntarily before stimulation. Thus, stimulation may immediately enhance the plasticity of marginally effective circuits, leading to improved voluntary function. Rehabilitation with stimulation may cement these improvements by causing persistent anatomical changes, also explaining persistence of functional improvement after withdrawal of cortical stimulation. A summary of possible mechanisms follows:
1-Reactivation of Plasticity This is likely responsible for the long (months) after-effect of ECS (Ward 2005; Nudo 2006). The study by Canavero et al (2006) suggests that GABA, an inhibitory neurotransmitter, is involved in releasing plasticity after surgical CS. There is also evidence for stroke-induced neurogenesis in the human brain (Jin et al 2006).
2-Local Enhancement of Perilesional Areas Patients who do recover usually rely on adaptations of perilesional areas, rather than remote areas of CNS. Although many areas of the brain show activation (Rossini et al 2003), the ipsilateral premotor cortex (PMC) appears to play a key role in the reorganization of sensorimotor pathways (Cramer et al 2005). PMC has a direct connection with spinal motor neurons, although of lesser magnitude than the primary motor pathways (Cao et al 1998): unmasking of this latent pathway may form a new motor pathway. Accordingly, this area has been selected as a target of ECS in subcortical infarcts (Canavero et al 2006; Kim et al 2008). New cortico-cortical connections mold a new local network by incorporating perilesional areas.
3-Enhancement of Network Function If the cortex is diffusely necrotized, perilesional networking would not be enough. Recruitment of distant cortices and corticofugal fibers in ipsi- or contralesional hemispheres are required for mutual interconnections. However, it remains unclear if the areas of fMRI activation are really a specific reorganization of motor representation or just a by-process unrelated to actual recovery. If the infarct area is large enough, distant areas will be further recruited, e.g. the parietal lobe. The parietal lobe is significantly activated in stroke patients compared to healthy controls. It is reciprocally connected with PMC, thus forming a
254
specialized network. Most parietal input to PMC originates from the superior parietal lobe, which is also known to play an important role in attentional tasks, visual vigilance, and spatial cognition. On the other hand, the inferior parietal lobe directly projects to the supplementary motor areas. The inferior parietal lobe is bilaterally activated in stroke patients versus normal controls, which indicates that this area may play an important role in the reorganization of sensory and motor systems leading to improvement of neurological deficit. ECS may thus achieve network stimulation (Kim et al and Canavero S, unpublished observations, and Canavero et al 2006). Combined stimulation of the perilesional and parietal lobe should be considered, for it may form a new and potential network to enhance motor recovery.
4-Amplification of Sensory Input It is known that target-specific training contributes to the recovery by increasing the magnitude of user-dependent plasticity, even in chronic stroke patients. Rehabilitative training modulates the somatosensory input, which not only influences the motor output, but also motor learning (Hummel and Cohen 2005, Celnik et al 2007). Enhancement of somatosensory input leads to improvement of motor performance in stroke patients. MI and SI display a striking capacity to reorganize and are anatomically and functionally highly interconnected. Cortical stimulation may enhance somatosensory input, leading to improved recovery. Direct electrical stimulation of the somatosensory cortex (SI) may be used as an amplifier to substitute for peripheral electrical stimulation in terms of strengthening the sensory input (Wolters et al 2005, Leal-Campanario et al 2006, Houweling and Brecht 2008).
MECHANISM OF APHASIA RECOVERY

The recovery of language is basically similar to that of motor recovery. Functional imaging studies show recruitment of perilesional tissue adjacent to Brocas or Wernickes areas, as well as involvement of the homologous language area in the contralateral hemisphere in chronic stroke patients. In the acute phase (first days) after a stroke, early weak activation of left Brocas area has been observed (Saur et al 2006). In the subacute stage (about 2 weeks following a stroke), fMRI revealed strong activation bilaterally in the language network, with peak activation in the right Broca-homologue. In the chronic stage (months after the stroke), fMRI activation was normalized and peak activation returned to the left hemisphere. Moreover, this normalization was associated with further significant improvement of language impairment. These observations indicate that recruitment of other language areas is essential during the recovery process. Meinzer et al (2008) found that language improvement after short-term intensive language training was significantly related to increased functional activity (fMRI) in the perilesional areas of the dominant hemisphere in chronic aphasia. Improvement was thus mediated by reintegration of, or increased connectivity within, these perilesional areas. Rehabilitative training may increase activity in perilesional and other remote areas. Still, the increased activity in these areas is unrelated to
255
improved language function; rather, it reflects activity subserving semantic processes like memory and attention, possibly mediated by a spread of perilesional activation to the related areas of both hemispheres. Therefore, the perilesional area adjacent to Brocas area is the single most important area for the recovery of expressive aphasia and seems to be a good target for plasticity induction by means of ECS (Kim et al. 2008). Intriguingly, there is growing evidence suggesting connections between the cortical hand motor area of the dominant hemisphere and regions subserving language: reading aloud provokes increased excitability in the hand motor area of the dominant hemisphere, or in the leg area of the non-dominant hemisphere in patients versus healthy subjects. A TMS study showed that the motor hand area is activated by language comprehension (Floel et al 2003). These findings support the notion that there is a functional connectivity between language areas (including Brocas and Wernickes) and regions mediating hand motor function. In stroke patients with aphasia, reading aloud enhanced the excitability of the right hand MI (Meister et al 2006). It remains to be clarified whether stimulation of the perilesional area adjacent to Brocas area may enhance cortical hand function or vice versa (Kim et al, unpublished observations), but stimulation outside the language areas can improve aphasia (Canavero et al 2006).
PART B: SURGICAL TECHNIQUE

Hyoung-Ihl Kim, Yong-Il Shin and Sung-Keun Moon
Brain Function Modulation Center, Presbyterian Medical Center, Jeonju, Department of Rehabilitation Medicine and Neurosurgery. Wonkwang University School of Medicine, Iksan, South Korea.
PREOPERATIVE EVALUATION
Candidates for ECS are chronic stroke patients who do not show further clinical improvements six months following a stroke. Some evidence suggests that ECS may be more helpful during the initial period: elapsed time between a stroke and rehabilitation is an important factor that influences outcome, but this still awaits validation. Although most patients reported in the literature underwent ECS no later than 3 years after stroke, at least one was 8 years after infarct. Perhaps, there is no upper limit to duration of chronic stroke for ECS. Currently, there is no way to tease out treatment responders preoperatively. Patients should be cognitively competent and motivated to adhere to long-term rehabilitation. Exclusion criteria are medical comorbidities that contraindicate rehabilitation or general anesthesia, epilepsy, other neurological diseases and compound strokes. Patients are assessed preoperatively with one or more scales (Table 2). Grip strength is a basic test that is often checked during the course of ECS. The UEFM scale does not reliably assess the speed of movement and the functional role of the upper extremity, but is useful for assessing the minima of arm function (floor effect). The JTHFT scale is only useful to assess higher-functioning patients that demonstrate good dexterity (ceiling effect). The LEFM scale does not appropriately represent the improvement of locomotor function as the UEFM scale does in evaluating upper extremity function. It is important to select the appropriate measurement depending upon the examinees locomotor ability. The LEFM and BBS scales have high ceiling effects in highly-performing subjects; conversely, walking speed may have
Correspondence concerning this article should be addressed to: Dr. Hyoung-Ihl Kim, M.D. Ph.D, e-mail: hyoungihl@hotmail.com.
258
a floor effect in stroke patients that walk slowly or require assistance. For example, if the patient regained walking ability (2/66) from lying down (1/66), the difference would appear to be just one point. Other tests should be developed to appropriately represent the improvement of locomotor function while avoiding the ceiling or floor effect outcomes. To assess a patients language function, methods must be simple and brief. Also, each test requires adjustments depending on the patients deficit. Most language-screening tests contain benchmarks of comprehension, expression, communication ability, reading, attention, and others. The most commonly used language screening tests include the AASP, the FAST and the MAST (Salter et al 2006). Selection of the test also depends on the language therapist that examines the patient. Table 2. Stroke outcome scales
General Stroke Scale National Institute of Health Stroke Scale (NIHSS) Modified Barthel index (BI) Functional Independence Measure (FIM) Mental Test (MT) Grip Strength (GS) Upper Extremity Fugl-Meyer Scale (UEFM) Wolf motor function test (WMFT) Arm Motor Ability Test (AMAT) Jebsen-Taylor hand function test (JTHFT) Gait Speed (GaS) Timed up and go (TUG) Lower Extremity Fugl-Meyer Scale (LEFM) Bergman Balance Scale (BBS) Acute Aphasia Screening Protocol (AASP) French Aphasia Screening Test (FAST) Mississippi Aphasia Screening Test (MAST)
Upper Extremity Motor impairment
Lower Extremity Motor Impairment
Language disorder
OPERATIVE MANAGEMENT
1A-Choice of Target for Motor Recovery The choice of the appropriate site for stimulation is key, as the electrode paddles usually employed are not large enough to cover extensive target areas singly. Neuroimaging may help in this regard. In hemiplegic stroke patients with a single subcortical infarct, significant increases in blood flow are observed in the inferior parietal cortex on both sides, the contralateral sensorimotor cortex, the ipsilateral prefrontal cortex, supplementary areas, and the cingulate cortex (Nelles et al 1999). In subcortical infarcts, asymmetrical bilateral activation of motor pathways, recruitment of additional sensory and secondary motor areas and extension of the sensorimotor cortex toward the face area are observed (Calautti and Baron 2003). In patients with cortical infarcts, a similar pattern of activation is seen, and yet peri-infarct activation and ipsilesional premotor cortex activation are prominent (Kwon et al
259
2007). Importantly, there is no significant correlation between volume of activation in the ipsilateral sensory motor cortex and residual motor deficit (Cao et al 1998) and, before any of these areas can be proposed as a target, metabolic changes in these cortical areas must be correlated with behavioral recovery. Although fMRI shows multiple areas of activation in chronic stroke patients, this does not imply a direct relation with functional improvement. At this time, the premotor cortex (PMC) can be suggested as a target for ECS, as well as MI and SI, especially if fMRI shows motor-task related activations (Canavero et al 2006, Brown et al 2006, Levy et al 2008). Several caveats are in order. Peri-infarct activation is not directly correlated with the magnitude of motor recovery (Cramer et al 2006). F-MRI motortask activations are usually bilateral in chronic stroke and surgical outcome does not correlate with neuroimaging results (Calautti and Baron 2003, Kim et al 2008). In other words, presence or absence of fMRI activation is not directly related to surgical outcome. In our experience, selection of target based on the results of fMR may be confusing. Moreover, perirolandic cortex stimulation produces variable results in patients with apparently similar neurologic conditions. Finally, it is compulsory to map the individual patients brain rather than rely on population averages when targeting for specific body segments, given interindividual variability (Cramer et al 2003). Clinical Cases A- A 51- year-old man with a chronic subcortical infarct 8 months before and moderately spastic arm and clumsy hand underwent ECS of the premotor and primary motor cortex. fMRI images on a wrist flexion extension task of the paretic arm failed to show activation of MI in both hemispheres. Six months stimulation with rehabilitative training improved the mobility and spasticity of his hands, thus enabling him to drive a car again. In addition, his Functional Independence Measure (FIM) score improved in self-care, mobility, and locomotion (Figure 1).
Figure 1. Chronic subcortical infarct in the right internal capsule. Diffusion imaging of MRI demonstrates the lesion in the right internal capsular area (A). This patients had left hemiparesis with spastic upper extremity. A paddle electrode was implanted to cover premotor and hand/arm MI (B). After 6 months stimulation and rehabilitative training, the patient can drive a car with improved FIM scores.
260
Hyoung-Ihl Kim, Yong-Il Shin and Sung-Keun Moon B- A 47-year-old patient with a chronic subcortical infarct showed similar neurological deficits. fMRI during the same task showed no activation of MI in the contralateral hemisphere. A paddle electrode was implanted in the same manner, and stimulation with concurrent rehabilitation was continued for 6 months. However, no objective meaningful improvement was observed in this case (Figure 2).
Figure 2. Chronic subcortical infarct in the left internal capsule. Diffusion imaging of MRI demonstrates the lesion in the left internal capsular area (A). The patient had right hemiparesis and spastic arm. fMRI on a wrist flexion extension task shows activation of the unaffected MI (C), but no activation of the affected MI (D). A paddle electrode was implanted to cover premotor and hand MI cortex (B). After stimulation and rehabilitation for 6 months, the patient was not improved, except in subjective mood.
C- A 61 year-old woman with right hemiplegia and aphasia showed a cortical sylvian infarct. Diffusion Tensor Tractography (DTT) showed complete disruption of the motor fibers in the left hemisphere. ECS along with rehabilitative training did not produce any improvement, except some stabilization of her mood (Figure 3). D- A 39-year old patient with a cortical infarct 18 months earlier presented severe dysphagia and hemiparesis. Two 4x4 cm paddle electrodes were implanted to cover Brocas area plus premotor and motor cortices. After 6 months of stimulation (1 to 5V over 1 month, 200 s, 50 Hz) and rehabilitation (two hours every day), both hemiparesis and language impairment were markedly improved (Figure 4).
261
Figure 3. Chronic cortical infarct in the left middle cerebral artery territory. T2-weighted image shows the cortical infarct in the left MCA territory (A). The patient showed right hemiplegia and aphasia. Diffusion Tensor Tractography shows complete disruption of left-sided motor fibers (B). A paddle electrode was implanted to cover the premotor and MI cortex, a little above the hand area (C). 6 months stimulation plus rehabilitation was fruitless.
Figure 4. Chronic cortical infarct in the left middle cerebral artery territory. T1-wighted image shows the cortical infarct in the left middle cerebral artery territory (A). The patient showed right hemiplegia and aphasia. Two electrodes were implanted to cover Brocas area, and premotor and MI cortex (B). After ECS and rehabilitation for 6 months, the patient regained communication ability and improved motor function of the right limbs.
These results indicate that ECS centered on the premotor cortex does not guarantee clinical improvement, even for small subcortical infarcts. Lack of motor-task activations on fMRI cannot be relied upon for selection of patients: patients A and B showed no activation on fMRI, yet one patient showed no clinical improvement, while the latter was markedly improved. It is fair to conclude that, while functional data can demonstrate the process of functional reorganization, they cannot be used to make a definitive surgical decision. Likewise, results of ECS in chronic stroke patients with cortical infarcts may show variable results.
262
Thus, the perirolandic cortex does not qualify as a universal target for all patients, despite similar presentation of imaging and clinical findings. The unaffected hemisphere is also known to be associated with recovery of motor function in the paretic extremities (Marshall et al 2000, Carey et al 2002). Increased activation of the sensorimotor cortex in the unaffected hemisphere correlates with recovery of the paretic hand, yet this gradually subsides in three to six months, while the intensity of overactivity in the unaffected hemisphere is unrelated to functional recovery (Loubinoux et al 2003). Thus, activation in the unaffected hemisphere does not seem to be a good indicator for ECS. Canavero reported that ECS with facilitatory parameters in the unaffected hemisphere produced mixed effects, both positive and negative (Canavero et al 2006), pointing to a complex role of the unaffected hemisphere in motor recovery.
1B-Choice of Target for Aphasia Recovery Targets different depending on the type of aphasia (expressive vs receptive) and the extent of the infarct. There is evidences that ECS over the perilesional area in Brocas area can promote language recovery (Canavero et al 2006, Kim et al 2008). The perilesional region near Wernickes area is a potential target for receptive aphasia, but no published experience has accrued so far.
2-Single Versus Multiple Stimulation At this time, no universally effective target can be offered. Single site stimulation has been generally used to cover the most commonly activated area, usually the contralateral sensorimotor or premotor cortex. These areas are supposed to be mainly responsible for motor planning, preparation, and execution of the movement. If there is anatomical damage to these areas, remaining intact parts in MI or premotor areas are recruited to restore function. Contralateral sensorimotor cortex and peri-infarct areas of activation, though mostly the premotor cortex, are proven useful targets for ECS in chronic subcortical stroke and in small cortical infarcts (Cramer et al 2005, Canavero et al 2006, Kim et al 2008). Nonetheless, premotor cortex or MI/SI stimulation often is ineffective for cortical infarcts in which corticocortical connections are widely disrupted. Combined stimulation including premotor + sensory cortex (or premotor) +parietal cortices may be beneficial for improving the motor outcome (unpublished result and Canavero et al 2006). As the electrode cannot cover the entire brain, the number and placement of electrodes should be modified depending on the goal of treatment. For instance, a patient with complete hemiplegia may wish to regain locomotor ability rather than fine hand motricity. In this situation, the electrode(s) may be placed more superiorly to cover leg representation areas in MI or peri-infarct areas adjacent to leg function. Bilateral parietal cortex and premotor cortices are activated in motor task fMRI protocols. These areas may be recruited together for functional reorganization in patients with larger infarcts. Significant motor recovery was found to be associated with the activation of
263
the superior posterior parietal lobe and premotor cortex in PET studies, indicating that a fronto-parietal network is mediating the recovery (Struppler et al. 2007). Most parietal input to premotor cortex and MI arise from the superior parietal lobule and posterior parietal cortex, forming the fronto-parietal visuomotor controll network (Wise et al 1997). Clinical Case E- A 67-year-old patient with massive cerebral infarct developed hemiplegia. He could only stand with maximal assistance, otherwise he was bed-ridden. fMRI on flexionextension of the knee showed diffuse activation in posterior parietal areas. Two paddle electrodes were implanted to cover the frontal (near supplementary motor cortex) and posterior parietal area (network stimulation) in order to achieve maximal coverage. After stimulation and rehabilitative training for 6 months, he began to stand independently and walk 200 m (Figure 5).
Figure 5. Massive cortical infarct in the right hemisphere. T1-weighted image shows a diffuse frontoparietal cortical infarct (A). The patient was not able to stand, despite long-term intensive rehabilitation. Two epidural electrodes were implanted extradurally on frontal and posterior parietal areas (B). After 6 months stimulation and rehabilitation, he was able to stand independently and walk long distance (with permission from Elsevier).
3-Parameters of Stimulation Given the limited experience, optimal parameters remain unknown. Since it takes several weeks to months to observe clinical recovery, choice of parameters cannot be based on acute clinical effects. Also, inter-individual variability makes it difficult to determine uniformly optimal parameters. If a motor response can be induced by cortical stimulation, an amplitude one-half or one-fourth of the threshold amplitude can be selected. However, if the cortical infarct involves the motor strip, motor contractions are unelicitable and amplitude must be decided empirically (6.5 mA in Brown et al 2006 and Levy et al 2008). We generally employ continuous-mode stimulation at 5 volt, 50Hz, and 200s of pulse duration. Canavero et al (2006) tried both low and high frequency stimulation and found mixed effects. We use anodal and unipolar stimulation, in order to reach deeper regions, in particular deeper leg
264
representation areas in the inter-hemispheric fissure. One important caveat is that brain atrophy is often observed among stroke patients and this must be taken into account when setting amplitude.
4-Duration of Stimulation Brown et al (2006) reported that three weeks of stimulation were enough to achieve improvements of upper Fugl-Meyer and hand function scores on the Stroke Impact Scale in chronic stroke patients. Levy et al (2008) stimulated for 6 weeks. On the other hand, we maintained cortical stimulation for six months (Kim et al. 2008), while Canavero et al (2006) performed four months of cortical stimulation in the affected hemisphere with positive gains; on the contrary, 3 months of ECS in the unaffected hemisphere gave mixed effects. Longer stimulation (months rather than weeks) appears more beneficial in our experience: some patients began to show improvement after four months of stimulation, with recovery continuing even after removal of the stimulation devices. Cortical stimulation is usually provided during the rehabilitative training, as it is believed to support brain plasticity and facilitate the learning process. Thus, about three hours of stimulation with concurrent rehabilitative training has been reported (Brown et al 2006, Levy et al 2008). Alternatively, continuous stimulation has been employed and found beneficial, despite possible neuronal fatigue (Canavero et al 2006, Kim et al 2008). Data indicate that two hours of stimulation every twelve hours may achieve the same effects of continuous stimulation (unpublished data). We think that the duration of stimulation must be individualized, depending on the extent and location of the stroke. For larger infarcts, longer durations of stimulation may be required.
5-Technique of Implantation A-Types of Electrodes The strip electrode as usually employed for ECS is adequate to recruit cortical areas involved in functional reorganization (Canavero et al 2006). Wide paddle electrodes are available and may provide a wider stimulation coverage. However, approximation of two or more strip electrodes is equally effective (Figure 6). B-Preoperative Mapping The reader should refer to chapters 1 and 2 for details. Cortical mapping is possible by identifying motor task-activated areas on fMRI. Anatomo-functional imaging that shows task-related activations and the extent of cortical infarcts provides useful clues as to the relationship between pathological nonfunctioning and intact functioning cortex, as somatotopic organization of motor function is often altered by cortical damage: motor taskactivated areas on fMRI shift to the remaining intact cortex. Identification of the intraparietal sulcus is important when planning fronto-parietal network stimulation. Similar to the junction between the precentral sulcus and superior and
265
inferior frontal sulci in the frontal lobe, the post central sulcus is joined by the horizontallyrunning intraparietal sulcus. This sulcus divides the parietal lobe into the superior and inferior parietal lobes. However, it is not easy to identify the sulci without the aid of neuronavigation and curvilinear plus multiplanar imaging. In the end, the neurosurgeons judgment is required to convert neuroimages to real-life surgery. If a wide electrode is used, simple anatomical landmarks (using MRI combined with fMRI) can be employed to determine the target, without the use of neuronavigation.
Figure 6. Electrodes for extradural stimulation. A: strip electrode (Resume, Medtronic Inc, MN,USA), B: paddle electrode for bipolar stimulation (NorthStar Neuroscience, Wa, USA), C: paddle electrode for unipolar stimulation (Cerastim, Oscor Inc, Fla, USA).
C-Craniotomy and Electrode Implantation Implantation of the electrode can be performed under local or general anesthesia. One or two burr holes are enough when inserting the strip electrode. Extradural placement of electrodes is safer and less invasive than intradural implantation (see chapter 2 and 3 for details). A burr hole is made based on the length of the strip electrode and the direction of placement over the cortex. Strip electrodes are usually positioned to cover the premotor and motor cortex at the level of the fMRI hot spot. A burr hole is fashioned immediately anterior to MI. If motor mapping using electrical stimulation and implantation of a wide electrode are simultaneously performed, a small (3-5 cm) flap craniotomy is needed. Intraoperative stimulation of MI may not always evoke muscle twitches, depending on damage to the corticospinal tract. If a 3cm wide electrode is epidurally placed to cover the premotor and part of motor cortex areas, the anterior edge of the electrode can be placed just behind the coronal suture, and the upper and lower edges can be adjusted depending on the activation pattern on fMRI. If the electrode is positioned on the inferior frontal (Brocas) or other cortical areas, a similar procedure is employed. In case of a flap, electrodes need to be tightly anchored to increase the contact, to keep blood out of the space between electrode and dura, and to prevent the migration of the electrode. Seepage of blood into the space between the dura and the electrode may hinder delivery of the electrical stimulation and therefore decrease the efficacy of treatment. The electrode is connected with a lead, which is
266
subcutaneously tunneled and connected to the neural stimulator, which is usually implanted subclavicularly in the anterior chest or axillary area.
POSTOPERATIVE MANAGEMENT
Antibiotics are administered according to local policy. Two or three days following the procedure, the patient may be discharged, but is rescheduled for a parameters-setting session. Initially, the threshold for eliciting a motor response in the contralateral extremity is determined by trains of stimulation (3 seconds of a train at 50 Hz, 100-250s, starting from 0.5 V). Amplitude is gradually increased until movement is observed or up to 10V. If the patient showed a positive response to movement during intraoperative testing, movement of the contralateral extremity can be expected; otherwise, even high voltages will be useless. Rehabilitation is usually started two weeks after the placement of electrode. Current is gradually increased up to half threshold for four weeks (up to 4 -5 volts), while searching for side effects of stimulation. Stimulation for aphasia does not elicit motor responses and thus is ramped up to 5V only, to avoid possible side effects. The dose of stimulation can be modified depending on the severity of the neurological deficit. If the patient has a mild neurological deficit, we elect to provide two hours of stimulation every twelve hours. Continuous stimulation is indicated in severely compromised brains. It is important to administer rehabilitation while the stimulator is on, for synergy to occur. We and Canavero make a strong case for longer periods of stimulation, i.e. 4-6 months (Canavero et al 2006, Kim et al 2008) versus 4-6 weeks reported in other series (Brown et al 2006, Levy et al 2008). No noticeable stimulationrelated complications such as seizures, abnormal behavior, or decline of neurological functions have been reported up to now, with the exclusion of one seizure in the trial of Levy et al (2008). Rehabilitation is usually performed using peripheral parts of the body. Use of the paretic limb is a major modulator of plasticity in intact areas of the brain adjacent to the cortical infarct (Use-dependent Plasticity -UDP) (Kreisel et al 2007). Rehabilitative methods based on UDP include Constraint-Induced Movement Therapy (CIMT), Robot-Assisted Motor Retraining (RAMR) or Interactive Robot Therapy (IRT), Functional Electrical Stimulation (FESS), Partial Body Weight Support Treadmill Training (PBWSTT), as well as other conventional physiotherapy and occupational therapies. Reinforcement of Somatosensory Input (RSI) is another useful technique (Wu et al 2005, Sawaki et al 2006). Repetitive practice is the main stimulus in UDP, while stimulation of the sensory system is the main stimulus in RSI. RSI includes Somatosensory Stimulation of the Median Nerve or Hands, and the Rood technique. Peripheral mechanisms include strengthening of muscle power by applying loads on supportive muscles to maintain balance and to execute locomotion and other meaningful movements of the extremities. Most conventional physiotherapic and occupational therapies can be included in this class. Unfortunately, given the limited scientific evidence, it remains unclear as to how much rehabilitation actually contributes to post-stroke improvement. Moreover, it is almost impossible to compare rehabilitative methods and establish superiority, with the possible exception of goal-specific rehabilitative training. Thus, no indication can be offered as to which rehabilitative technique is best
267
associated to ECS, except to underscore the need of individualization of the rehabilitative process to the patients physical and cognitive disabilities. Katayama et al (1994) described a post-bulbar stroke patient with no analgesia but marked improvement in his motor weakness and Katayama et al (1998) found that MI ECS improved hemiparesis in 19% of stroke patients submitted to MI ECS for pain. Other authors had similar experiences. Thus, ECS per se may partially contribute to overall improvement.
OUTCOME OF ECS AND PREDICTORS OF GOOD OUTCOME

Prognostic signs for post-stroke recovery include initial degree of motor impairment (less is better), initiation of active movements in distal joints (i.e. finger and wrist), sensory deficit (poor prognostic sign), integrity of MI and the pyramidal tract, site of stroke (subcortical better than cortical), presence (favorable sign) or absence (unfavorable sign) of motor-evoked potentials (MEPs) in the affected limb, MI laterality index and degree of hand motor cortex activation (Dominkus et al 1990, Duncan et al 1992, Jorgensen et al 1995, Feys et al 1998, Hendricks et al 2002, Kwakkel et al 2003, Loubinoux 2007). Recovery of function after a stroke is also better for proximal arm function than distal. Mehta and colleagues were able to establish a motor threshold in about 16 percent of their investigational patients: these achieved statistically better outcomes compared to the whole cohort of patients, with results persisting through the 6-month follow-up (Dr. Nawzer Mehta, NorthStar Neuroscience 2008, personal communication). This finding dovetails with TMS studies (Feys et al 1998). Unfortunately, TMS data is too rough a guide for accurate selection of target, but may be predict outcome. Single stimulation of the perilesional area has been the major option in ECS. This strategy may work if the cortical infarct is not extensive or limited to the subcortical area. If the infarct area is large or involves cortical and subcortical areas simultaneously, single, perilesional stimulation is not likely to improve the outcome. ECS of a larger area, or even network stimulation, is required (see clinical case E). It is also important to determine the level of excitability of the cortex ( hypo- or hyper), in order to optimize parameters of stimulation. Duration of ECS and concurrent rehabilitation may be another factor bearing on outcome. In our experience, 3-6 months of ECS is required to ensure full treatment (see also Canavero et al 2006). Merely extending the treatment period is not expected to work in all cases: other factors (e.g. interval between stroke and ECS, type of rehabilitation, sex, age) may play an undefined role. Rehabilitative training is an integral part of post-stroke recovery, with focus on taskoriented training. ECS plus rehabilitation is said to promote recovery at a level above what is observed after CIMT (Levy et al 2008). Sensorimotor Stimulation (Tarkka et al 2008) is another technique that should be assessed along with ECS.
268
PROSPECTS
Although investigational, ECS may be indicated for chronic stroke patients whose neurological deficit has not improved after initial rehabilitative training. It led to 5- 40% improvement of the original neurological deficit and activities of daily living in successful cases (unpublished data), with benefit persisting > 6 months in almost all cases. However, a pall has been cast on the whole technique following the negative results of the large pivotal study (Harvey et al 2009, Levy et al 2008b). At the 4-week follow-up, 30.8% of cortical or capsular ischemic stroke patients receiving ECS achieved the threshold of clinically meaningful improvement on the composite primary end point (UEFM plus AMAT) versus 29.1% of controls (= 1.7, P 0.41= NS; preset primary end point: 20% absolute difference), in disagreement with favorable animal test studies (Plautz et al 2003). The study enrolled stroke patients with moderate to moderately severe motor impairment (UEFM range: 20-50 points) at least 4 months after hemiparetic stroke. According to Canavero (personal communication, 2008), the failure of this trial is likely due to too short rehabilitation and faulty inclusion criteria. For instance, patients who lacked significant cortical activation in the affected perirolandic region or with too diffuse a pattern of activation on a motor task (index-finger tapping, simultaneous tapping of 4 fingers or wrist extension) were excluded from implantation. Patients were only implanted on the affected MI/SI. A fixed frequency was set for all patients. As shown by available data (Canavero et al 2006, Kim et al 2008), in several patients other cortical areas must be included, such as the premotor cortex and posterior parietal areas, even bilaterally or contralaterally tout court. A TMS study found that bilateral rTMS exceeded the algebraic sum of the separate stimulation of each MI (Strens et al 2003). Importantly, MI/SI hot spots may not portend successful implantation: data suggest that patients presenting near normal activation would not be good candidates, as maximal usage implies no further reserve (Loubinoux 2007). In this sense, a low baseline cortical activity may represent underuse of surviving cortical resources (but also damage with little available resource to activate or enhance). In many patients, intraoperative stimulation elicited no motor response (Levy et al 2008), despite fMR hot spots. These patients should have been implanted elsewhere. Dual or triple stimulation of functionallyrelated areas may salvage patients that might not benefit from single stimulation. There is also some initial attempt to enhance ECS effects with concurrent drug administration (Canavero S, personal communication, 2008). rTMS tuned to individual alpha EEG frequency (10-12 Hz + 1 Hz) enhanced cognitive performance (Klimesh et al 2003), and a similar protocol might be adapted to ECS, as suggested by Canavero. General anesthesia may be risky in older patients with associated comorbidities; neuroleptoanalgesia is an option, particularly with non-flap techniques. Subdural placement would be indicated in severely atrophic brains, but is unsafe. It is unclear whether excessively early modulation of interhemispheric interaction might worsen the clinical outcome, but early post-insult implantation should be explored: besides its plasticity-inducing effects, ECS has the potential to suppress the inflammatory phase, or to prevent apoptosis. The whole procedure (ECS plus rehabilitation) is costly and clear benefit must be established before this procedure becomes routine. ECS faces direct competition from
269
noninvasive techniques (see chapter 14). However, unlike these latter techniques, ECS can provide longer even continuous- stimulation, because the stimulator is fully implanted. Nonetheless, noninvasive stimulation may have prognostic value as of subsequent ECS and help select the best parameters by highlighting facilitatory or inhibitory effects. A variety of disabilities follow stroke attacks, including cognitive impairment and depression: simple correction of motor or language deficits may fail to improve quality of life, but multi-target ECS might. In conclusion, cortical stimulation to enhance recovery after ischemic stroke represents a new treatment paradigm that may extend our ability to reverse the devastating effects of cerebrovascular disease. Long-term stimulation is required to observe the effect of functional reorganization. Once the stimulation effect plateaus, further stimulation is unnecessary. Such stimulation may be effective long after recovery from the acute phase has ceased. To date, clinical studies have focused on hand/arm function, but opportunity for the treatment of depression and aphasia now exists as well.
REFERENCES
Bang OY, Saver JL, Buck BH, et al. Impact of collateral flow on tissue fate in acute ischemic stroke. J Neurol Neurosurg Psychiatry 2008;79:625-9. Bonita R, Beaglehole R. Recovery of motor function after stroke. Stroke 1988;19:1497-1500. Brown JA. Recovery of motor function after stroke. Prog Brain Res 2006;157:223-228. Brown JA, Lutsep H, Cramer SC, Weinand M. Motor cortex stimulation for enhancement of recovery after stroke: case report. Neurol Res 2003;25:815-818. Brown JA, Lutsep HL, Weinand M, Cramer SC. Motor cortex stimulation for the enhancement of recovery from stroke: a prospective, multicenter safety study. Neurosurgery 2006;58:464-473. Calautti C, Baron JC. Functional neuroimaging studies of motor recovery after stroke in adults. Stroke 2003;34:1553-1566. Canavero S, Bonicalzi V, Intonti S, Crasto S, Castellano G. Effects of bilateral extradural cortical stimulation for plegic stroke rehabilitation. Neuromodulation 2006; 9:28-33. Canavero S, Bonicalzi V: Extradural cortical stimulation for central pain. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: Springer-Verlag, 2007a, pp 27-36. Canavero S, Bonincalzi V. Extradural cortical stimulation for movement disorders. In: D Sakas, B Simpson, E Krames, eds. Operative Neuromodulation. Vol.2. Wien: SpringerVerlag, 2007b, pp 223-232. Cao Y, D'Olhaberriague L, Vikingstad E, Levine S, Welch K. Pilot study of functional MRI to assess cerebral activation of motor function after poststroke hemiparesis. stroke 1998;29:112-122. Carey JR, Kimberley TJ, Lewis SM, et al. Analysis of fMRI and finger tracking training in subjects with chronic stroke. Brain 2002; 125(Pt 4): 773-788.
270
Celnik P, Hummel F, Harris-Love M, Wolk R, Cohen LG. Somatosensory stimulation enhances the effects of training functional hand tasks in patients with chronic stroke. Arch Phys Med Rehabil 2007;88:1369-1376. Cramer SC, Benson RR, Burra VC, et al. Mapping individual brains to guide restorative therapy after stroke: rationale and pilot studies. Neurol Res 2003; 25: 811-814. Cramer SC, Benson RR, Himes DM, et al. Use of functional MRI to guide decisions in a clinical stroke trial. Stroke 2005;36(5):e50-52. Cramer SC, Parrish TB, Levy RM, et al. Predicting functional gains in a stroke trial. Stroke. 2007;38:2108-14. Cramer SC, Shah R, Juranek J, Crafton KR, Le V. Activity in the peri-infarct rim in relation to recovery from stroke. Stroke 2006;37(1):111-115 Dominkus M, Grisold W, Jelinek V. Transcranial electrical motor evoked potentials as a prognostic indicator for motor recovery in stroke patients. J Neurol Neurosurg Psychiatry 1990;53:745-748. Duncan P, Goldstein L, Matchar.D., Divine G, Feussner J. Measurement of motor recovery after stroke. Outcome assessment and sample size requirement. Stroke 1992;23:10841089. Duncan P, Lai S, Keighley J. Defining post-stroke recovery: implications for design and interpretation of drug trials. Neuropharmacology 2000;39:835-841. Feys HM, De Weerdt WJ, Selz BE, et al. Effect of a therapeutic intervention for the hemiplegic upper limb in the acute phase after stroke: a single-blind, randomized, controlled multicenter trial. Stroke 1998; 29:785-792. Floel A, Ellger T, Breitenstein C, Knecht S. Language perception activates the hand motor cortex: implications for motor theories of speech perception. Eur J Neurosci 2003;18:704-708 Gresham GE, Alexander D, Bishop DS, et al. American Heart Association Prevention Conference. IV. Prevention and Rehabilitation of Stroke. Rehabilitation. Stroke 28:15221526, 1997. Harvey RL, Winstein CJ, for the Everest trial group. Design for the Everest randomized trial of cortical stimulation and rehabilitation for arm function following stroke. Neurorehabil Neural Repair 2009; 23: 32-44 Hendricks HT, Limbeek J, Geurts AC. Motor Recovery After stroke: A Systematic Review of the Literature. Arch Phys Med Rehabil 2002;83:1629-1637. Houweling A, Brecht M. Behavioral report of single neuron stimulation in somatosensory cortex. Nature 2008;451:65-68. Huang M, Harvey RL, Stoykov ME, et al. Cortical Stimulation for Upper Limb Recovery Following Ischemic Stroke: A Small Phase II Pilot Study of a Fully Implanted Stimulator. Top Stroke Rehabil 2008;15:160-172. Hummel FC, Cohen LG. Drivers of brain plasticity. Curr Opin Neurol 2005;18(6):667-674. Jin K, Wang X, Xie L, et al. Evidence for stroke-induced neurogenesis in the human brain. Proc Natl Acad Sci U S A 2006;103:13198-13202. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil 1995;76:27-32.
271
Katayama Y, Tsubokawa T, Yamamoto T. Chronic motor cortex stimulation for central deafferentation pain: experience with bulbar pain secondary to Wallenberg syndrome. Stereotact Funct Neurosurg 1994; 62: 295-299. Katayama Y, Fukaya C, Yamamoto T. Post-stroke pain control by chronic motor cortex stimulation: Neurological characteristics predicting a favorable response. J Neurosurg 1998;89:585-591. Kim HI, Shin YI, Moon SK, Chung GH, Lee MC, Kim HG. Unipolar and continuous cortical stimulation to enhance motor and language deficit in patients with chronic stroke: report of 2 cases. Surg Neurol 2008; 69:77-80. Klimesh W,Sauseng P, Gerloff C. Enhancing cognitive performance with repetitive transcranial magnetic stimulation at human individual alpha frequency. Eur J Neurosci 2003; 17: 1129-1133. Kreisel SH, Bazner H, Hennerici MG. Pathophysiology of stroke rehabilitation: temporal aspects of neuro-functional recovery. Cerebrovasc Dis 2006;21:6-17. Kreisel SH, Hennerici MG, Bazner H. Pathophysiology of stroke rehabilitation: the natural course of clinical recovery, use-dependent plasticity and rehabilitative outcome. Cerebrovasc Dis 2007;23:243-255. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJ. Probability of regaining dexterity in the flaccid upper limb: impact of severity of paresis and time since onset in acute stroke. Stroke 2003;34 :2181-2186. Kwon YH, Lee MY, Park JW, et al. Differences of cortical activation pattern between cortical and corona radiata infarct. Neurosci Lett 2007;417:138-142. Leal-Campanario R, Delgado-Garcia J, A G. Microstimulation of the somatosensory cortex can substitute for vibrissa stimulation during Pavlovian conditioning. Proc Natl Acad Sci USA 2006;103:10052-10057. Levy R, Ruland S, Weinand M, Lowry D, Dafer R, Bakay R. Cortical stimulation for the rehabilitation of patients with hemiparetic stroke: a multicenter feasibility study of safety and efficacy. J Neurosurg 2008;108:707-714. Levy RM, Benson RR, Winstein CJ. Cortical stimulation for upper-extremity hemiparesis from ischemic stroke. Stroke 2008; 39: 568, A61 Loubinoux I, Carel C, Pariente J, et al. Correlation between cerebral reorganization and motor recovery after subcortical infarcts. Neuroimage 2003;20:2166-2180. Loubinoux I. Can fMRI measures of brain motor activation add significantly to other variables in the prediction of treatment response? Stroke 2007; 38: 2032. Marshall RS, Pertera GM, Lazar RM, Krakauer JW, Constantine RC, DelaPaz RL. Evolution of cortical activation during recovery from corticospinal tract infarction. Stroke 2000;31:656-661. Meinzer M, Flaisch T, Breitenstein C, Wienbruch C, Elbert T, Rockstroh B. Functional rerecruitment of dysfunctional brain areas predicts language recovery in chronic aphasia. Neuroimage 2008;39:2038-2046. Meister IG, Sparing R, Foltys H, et al. Functional connectivity between cortical hand motor and language areas during recovery from aphasia. J Neurol Sci 2006;247:165-168.
272
Nelles G, Spiekermann G, Jueptner M, et al. Reorganization of Sensory and Motor Systems in Hemiplegic Stroke Patients; A Positron Emission Tomography study. Stroke 1999;30:1510-1516. Nudo RJ. Plasticity. NeuroRx 2006;3:420-427. Pascual-Leone A, Amedi A, Fregni F, Merabet LB: The plastic human brain cortex. Annu RevNeurosci 28:377-401, 2005. Plautz EJ, Barbay S, Frost SB, et al. Post-infarct cortical plasticity and behavioral recovery using concurrent cortical stimulation and rehabilitative training: a feasibility study in primates. Neurol Res 2003;25:801-810. Rossini PM, Calautti C, Paurl F, Baron J. Post-stroke plastic reorganization in the adult brain. Lancet Neurology 2003;2:493-502. Salter K, Jutai J, Foley N, Hellings C, Teasell R. Identification of aphasia post stroke: a review of screening assessment tools. Brain Inj 2006;20:559-568. Saur D, Lange R, Baumgaertner A, et al. Dynamics of language reorganization after stroke. Brain 2006;129(Pt 6):1371-1384. Sawaki L, Wu CW, Kaelin-Lang A, Cohen LG. Effects of somatosensory stimulation on usedependent plasticity in chronic stroke. Stroke 2006;37(1):246-247. Strens LH, Fogelson N, Shanahan P, Rothwell JC, Brown P. The ipsilateral human motor cortex can functionally compensate for acute contralateral motor cortex dysfunction. Curr Biol. 2003;13:1201-5. Struppler A, Binkofski F, Angerer B, et al. A fronto-parietal network is mediating improvement of motor function related to repetitive peripheral magnetic stimulation: A PET-H2O15 study. Neuroimage 2007;36 Suppl 2:T174-186. Tarkka IM, Kononen M, Pitkanen K, Sivenius J, Mervaala E. Alterations in cortical excitability in chronic stroke after constraint-induced movement therapy. Neurol Res 2008; 30: 504-510. Turney TM, Garraway WM, Whisnant JP. The natural history of hemispheric and brainstem infarction in Rochester, Minnesota. Stroke 1984; 15:790-794. Ward NS: Neural plasticity and recovery of function. Prog Brain Res 150:527-535, 2005. Wise SP, Boussaoud D, Johnson PB, Caminiti R. Premotor and parietal cortex: corticocortical connectivity and combinatorial computations. Annu Rev Neurosci 1997;20:25-42 Wolters A, Schimidt A, Schramm A, et al. Timing-dependent plasticity in human primary somatosensory cortex. J Physiol 2005;565:1039-1052. Wu CW, van Gelderen P, Hanakawa T, Yaseen Z, Cohen LG. Enduring representational plasticity after somatosensory stimulation. Neuroimage 2005;27(4):872-884.
COMA REHABILITATION
Chapter 16
EXTRADURAL BIFOCAL CORTICAL STIMULATION FOR THE POST-TRAUMATIC PERMANENT VEGETATIVE STATE
Sergio Canavero, Barbara Massa-Micon, Franco Cauda and Federico DAgata
Turin Advanced Neuromodulation Group (TANG), and Brain Imaging Center, Koelliker Hospital, Turin, Italy.
INTRODUCTION
The vegetative state (VS), an artifact of intensive care units introduced in the 1960s, was described by Jennett and Plum in 1972 (Jennett 2002, RCP 2003). The VS is diagnosed when, after some days to weeks of brain injury, comatose patients open their eyes, but their motor activity is reflexive only, devoid of any voluntary interaction with the environment. The VS may be a transition to further recovery, including the minimally conscious state (MCS) (Giacino et al 2002), or not. Clinical differentiation between VS and MCS is very difficult and based on the very subtle distinction between reflexive reactions in VS and sporadic, weak and inconsistent intentional actions in MCS. Estimation of the presence of consciousness requires expert clinical interpretation of motor responsiveness. VS patients can move extensively: differentiating reflex or automatic from voluntary or willed movements is thus hard. This results in an underestimation of behavioral signs of consciousness and hence misdiagnosis (up to 50% of vegetative cases) (Andrews et al 1996), with diagnostic error rates of 30-40% of patients in VS, who might otherwise be in MCS (Andrews et al 1996, Laureys et al 2005). Visual fixation is one of the more controversial clinical signs of consciousness, which is considered either as compatible but atypical feature or a purposeful sign of the minimal conscious state if it is sustained, but not brief. Other controversial clinical
276
Sergio Canavero, Barbara Massa-Micon and Franco Cauda

signs are visual tracking, blinking in response to visual threat and orientated motor responses to noxious stimuli. Assessment of motor or autonomic signs (heart rate, papillary diameter, skin conductance, respiratory frequency, blood pressure) are unreliable indicators of conscious perception of pain. Patients remaining in VS demonstrate widespread death of thalamic and cortical (typically less injured) neurons and deafferentation and compression injuries to the thalamus and midbrain (Kampfl et al 1998), with impaired thalamo-cortical and corticocortical effective connectivity (Laureys 2005; Laureys et al 2007). Traumatic VS has an incidence of 1-10/100000, with a prevalence lying between 56 and 140 per million. In 1994, the Multi-Society Task Force on Persistent Vegetative State concluded that the chance of recovery from the VS 1 year after traumatic brain injury (TBI) is close to zero, i.e. it is permanent - PVS (Wijdicks and Cranford 2005). Recovery mechanisms from VS and MCS are poorly understood. Nonetheless, cases of late emergence from PVS are on record (Andrews 1993, Childs and Merger 1996), sometimes even 5 years after the incident (Dyer 1997). A patient emerged from traumatic VS 20 months after the event: beginning from the sixth month, event related potentials (ERPs) to complex sensory and verbal stimulation started to improve, although the clinical examinations remained unchanged (Faran et al 2006). In Kotchoubey et al (2005)s series, all PV patients who recovered did so within 10 months of injury. However, according to Laureys (2005), up to 40% of patients in apparent VS have some signs of consciousness (see also Laureys et al 2005). Thus, recovery from PVS, unlike MCS (Lamni et al 2005), is extremely rare (about 10 cases: Wijdicks and Cranford 2005) and no better than a severely disabled fully dependent state of living. Since the 1970s, multiple attempts have been made to use deep brain stimulation (DBS) of the midbrain tegmentum, nonspecific posterior intralaminar thalamic nuclei (CM/Pf) and globus pallidus to improve arousal and awareness in VS patients. Although eye opening and some fragmentary movements were generally observed, consistent with an arousal effect, no examples of recovery of sustained interactive behavior were noted; neither were formal behavioral assessments conducted to link DBS to improvement. In the 1980s, a large multicenter, multinational study of CM DBS plus cervical spinal cord stimulation in 50 VS patients (including Terry Schiavo) did not yield convincing results. Despite clear, clinically judged increases in arousal and physiologic responses to brain stimulation in many patients, including changing of the frequency content of the EEG and increases in cerebral metabolic rates measured with PET, substantive clinical improvements were not identified. Although a small number of patients with traumatic brain injury exhibited significant improvement, all were studied within 6 months of injury, well within the expected recovery window. A large series is that of Yamamoto and Katayama (2005). Twenty-one cases of the vegetative state (VS) and 5 cases of the minimally conscious state (MCS) caused by various kinds of brain damage were treated by deep brain stimulation (DBS) therapy at 3 months after brain injury and followed up for over 10 years. The mesencephalic reticular formation was selected as a target in 2 cases of VS, and the CM-pf complex in the other 19 cases of VS and all 5 cases of MCS. Eight of the 21 patients emerged from the VS, and became able to obey verbal commands, but remained bedridden, except for one case. Of course, given the short interval from injury, emergence could have been due to spontaneous recovery. On the other hand, four of the 5 MCS patients emerged from the bedridden state, and were able to enjoy their life in their own home. Thus, the application of DBS to VS patients in general does not appear to be supported by available data. DBS may damage deep nuclei and further impair possible recovery and most importantly- cannot resynchronize widely disconnected and damaged neuronal tissue.
Extradural Bifocal Cortical Stimulation
277
RATIONALE FOR ECS

We tried ECS in two patients in PVS in the summer of 2007. ECS can enhance neuroplasticity following brain stroke, by altering, among others, neurotransmitter levels in the brain (chapter 15). The rationale of this trial was based on research carried out over the past 10 years or so. SPECT, PET and fMRI studies rejected the ancient view that VS patients are apallic i.e. neocortically dead: cerebral activation in isolated and disconnected metabolic islands of lower level cortices or pallium in response to auditory, visual, somatosensory and noxious stimuli has been observed, suggesting the potential for cognitive processing in a subset of patients and potentially recruitable cortical regions (Laureys et al 2004, 2005). While the initial descriptions of PVS (apallic syndrome) by Kretschmer in 1940 and Jennett and Plum in 1972 assumed at least functional decortication in PVS, some level of cortical processing may actually remain in occasional PVS patients. PET studies demonstrate a very low metabolic rate in PVS comparable to that in deep anesthesia (Alkire and Miller 2005), particularly in the precuneus, posterior cingulate cortex, BA19, BA22, BA30 and BA39, although in some patients islands of well functioning cortex may survive (Laureys et al 2004,2005,2007), as already suggested by ERP studies (Laureys et al 2004, 2005). Kotchoubey et al (2005) found that, among patients with prevailing theta or slow alpha EEG background activity, all patients diagnosed as PVS exhibited some cortical responsivity (13.4%). In some of them, only primary cortical components N1 and/or P2 were reliably proven. More complex patterns were also present: the mismatch negativity was found in about 50% of the patients, an oddball-P3 in about 33% and cortical evidence for semantic differentiation in about 25%, confirming activity in associative auditory cortex. All these numbers were probably underestimations. A patient may lack the simplest cortical responses, but still exhibit the more complex responses. There were no major differences in ERPs between PVS with the above EEG and MCS. The highly significant differences between PVS with very severe versus only moderate disturbances in the background EEG underscores the importance of the activity of thalamocortical gating systems mediating neural mechanisms of perception. Isolated thalamocortical circuits may be working in PVS indicating some specialized processing modules (islands) separated from larger networks in which they are normally involved, although this processing is unrelated to conscious experience (Laureys et al 2005). Machado et al (2007) reported imaging evidence of speech recognition in PVS. It is important to underscore the fact that the VS and the MCS are truly different physiological entities (see also Kotchoubey et al 2005, Laureys and Boly 2007). Although overall cerebral metabolism is decreased to values slightly higher, but comparable to those observed in VS (Laureys et al 2004), activation studies show large-scale higher-order cortical activation in MCS, but normally not in VS patients (Laureys et al 2004, Laureys 2005, Laureys and Boly 2007).VS and MCS patients show differences in both functional segregation and integration (e.g. Boly et al 2004) and neuro-metabolic coupling between neuronal electrical function and cerebral metabolism, i.e. preservation in MCS, but not PVS (Coleman et al 2005). Nonetheless, recent studies strongly suggest that some VS patients not only retain islands of preserved function, but also full-blown self-awareness, which, for some reasons, cannot be made explicit to others (Owen and Coleman 2007). In one highly publicized case, repeated and prolonged activation in response to an instruction to perform a
278
mental imagery task provided evidence of voluntary task-dependent brain activity and hence consciousness (Owen et al 2006). In another study, one of 8 VS and both MCS patients showed an increased hand EMG signal specifically linked to a verbal command (Bekinschtein et al 2008). Several caveats concerning instrumental differentiation of VS cases are in order. ERP and neuroimaging activation paradigms may merely identify neural activation reflecting passive automatic processing, rather than demonstrating preserved awareness (Laureys et al 2007) and available behavioral assessment tools (Laureys et al 2005, 2007) do not have predictive power at the single patient level. The absence of cortical somatosensory evoked potentials (SSEPs) predict an unfavorable outcome, the outcome being worse for patients with hypoxia-ischemia than trauma, whereas their presence moderately predicts a good outcome. The presence of the mismatch negativity (MMN) that appears in auditory oddball paradigms seems a promising predictor of re-emergence from coma, whereas P300 and N400 ERP responses do not necessarily reflect conscious perception and cannot be used to differentiate vegetative state from MCS; moreover, ERPs in general have a rather low spatial resolution. PET and fMRI studies are not reliable markers of recovery of consciousness (Owen and Coleman 2007). In the absence of a full understanding of the neural correlates of consciousness, a normal activation in response to passive sensory stimulation cannot be taken as incontestable proof of consciousness. The power of PET studies to detect statistically significant responses is low and group studies are often needed to satisfy standard statistical criteria. FMRI offers increased statistical power, improved spatio-temporal resolution, with no associated radiation burden. However, the acquisition, analysis and interpretation of fMRI data from patients with severe brain damage are complex. The coupling of neuronal activity and local hemo-dynamics is likely to differ from healthy controls, making interpretation of such data sets extremely difficult. For one, the regional distribution of the cortical metabolic dysfunction is unrelated to the regional distribution of the structural damage (Juengling et al 2005). Higher-order activations following the patients own name spoken by a familiar voice without appropriate controls lack cognitive specificity, and may be due to e.g. an orienting response to speech in general or an emotional response to the speaker, among others. As a result, the interpretation hinges on a reverse inference, by which the engagement of a given cognitive process is inferred solely on the basis of the observed activation in a particular brain region. Yet a response to one owns name is very basic and may not depend on higherlevel linguistic processes underpinning comprehension. Responses to faces, speech and pain are relatively automatic and normal activation patterns are thus of limited significance. Complex and incompletely reliable activation paradigms have been developed and tested in order to ascertain possible preserved self-awareness. Besides resting studies, activation studies offer the possibility to link residual neural activity to controlled external stimulation in the absence of any overt motor response of the patient, but normal neural responses in these patients do not necessarily indicate that they have conscious experience associated with processing external stimuli. The choice of the experiment is also crucial: if the auditory pathway is damaged, one must investigate the visual pathway. The investigation should be complex enough that the cognitive processes of interest will be studied, yet not so complex to overload the cognitive capacities of a tired and inattentive patient. Control studies are essential which must produce well documented, anatomically specific, robust and
279
reproducible activation patterns in healthy subjects. In VS and MCS, episodes of low arousal and sleep are common (EEG during scanning is essential in order to avoid these episodes during fMRI). Spontaneous movements during the scan may compromise the interpretation. The presence of gross hydrocephalus or focal pathology may complicate the fitting of functional imaging data to structural imaging data and the normalization of these images through reference to a healthy brain. Under these circumstances, statistical assessment of activation patterns is complex and interpretation of activation foci with standard stereotaxic coordinates may be impossible. Thus, negative findings in VS/MCS cases must never be used as evidence of lack of awareness: false negatives in functional neuroimaging are common, even in healthy volunteers. However, fMRI can complement clinical assessment of patients in whom voluntary movements may be very small, inconsistent and easily exhausted (Galanaud et al 2007). Finally, the effects on the fMRI activation patterns of deep sedation used to keep the patients motionless are still poorly understood.
CHOICE OF TARGET
The most impaired network in imaging studies is the frontoparietal network of polymodal associative cortices (Laureys et al 2004). This network encompasses the lateral posterior parietal, prefrontal and parieto-temporal regions and midline precuneal, posterior cingulate and mesiofrontal areas and is known to be the most active by default in resting nonstimulated conditions in conscious controls (Northoff and Bermpohl 2004; Fox and Raichle 2007). The precuneus has a central role in visuo-spatial imagery, episodic memory retrieval and first-person perspective taking and the experience of agency; the precuneus and the posterior cingulate areas have been proposed to be part of the neural network (BA 39, 40, 22 and 19) subserving self-awareness and conscious experience (Cavanna and Trimble 2006), and this is markedly impaired in PV (Laureys et al 2004). Frontoparietal association areas could have a special relationship with consciousness, even though they do not support the contents of sensory experience: they may support a self system that is disrupted in PV (Baars et al 2003). The parietal cortex supports the first person perspective on the visual world, unconsciously framing the visual object stream, while some prefrontal areas (including the dorsolateral prefrontal cortex) select and interpret conscious events for executive control (Baars et al 2003). Restoration of functional disconnections between the intralaminar thalamus and frontal cortices paralleled recovery of consciousness from VS (Laureys et al 2004, 2005, 2007), whereas a patient that emerged from MCS 19 years later showed parieto-occipital white matter anisotropy renormalization (due to denser distribution of axons or stronger myelination) (Voss et al 2006). On this basis, we thus elected to stimulate the parietal gyri P1 and P2 and the middle frontal sulcus (F2), including Brodmanns areas 8 and 46 in order to recruit this consciousness network, connected via the superior longitudinal fasciculus. Stimulation of DLPFC, via its connections with the supplementary motor area (SMA), was also expected to influence swallowing and axial tone. Neurological and neuroimaging evidence underscores the importance of the right hemisphere in the processes of self-awareness (e.g. Serafenides 1995, Keenan et al 2001,
280
Northoff and Bermpohl 2004). In our study, we stimulated the female on the left side and the male on the right to confirm the role of laterality in possible consciousness recovery.
STUDY HIGHLIGHTS
In the summer of 2007, we applied extradural bifocal cortical stimulation in two patients in the permanent vegetative state for 19-20 months (i.e. beyond the maximum time of possible spontaneous recovery) since traumatic brain injury (for which decompressive hemicraniectomy was performed only in the female case) (Canavero et al, 2009). Experimental studies must be conducted after spontaneous recovery is statistically unlikely and formal assessment demonstrates a flat behavioral baseline. The N20/P25 components of the SSEPs were absent bilaterally in both cases. The female (born 1988) was scored 25 on the Disability Rating Scale/DRS (Category 9) and the male (born 1985) 23 (category 8). The male patient had been on intrathecal baclofen for severe spasticity for several months. His defensive blink reflex was present and brisk, whereas it was completely absent in the female patient. While the male could be fed regularly, the female only with great difficulty. After induction of general anesthesia, a double sigmoid incision of the skin overlying the target areas was performed. Specifically, the left superior parietal lobule and dorsolateral prefrontal cortex were targeted (figures 1 and 2). Four burr holes were fashioned and two stimulating paddles were inserted extradurally (Lamitrode 4, MOD. 3240, ANS, Plano, Texas). The paddles were linked via a dual extension to a subclavearly pocketed pulse generator (Genesis, MOD. 3608, ANS, Plano, Texas). Stimulation was started at a low power and then gradually increased (8-12 mA). It consisted of daily stimulation with switching off at night. Parameters were determined empirically, on the basis of our previous experience with ECS for other disorders (chapters 9,10, 13,13b and 15). Low frequency stimulation was elected (6-16 Hz) with pulse width trials ranging between 52 and 455 s in the female patient. In the male patient, assessed parameters were 6-100 Hz, 65-455 s, 8-13 mA, 0+1-23+ /0+1-2-3+. Clinical progression was evaluated over the following 10 months on 9 occasions by means of the Coma Recovery Scale-Revised and the Levels of Cognitive Functioning Scale. The CRS-R is a standardized behavioral rating scale comprised of six subscales designed to assess auditory, visual, motor, oromotor/verbal, communication and arousal functions. Subscale items are arranged hierarchically in that the lowest item reflects reflexive activity and the highest item represents cognitively-mediated behavior. Interrater and test-retest reliability for the CSR-R score are high. The scale has a high concurrent validity with the Disability Rating Scale, a well-established measure that gauges neurobehavioral responsiveness and functional capacity (Giacino et al 2006). Within 48 hours of switching the stimulator on, a few days after surgery, both showed increased arousal during follow-up. On changing parameters during follow-up, it was observed that arousal, spasticity and other vegetative parameters could dramatically change (improve or worsen) within 12 hours. High frequency (100 Hz) stimulation was not tolerated (spasticity increased) in the male patient. Best parameters were 50-60 Hz , 65-208 s, 8-10 mA. In the female, best parameters were 8-10 Hz, 65 s, 11 mA, ++--/--++. Intensity was higher than
281
(a)
(b)
(c)
Figure 1. case 1 (female). Neuronavigation images showing the parietal (a) and frontal (b) targets. Lateral skull x-rays showing the position of the two stimulating strips (c).
282
(a)
(b)
(c)
Figure 2. case 2 (male). Neuronavigation images showing the parietal (a) and frontal (b) targets. Lateral skull x-rays showing the position of the two stimulating strips (c).
283
reported in DBS studies (2-3 mA versus 8-13 mA). Effects emerged immediately, but strengthened in time. The female showed increased vigilance with clear improvements of swallowing and self-management of oral secretions. Oral feeding with both solids and liquids has become possible and episodes of aspiration have not been reported; weight has increased by 4 Kg. Axial tone has increased dramatically. Most importantly, occasionally after the first month and on a more repeatable basis at study end, she could lift her left arm and hand on command, a clear sign of consciousness. On several occasions, the physiotherapists had the clear impression of the patient being conscious and cooperative. The male could respond to emotionally charged stimuli with appropriate facial expressions. Consistent interaction with family was the most important change cited by family members. Neuro-imaging by default-state MRI (Fox and Raichle 2007), which we first validated in a previous study as providing a marker of self-consciousness (see Cauda et al 2009 for details), showed a clear improvement, with a pattern towards normalization. Diffusion tensor imaging (DTI) suggested fiber regrowth in the superior longitudinal fasciculus and corpus callosum in the female patient, but not the male. In the male, Magnetic Resonance Spectroscopy (MRS) showed signs of altered neuronal metabolism (Table 1). Importantly, at the end of study, stimulators were deactivated: the benefits persisted for a few weeks, a sign of neuroplastic effects seen also in Parkinson Disease and central pain, socalled after-effect (chapter 9 and 13). In the female patient, benefits dwindled thereafter, only to be recaptured upon replacing the battery 4 months later. Table 1. MRS (posterior frontal white matter) data (male patient) PREOPERATIVE MRS NAA/Cr 1.28 NAA/Cr(h) 1.32 Cho/Cr 0.75 Cho/Cr(h) 0.69 NAA/Cho 1.72 NAA/Cho(h) 1.92 Cho/NAA 0.58 Cho/NAA(h) 0.52 POSTOPERATIVE MRS (6 weeks after surgery) NAA/Cr 1.94 NAA/Cr(h) 1.83 Cho/Cr 0.69 Cho/Cr(h) 0.80 NAA/Cho 2.82 NAA/Cho(h) 2.28 Cho/NAA 0.35 Cho/NAA(h) 0.44
In conclusion, given the results in two patients and loss and later recapture of benefit in one patient, chance recovery can be dismissed. Along with reports of spontaneous late emergence from VS, these preliminary data indicate that the permanent VS may not be permanent after all, at least in some patients, and that the severely injured brain has a capacity for recovery that exceeds current expectations, even after the no-return point as currently assessed has been passed: this highlights the need for continuation of treatment efforts even after the 1 year mark. The high cost of the stimulation apparatus could be offset by the reduced cost of care. These results could well apply to a substantial share of PVS patients. In the future, ECS might be combined with DBS, spinal cord stimulation or other techniques. Recovery from PVS is hampered by damaged white matter tracts (e.g. following contusions, hematomas), but
284
these preliminary data suggest that ECS can trigger neuroplastic changes (chapters 9,13,15) and can stimulate fiber regrowth (as seen in the female, but not in the male patient). Potentially, ECS could promote survival and growth of stem cells in future trials. Possible ongoing degenerative processes months after the initial injury might be delayed. Importantly, ECS can force into resynchronization rebind- (Slewa-Younan et al 2002) and rebalance activity across wide swaths of damaged hemispheres bilaterally (as seen during ECS for Parkinsons disease: chapter 13 and 13B), by altering thalamocortical transmission ipsilaterally and contralaterally via the corpus callosum and other deep structures (He et al 2007; chapter 13B). Cortico-cortical coherence between distant brain areas has been selectively enhanced by simultaneous bifocal 10Hz rTMS, with an increase of interregional coupling in the alpha and lower beta bands on the stimulated side without effects on spectral power (Plewnia et al 2008). ECS may also compensate for a loss of arousal regulation that is normally controlled by the frontal lobe in the intact brain. Future stimulation efforts will also consider adding MI in a trifocal stimulation, bilaterally in nonresponders.
REFERENCES
Alkire MT, Miller J. General anesthesia and the neural correlates of consciousness. Prog Brain Res 2005;150:229-44. Andrews K, Murphy L, Munday R, Littelwood C. Misdiagnosis of the vegetative state: retrospective study in a rehabilitation unit. BMJ 1996; 313: 13-16 Baars BJ, Ramsoy TZ, Laureys S. Brain, conscious experience and the observing self. TINS 2003; 26: 671-5 Bekinschtein TA, Coleman MR, Niklison III J, Pickard JD, Manes FF. Can electromyography objectively detect voluntary movement in disorders of consciousness? JNNP 2008; 79: 826-828 Boly M, Faymonville ME, Peigneux P, et al. Auditory processing in severely brain injured patients: differences between the minimally conscious state and the persistent vegetative state. Arch Neurol 2004; 61: 233-238 Canavero S, Massa-Micon B, Cauda F and Montanaro E. Cortical stimulation-induced recovery of consciousness in the permanent post-traumatic vegetative state. J Neurol 2009 (in press) Cauda F, Massa-Micon B, Sacco K, Duca S, DAgata F, Geminiani G, Canavero S. Disrupted intrinsic functional connectivity in the vegetative state. J Neurol Neurosurg Psych 2009 (in press) Cavanna AE, Trimble MR. The precuneus: a review of its functional anatomy and behavioural correlates. Brain 2006; 129: 564-583 Childs NL, Merger WN. Late improvement in consciousness after posttraumatic vegetative state. NEJM 1996; 334: 24-25 Coleman MR, Menon DK, Fryer TD, et al. Neurometabolic coupling in the vegetative and minimally conscious states: preliminary findings. J Neurol Neurosurg Psych 2005; 76; 432-434
285
Dyer C. Hillsborough survivor emerges from permanent vegetative state. BMJ 1997; 314: 993 Faran S, Vatine JJ, Lazary A, Ohry A, Birbaumer N, Kotchoubey B. Late recovery from permanent traumatic vegetative state heralded by event-related potentials. JNNP 2006;77:998-1000. Fox MD, Raichle ME. Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nat Rev Neurosci 2007;8:700-11. Galanaud D, Naccache L, Puybasset L. Exploring impaired consciousness: the MRI approach. Curr Op Neurol 2007; 20: 627-631 Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state. Definition and diagnostic criteria. Neurology 2002; 58: 349-353 Giacino J.T., Kalmar K., Whyte J. The JFK Coma Recovery Scale-Revised: Measurement Characteristics and Diagnostic Utility. Arch Phys Med Rehab 2006; 85 : 2020-2029 He BJ, Shulman GL, Snyder AZ, Corbetta M. The role of impaired neuronal communication in neurological disorders. Curr Op Neurol 2007; 20: 655-660 Jennett B. The vegetative state. Cambridge: Cambridge University Press, 2002 Juengling FD, Kassubek J, Huppertz HJ, et al. Separating functional and structural damage in persistent vegetative state using combined voxel-based analysis of 3-D MRI and FDGPET. J Neurol Sci 2005; 228: 179-184 Kampfl A, Schmutzhard E, Franz G, et al. Prediction of recovery from post-traumatic vegetative state with cerebral magnetic-resonance imaging. Lancet 1998; 351 1736-67 Keenan JP, Nelson A, OConnor M, Pascual-Leone A. Self recognition and the right hemisphere. Nature 2001; 409: 305 Kotchoubey B, Lang S, Mezger G, et al. Information processing in severe disorders of consciousness: vegetative state and minimally conscious state. Clin Neurophysiol 2005; 116: 2441-2453. Lammi MH, Smith VH, Tate RL, Taylor CM. The minimally conscious state and recovery potential: a follow-up study 2 to 5 years after traumatic brain injury. Arch Phys Med Rehabil 2005; 86: 746-754 Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disorders. Lancet Neurol 2004; 3: 537-46. Laureys S, Perrin F, Schnakers C, Boly M, Majerus S. Residual cognitive function in comatose, vegetative and minimally conscious states. Curr Op Neurol 2005 18; 726-733 Laureys S. The neural correlate of (un)awareness: lessons from the vegetative state. Trends Cogn Sci 2005; 9:556-9. Laureys S, Boly M. What is it like to be vegetative or minimally conscious? Curr Op Neurol 20 609-613, 2007 Laureys S, Perrin F, Bredart S. Self-consciousness in non-communicative patients. Conscious Cogn 2007; 16: 722-41 Machado C, Korein J, Aubert E, et al. Recognizing a mother's voice in the persistent vegetative state. Clin EEG Neurosci 2007; 38:124-6. Northoff G, Bermpohl F. Cortical midline structures and the self. Trends Cogn Sci 2004; 8:102-107
286
Owen AM, Coleman MR, Boly M, Davis MH, Laureys S, Pickard JD. Detecting awareness in the vegetative state. Science 2006; 313: 1402 Owen AM, Coleman MR. Functional MRI in disorders of consciousness: advantages and limitations. Curr Op Neurol 2007; 20: 632-637 Plewnia C, Rilk AJ, Soekadar SR, et al. Enhancement of long-range EEG coherence by synchronous bifocal transcranial magnetic stimulation. Eur J Neurosci 2008; 27: 15771583 RCP, Royal College of Physicians . The vegetative state. Guidance on diagnosis and management. Clin Med 2003 3: 249-254 Serafetinides EA. Cerebral laterality and consciousness. Arch Neurol 1995;52:337-8. Slewa-Younan S, Green AM, Baguley IJ, Felmingham KL, Haig AR, Gordon E. Is gamma(40 Hz) synchronous activity disturbed in patients with traumatic brain injury? Clin Neurophysiol 2002; 113: 1640-1646 Voss HU, Uluc A, Dyke J, et al. Possible axonal regrowth in late recovery from the minimally conscious state. J Clin Invest 2006; 116: 2005-2011 Wijdicks EF, Cranford RE. Clinical diagnosis of prolonged states of impaired consciousness in adults. Mayo Clin Proc 2005; 80: 1037-1046 Yamamoto T, Kobayashi K, Kasai M, Oshima H, Fukaya C, Katayama Y. DBS therapy for the vegetative state and minimally conscious state. Acta Neurochir Suppl. 2005;93:101-4.
PSYCHIATRIC DISORDERS
Chapter 17
INVASIVE AND NON-INVASIVE STIMULATION FOR PSYCHIATRIC CONDITIONS

P. Eichhammer1,, B. Langguth1, M. Landgrebe1, S. Canavero2 and G. Hajak1
1
Department of Psychiatry, Psychosomatics and Psychotherapy, University of Regensburg, 93053-Regensburg, Germany; 2 Turin Advanced Neuromodulation Group, Turin, Italy.
1- MAJOR DEPRESSION
The World Health Organization estimates that 340 million people worldwide suffer from an episode of major depression each year, accounting for 4.4% of the overall global disease burden. In the USA, almost 20 million people are affected by a depressive disorder, with a lifetime risk of about 20-25% (females) or 7-12% (men) for a major depressive episode. Up to 20% of patients completely fail to respond to standard therapy and nearly 60% may not achieve adequate response: on the whole, two million Americans suffer treatment resistant depression. Electroconvulsive therapy (ECT) is effective in about 70% of cases where 3-4 medications failed, but periodic maintenance is indicated in many patients every few months. The use of transcranial magnetic stimulation (TMS) in major depression has been the most-studied clinical application in psychiatry. The studies have ranged from uncontrolled clinical observations of therapeutic effects to randomized, controlled clinical trials. Taken together, despite the high variability in terms of diagnostic and treatment differences, the findings are promising and supported by meta-analyses (e.g. Schutter, 2008). The theoretical background for applying fast-frequency (> 5Hz) repetitive TMS (rTMS) to the left prefrontal cortex in the treatment of depression may find its neurobiological basis in observations that patients with left prefrontal cortex damage develop depressive symptoms.
Correspondence concerning this article should be addressed to: Dr. Peter Eichhammer, MD, E-mail: peter.eichhammer@medbo.de
290
P. Eichhammer, B. Langguth, M. Landgrebe et al.
A substantial body of literature supports the theory that mood is regulated by a network of brain regions (including the prefrontal, cingulate, parietal, and temporal cortical regions as well as parts of the striatum, thalamus, and hypothalamus) and that focal lesions in this network (infarction, tumor, or transient disruption, e.g. by the Wada procedure) can result in mood disturbances. Further support for the involvement of the left prefrontal cortex in depression was provided by functional neuroimaging studies pointing to left anterior hypoactivity in depressive patients (Baxter et al., 1989). Current views give strong support to the fact that restoring the balance between left and right prefrontal cortices may be more important than establishing absolute increases in left-sided activity per se. In favor of this hypothesis, inhibitory slow-frequency (1 Hz) rTMS over the right dorsolateral prefrontal cortex (DLPFC) has proven to possess antidepressant properties (Klein et al., 1999). Moreover, functional imaging studies show that rTMS over the DLPFC not only modulates this area, but also the subgenual cingulate cortex (BA25), thereby suggesting that frontocingular networks may be involved in mediating the antidepressant effects by TMS. Yet, CBF changes in depression are distributed, vary with coil placement, intensity of stimulation, frequency and number of pulses administered; in addition, the number of treatments before imaging, time between imaging and TMS and whether stimulation was performed before or during neuroimaging may also affect results (Lisanby et al 2002). Thus, the selection of the DLPFC as the only target may have been premature, although deeper areas (cingulate cortex, orbitofrontal cortex/ OFC) may become amenable to deep H-coil TMS in the future.
Review of Clinical Trials The vast majority of rTMS in depression focuses on high-frequency stimulation over the left DLPFC (Schutter 2008) (Table 1). A recently published meta-analysis supports the antidepressant efficacy of rTMS beyond sham treatment. An effect size (d=0.39) demonstrates moderate antidepressant effects on frequently-relapsing drug-refractory depression (Schutter, 2008). For this reason, there is meta-analytic evidence indicating that the moderate effect size presently observed is comparable to effect sizes seen in active placebo-controlled trials with drug treatments (Moncrieff et al., 2004). Moreover, the effect size of five more recent rTMS studies was estimated to be 0.76 (Gross et al., 2007), suggesting that newer rTMS stimulation protocols are more effective. The body of literature on the use of rTMS in depression is rapidly growing. Currently, one may try with low frequency right DLPFC rTMS followed, in case of failure, by high frequency left DLPFC rTMS, for at least 1 month (Fitzgerald et al 2003, Januel et al 2006). rTMS is approved as add on treatment for depression in Canada and Israel. Given the slight effects of left vagal nerve stimulation (VNS), on the basis of metaanalytic evidence, rTMS appears superior to VNS. A study suggests that there may be a differential antidepressant response dependent on baseline level of cerebral glucose metabolism, with baseline hypometabolism responding better to high-frequency excitatory stimulation and baseline hypermetabolism responding better to low-frequency inhibitory stimulation (Speer et al 2000).
Invasive and Non-Invasive stimulation for Psychiatric Conditions
291
Future studies should extend the duration of rTMS treatment to 15 sessions or more (over 1 month) (Rumi et al., 2005). Identifying genetic predictors of rTMS efficacy may help to improve antidepressant effects, especially in drug resistant depression (Bocchio-Chiavetto et al., 2008). Neuroimaging may help identify responders (Dougherty et al 2003). Combining rTMS with antidepressant drugs may also improve clinical efficacy (Bretlau et al., 2008). The combination of two biophysical treatment modalities, such as priming stimulation with 6-Hz TMS followed by low-frequency rTMS over the right DLPFC, enhanced clinical efficacy in major depression (Fitzgerald et al., 2008). On the other hand, bilateral rTMS (high frequency on left DLPFC plus low frequency on right DLPFC) is no more effective than unilateral stimulation (Loo et al 2003). The more powerful, less focal round coil may be worth reconsidering. Post-stroke depression too should be submitted to further trials (Jorge et al 2004). All studies only evaluated the short-term effects of rTMS for a few months only, but a single study examined long-term maintenance with high-frequency (10 Hz) rTMS of the left prefrontal cortex in 10 adult patients with major depression, for periods ranging from 6 months to 6 years. rTMS sessions averaged one to two per week. Seven of the 10 subjects experienced either marked or moderate benefit, which was sustained without the need of concomitant antidepressant medications in three cases (OReardon et al 2005). These data, while open label, suggest that maintenance rTMS might have a positive effect in some patients with unipolar depression. Equally important, in patients not refractory to medication, the effect size of rTMS was almost twice as large as the second largest effect size (Januel et al 2006). RTMS might prove particularly useful in children and adolescents (due to their poor response to antidepressants and concerns with increased suicidal risk) and in the elderly (with inadequate dosing and the issue of cerebral atrophy). tDCS could even replace rTMS due to its better safety (Boggio et al 2008). The convention in all studies is to administer an intensity that is set relative to the threshold for eliciting motor responses, but its relevance to areas other than MI is unknown. Clinical trials of extradural cortical stimulation using surgically implanted electrodes on the brain surface (ECS) are underway with some initial promising results. A 12 treatmentresistant patient pilot study (NCT00380042) of ECS of the left DLPFC has been completed. Another 5 patient study is comparing cortical stimulation using surgically implanted electrodes on the medial prefrontal cortex versus ECT (NCT00565617). Continuous stimulation is possible, unlike non-invasive approaches. Stimulation or inhibition can be adjusted on the basis of PET hypo- or hyper-metabolism. Surgical approaches also have the advantage of precision in localization. For instance, most trials of DLPFC targeted an area 5 cm anterior to the optimal site for stimulation of a distal hand muscle, but actually several patients may have been treated more dorsally, over the premotor cortex (Herwig et al 2001). Neuronavigated TMS may overcome this problem. Additionally, current figure-of-8 coils couple poorly to the curved surface of the scalp, a problem not encountered with surgically implanted electrodes. Likewise, these latter overcome the problem of distance between stimulating source and target, as experienced in non-invasive stimulation (Kozel et al 2000). Finally, MI ECS also activates the DLPFC (chapter 11) and this may become an option in future trials (Canavero S, personal observation).
292
P. Eichhammer, B. Langguth, M. Landgrebe et al. Table 1. rTMS studies of Major Depression
Authors/ Year
Patients
George et al.., 1997 Haag et al.., 1997 Avery et al., 1999 Kimbrell et al., 1999 Loo et al., 1999 Padberg et al., 1999 Berman et al., 2000 Eschweiler et al., 2000 George et al., 2000 Garcia-Toro et al., 2001 Manes et al., 2001 Nahas et al., 2003 Szuba et al., 2001 Boutros et al., 2002 Padberg et al., 2002 Fitzgerald et al., 2003 Hppner et al., 2003 Holtzheimer et al., 2004 Jorge et al., 2004 Koerselman et al., 2004 Mosimann et al., 2004 Hausmann et al, 2004* Poulet et al., 2004 Miniussi et al., 2005 Rossini et al., 2005 Rumi et al., 2005 Avery et al., 2006 Herwig et al., 2007 Loo et al., 2007 Mogg et al., 2008 OReardon et al., 2007
12 12 6 8 18 12 20 10 20 35 20 23 14 20 30 40 20 15 20 51 24 38
StimulationSide left DLPFC left DLPFC left DLPFC left DLPFC
rTMSFrequency 20 Hz 10 Hz 10 Hz 20 Hz
StimulationIntensity 80 % MT 90 % MT 80 % MT 80 % MT
Pulses per Session 800 1250 1000 800
Study Design
Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled
left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC Left DLPFC + right DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC left DLPFC
10 Hz 10 Hz 20 Hz 10 Hz 20 Hz 20 Hz 20 Hz 5 Hz 10 Hz 20 Hz 10 Hz 10 Hz 20 Hz 10 Hz 10 Hz 20 Hz 20 Hz 20 Hz
110 % MT 90 % MT 80 % MT 90 % MT 100 % MT 90 % MT 80 % MT 110 % MT 100 % MT 80 % MT 90 % MT 100 % MT 90 % MT 110 % MT 110 % MT 80 % MT 100 % MT 100% MT
1500 250 800 250 1600 1200 800 1600 1000 800 1500 1000 800 1600 1000 800 1600 2000
19 29 96 46 68 105 37 59 277
10 Hz 17 Hz 15 Hz 5 Hz 17 Hz 10 Hz 10 Hz 10 Hz 10 Hz
80 % MT 110 % MT 100 % MT 120 % MT 110 % MT 110 % MT 110 % MT 110 % MT 120 % MT
400 2040 900 1000 2040 2000 1500 1000 3000
Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled Sham-controlled
DLPFC: dorsolateral prefrontal cortex; *: negative study.
As concerns mania, Grisaru et al (1998) randomized 16 manic patients to high-frequency TMS to the left or right DLPFC with a round coil for 2 weeks. Right, but not left, stimulation significantly improved symptoms, an effect opposite to depression. On the other hand,
293
hypomania/mania has been triggered during TMS for depression (Dolberg et al 2001) and this potential risk should be kept in mind, also in surgical trials.
2-SCHIZOPHRENIA
Schizophrenia is a devastating psychiatric illness affecting approximately 0.5% of the population. Since Eugen Bleuler, schizophrenia has been regarded as heterogeneous, encompassing a variety of distinct disease entities. The essential features of schizophrenia are a mixture of characteristic signs and symptoms, positive and negative. This clinical distinction may provide the rationale for using different stimulation protocols in order to treat specific syndromal patterns of schizophrenia Table 2a. rTMS studies of Schizophrenia Positive Symptoms
Authors/Year Patients StimulationSide I-TP rTMSFrequency 1 Hz StimulationIntensity 80 % MT Pulses per Session 2400 Study Design
Hoffmann et al. 2000 DAlfonso et al. 2002 Hoffmann et al. 2003 McIntosh et al. 2004 SchnfeldtLecuona et al. 2004 Poulet et al. 2005 Lee et al. 2005 Saba et al. 2005 Chibbaro et al. 2005 Hoffmann et al. 2005 Brunelin et al. 2005 Fitzgerald et al. 2005
12
9 24
I-AC I-TP
1 Hz 1 Hz
80 % MT 90 % MT
12.200 7920
Sham-controlled Double-blind; Cross-over Open Sham-controlled Double-blind; Cross-over Sham-controlled; Cross-over Sham-controlled; Cross-over Sham-controlled; Cross-over Sham-controlled; Cross-over Sham-controlled; Double-blind Sham-controlled; Double-blind Sham-controlled; Double-blind Sham-controlled; Double-blind Sham-controlled; Double-blind
16 12 10 39 18 16 50 24 33
I-TP I-STG I-TP I-TP I-TP I-TP I-TP I-TP I-TP
1 Hz 1 Hz 1 Hz 1 Hz 1 Hz 1 Hz 1 Hz 1 Hz 1 Hz
80 % MT 90 % MT 90 % MT 100 % MT 80 % MT 90 % MT 90 % MT 90 % MT 90 % MT
2400 4800 10.000 12.000 3000 3600 7920 10.000 9000
I-AC: left acoustic cortex; l-STG: left superior temporal cortex; l-TP: left temperoparietal; MT: motor threshold.
Positive symptoms are associated with delusion, severe thought disorders and auditory hallucinations. Functional imaging studies demonstrate hyperactivation of the right and left temporal cortex including the auditory cortex. Based on these findings, a so-called
294
hyperexcitability-syndrome was postulated, which should be especially sensitive to lowfrequency inhibitory rTMS. In contrast, negative symptoms are associated with hypoactivity in frontal regions of the cortex. The prefrontal cortex shows bidirectional connections with a variety of cortical and subcortical structures including the insula, limbic regions as well as thalamic areas and the basal ganglia. Deranged prefrontal activity is also linked to dysfunctional dopaminergic neurotransmission in mesolimbic structures, which is thought to be involved in generating depressive symptoms and negative symptoms. For this reason, high-frequency stimulation should promote clinical improvement, especially reduction in negative symptoms.
Review of Clinical Trials Positive Symptoms The use of low-frequency rTMS in the treatment of positive symptoms focuses mainly on the reduction of drug-resistant acoustic hallucinations. Pioneering work was done by Ralph Hoffman in Yale (Hoffmann et al., 2000). In a double-blind cross-over study on 12 patients Table 2b. rTMS studies of Schizophrenia - Negative Symptoms
Authors/Year Patients StimulationSide rTMSFrequency StimulationIntensity Pulses per Session 30 1200 1200 Study Design
Geller et al. 1997 Feinsod et al. 1998 Klein et al. 1999 Cohen et al. 1999 Nahas et al. 1999
10 10 31 6 8
FA bilateral r-PFC r-PFC OA I-DLPFC
1 Hz 1 Hz 1 Hz 20 Hz 20 Hz
100 % MT 100 % MT 110 % MT 80 % MT 100 % MT
8000 1600
Rollnik et al. 2000
12
DLPFC dominant I-DLPFC I-DLPFC I-DLPFC DLPFC bilateral I-DLPFC I-DLPFC
20 Hz
80 % MT
8000 10 10.000 3500 10.000 8000
Hajak et al. 2004 Jandl et al. 2004 Holi et al 2004 Jin et al. 2005
20 10 22 20
10 Hz 10 Hz 10 Hz 20 Hz
110 % MT 100 % MT 100 % MT 80 % MT
Sachdev et al. 2005 Novak et al. 2006
4 16
15 Hz 20 Hz
90 % MT 90 % RMT
36.000 20.000 10
Open Open Sham-controlled; Dpouble-blind Open Sham-controlled; Double-blind; Cross-over Sham-controlled; Double-blind; Cross-over Sham-controlled; Double-blind Open Sham-controlled: Double-blind Sham-controlled; Double-blind; Cross-over Open Sham-controlled: Double-blind
FA: frontal area; DLPFC: dorsolateral prefrontal area; OA: orbital area; r-PFC: right prefrontal area; MT: motor threshold.
295
with schizophrenia, 1-Hz-rTMS over the left parietotemporal cortex led to a significant reduction of acoustic hallucinations. These results were confirmed in further studies, but cross-over studies conducted by other groups failed (Table 2a). Neuronavigated rTMS using functional imaging to localize hyperactivity within the temporal cortex in the single subject may improve rTMS efficacy in the treatment of acoustic hallucinations (Schnfeldt-Lecuona et al., 2004). In this context, negative results were attributed to the small number of total TMS stimuli. In favor of this hypothesis, studies using a larger number of magnetic impulses efficiently reduced acoustic hallucinations.
Review of Clinical Trials Negative Symptoms The first clinical findings suggesting low-frequency rTMS in the treatment of negative symptoms were published by Geller (1997) and Feinsod (1998). Both open studies led to an improvement of anxiety and anhedonic feelings without influence on positive symptoms. In 1999, the first double-blind sham-controlled study was published investigating 31 patients with schizoaffective or schizophrenic disorder. Patients were treated with low-frequency rTMS over the right DLPF, without demonstrating any significant therapeutic effect. In contrast, studies using high-frequency rTMS could induce an improvement of negative symptoms. The use of high-frequency rTMS was chosen on the basis of functional imaging pointing to hypoactivity in the prefrontal cortex of schizophrenic patients with negative symptoms. These positive results were confirmed in a sham-controlled study using a 10-Hz stimulation protocol (Hajak et al., 2004). 10-Hz-rTMS has been shown to efficiently modulate mesolimbic and mesostriatal dopamine turn-over, thereby influencing subcortical processes closely linked to the generation of negative symptoms. This work was recently confirmed by a further study (Cordes et al., 2005). Unfortunately, other studies failed to replicate these positive results (Table 2b). Future rTMS studies should focus on measuring quality of life indices after stimulation. Moreover, studies with longer follow-up periods are needed. In addition, to reduce the etiological heterogeneity of schizophrenia, biological sub-typing based on neurobiological findings is warranted. The modes of sham-stimulation and the specific advantages and disadvantages of different placebo-stimulation techniques should be better researched, especially in light of the effect of low-frequency rTMS in treating drug-resistant negative symptoms. Surgical stimulation by implanted electrodes should be attempted.
3-ANXIETY AND PANIC DISORDER

Anxiety disorders are the most frequent psychiatric disorders. These disorders share a common clinical and neurobiological basis with depressive disorders. This close relationship is responsible for a high comorbidity of both disease entities. Anxiety disorders are associated with dysfunctional processing in frontal and limbic cortical areas. For this reason, modulation of activity within prefrontal regions is thought to modify a cortical network closely involved in the pathophysiology of anxiety disorders.
296
Review of Clinical Trials The first rTMS study was published in 2002 on three patients with drug- and psychotherapy resistant panic attacks (Garcia-Toro et al 2002). Low-frequency rTMS over the right DLPFC induced no significant clinical improvement. In contrast to this work, Zwanzger et al (2002) reported a significant reduction of anxiety symptoms in a patient with panic attacks using low-frequency rTMS over the right DLPFC (see also Guaiana 2005). In another work, low-frequency rTMS failed to reduce anxiety symptoms in probands with CCK (Cholecystokinin-Tetrapeptide) induced panic attacks (Zwanzger et al 2007). Additionally, high-frequency rTMS over the left DLPFC may help to reduce anxiety as shown in a patient suffering both from depressive symptoms and a myocardial infarction (Sakkas et al 2006). Large, sham-controlled studies are urgently needed.
4-POSTTRAUMATIC STRESS DISORDER (PTSD)

Cardinal symptoms of PTSD are the reexperience of events (such as visual flashbacks), avoidance behavior, and hyperarousal accompanied by marked anxiety. These symptoms cause significant interference in occupational and social functioning, particularly in soldiers with combat trauma. Recent theories of PTSD pathogenesis suggest that mechanisms involved in normal threat assessment become dysregulated. These hypotheses propose that brain regions associated with fear conditioning and extinction are important in PTSD. These areas include the amygdala, the hippocampus, the medial prefrontal cortex and the affective division of the anterior cingulate. Neuroimaging research in PTSD using MRI (magnetic resonance imaging) or PET (positron emission tomography) suggest that temporal and frontal areas of the right cortex are especially associated with the core symptomatology of PTSD. Moreover, magnetic resonance spectroscopy points to a cell loss in limbic cortical areas as well as in the anterior cingulum. For this reason, modulation of frontal cortex activity may provide a neurobiological avenue to influence a cortical network associated with the generation of PTSD.
Review of Clinical Trials Grisaru et al (1998) treated ten patients with PTSD over the motor cortex using a lowfrequency stimulation protocol. After rTMS, a significant, but transient improvement of PTSD symptoms could be detected. This study was not sham-controlled. In contrast, McCann et al (1998) used a low-frequency rTMS over the right DLPFC in agreement with functional imaging results. Using these stimulation parameters patients showed a reduction in PTSD symptoms paralleled by a reduction in regional cerebral metabolism as evidenced by Glucose-PET. In a placebo-controlled, double-blind study, low-frequency (1 Hz) rTMS was compared with a high-frequency (10 Hz) rTMS over the right DLPFC (Cohen et al., 2004). Only under high-frequency rTMS a significant clinical improvement could be achieved.
297
These positive results were interpreted as a consequence of a successful activation of the right prefrontal cortex, leading to a forced inhibitory action on limbic areas, including the amygdala (Table 3). Future studies should verify the clinical relevance of high-frequency rTMS over the right DLPFC. Moreover, rTMS should be guided by functional imaging results, thereby leading to individualization and optimization of treatment effects. Cortical patterns of activity may allow to choose both the optimal target area as well as the best stimulation frequency in order to efficiently modify the underlying network activity.
Table 3. rTMS studies of Post-traumatic Stress Disorder

Authors/Year Patients StimulationSide M1 bilateral r-DLPFC I-DLPFC rTMSFrequency 0,016 Hz 1 Hz 1 Hz and 5 Hz 1 Hz and 10 Hz StimulationIntensity 100 % MT 80 % MT 90 % MT Pulses per Session 30 20.400 6000 Study Design Open Open Open
Grisaru et al. 1998 McCann et al. 1998 Rosenberg et al. 2002 Cohen et al. 2004
10 2 21 24
r-DLPFC
80 % MT
1000 (1 Hz) 4000 (10 Hz)
Shamcontrolled; Double-blind
l-DLPFC: left dorsolateral prefrontal cortex; r-DLPFC: right dorsolateral prefrontal cortex; M1: primary motor cortex; MT: motor threshold.
5- OBSESSIVE-COMPULSIVE DISORDER
Obsessive compulsive disorder is characterized by recurrent intrusive thoughts, images, or feelings that lead to repetitive behaviors. The intrusions (obsessions) persist against the patient`s attempts to eradicate them and are accompanied by marked and overwhelming anxiety. The symptoms are associated with significant, and often dramatic, impairment in the abilities of affected individuals to carry out their social roles. In the early 1990s, the orbitofrontal cortex was identified as a major player in the etiopathogenesis of obsessive-compulsive symptoms (OCD). Further studies pointed to a significant hyperactivity both in orbitofrontal cortical areas and within the basal ganglia. These results are supported by studies interpreting OCD symptoms as a consequence of leftsided hyperexcitability of certain cortical structures including motor areas, causing a reduction of subcortical inhibitory mechanisms (Rossi et al., 2005). Magnetic resonance spectroscopy points to neuronal cell loss within the basal ganglia. Successful interventions with pharmaceutical agents or psychotherapy are associated with normalization of hyperactive frontal cortical areas and deep brain stimulation has been shown to act by modifying the activity of subcortical structures. Thus, rTMS may be a promising tool applied both over frontal cortex areas and over MI. MI or prefrontal ECS should be tested in refractory patients (Canavero S, personal observation).
298
Review of Clinical Trials Greenberg et al (1997) published the first study on rTMS treatment of OCD symptoms. Twelve patients were treated using a cross-over design with high-frequency 20-Hz stimulation over the right and the left DLPFC. Occipital rTMS served as a sham condition. Right-sided rTMS led to a significant clinical improvement in comparison with left-sided stimulation as well as with the sham condition. In this study 800 magnetic impulses were delivered inducing a therapeutic effect most prominent after eight hours of stimulation. In a double-blind placebo-controlled study, low-frequency rTMS over the right DLPFC did not cause a significant improvement after a six week stimulation period. In contrast, highfrequency rTMS over the right or left DLPFC induced a clinical improvement in four of 12 patients. A Cochrane analysis found no conclusive evidence to support rTMS in the treatment of OCD (Martin et al., 2003). In 2006, two additional rTMS studies were published. In a placebo-controlled study, low-frequency rTMS over the left DLPFC failed to induce a significant clinical improvement (Prasko et al 2006). As a limitation of this study, TMStreated patients showed significantly more severe clinical symptoms than sham-treated patients at the time point of inclusion. In the second study, low-frequency rTMS applied bilaterally over the supplementary-motor area (SMA) led to a significant reduction of OCD symptoms already after ten rTMS sessions (Mantovani et al., 2006). The clinical improvement was evident even three months after stimulation. This was associated with a significant increase in motor cortex thresholds, a neurophysiological parameter pointing to a reduction of axonal excitability in the setting of successful rTMS treatment (Table 4). Table 4. rTMS studies of Obsessive-Compulsive Disorder
Authors/Year Patients StimulationSide DLPFC bilateral rTMSFrequency 20 Hz StimulationIntensity 80 % MT Pulses per Session 800 Study Design
Greenberg et al. 1997
12
Alonso et al. 2001
18
r-DLPFC
1 Hz
110 % MT
21.600
Sachdev et al. 2001 Prasko et al. 2006
12 30
I-DLPFC r-DLPFC I-DLPFC
10 Hz 1 Hz
110 % MT 110 % MT
15.000 18.000
Shamcontrolled; Double-blind; Cross-over Shamcontrolled; Double-blind Double-blind Shamcontrolled; Double-blind Double-blind
Mantovani et al.
10
SMA-bilateral
1 Hz
100 % MT
12.000
DLPFC: dorsolateral prefrontal cortex; r-DLPFC: right dorsolateral prefrontal cortex; l-DLPFC: left dorsolateral prefrontal cortex; SMA: Supplementary motor area; MT: motor threshold.
Taken together, both high-frequency rTMS over frontal cortical areas and low-frequency rTMS over the SMA may be effective in treating OCD, with more rapid effects than psychosurgery. Nevertheless, studies with more patients are urgently needed to verify these
299
results. Since the SMA is closely connected with cortical areas encompassing the thalamus and the basal ganglia, neuroanatomical data suggest that this stimulation strategy may be especially effective in treating OCD. ECS remains to be explored.
6-CRAVING
Craving describes a state of intensive desire to consume and is an integral part of addictional behavior. This desire is hard to treat with centrally-acting drugs. Craving is associated with dysmodulated activity within the mesolimbic dopaminergic system as evidenced by Raclopride-PET in humans (Strafella et al., 2003). Dopamine drugs can trigger impulsivity, e.g. gambling and hypersexuality, and interfere with the ability to learn from bad experiences; parenthetically, subthalamic nucleus DBS promotes faster choices by silencing STN, which acts as a brake in decision making.
Review of Clinical Trials In a placebo-controlled, cross-over study high-frequency rTMS (20 Hz) was successful in reducing nicotine craving and decreasing cigarette smoking (Eichhammer et al., 2003). Uher et al (2005) too demonstrated reduction of food craving after high-frequency rTMS over prefrontal areas. Moreover, cocaine-craving can be transiently reduced using high-frequency rTMS over the right DLPFC (Camprodon et al., 2007). Three sham-controlled studies of transcranial direct current stimulation reported a reduction of smoking, food and alcohol craving (Boggio et al 2008, Fregni et al 2008a,b). Taken together, these studies point to a promising way in the treatment of addictive disorders. Future studies have to confirm these results using larger sample sizes and visualizing treatment effects by functional imaging. ECS might become an option.
REFERENCES
Alonso P, Pujol J, Cardoner N, et al. Right prefrontal repetitive transcranial magnetic stimulation in obsessive-compulsive disorder: a double-blind, placebo-controlled study. Am J Psychiatry 2001; 158:1143-1145. Avery DH, Claypoole K, Robinson L, et al. Repetitive transcranial magnetic stimulation in the treatment of medication-resistant depression: preliminary data. J Nerv Ment Dis 1999; 187:114-117. Avery DH, Holtzheimer PE, 3rd, Fawaz W, et al. A controlled study of repetitive transcranial magnetic stimulation in medication-resistant major depression. Biol Psychiatry 2006; 59:187-194. Berman RM, Narasimhan M, Sanacora G, et al. A randomized clinical trial of repetitive transcranial magnetic stimulation in the treatment of major depression. Biol Psychiatry 2000; 47:332-337.
300
Bocchio-Chiavetto L, Miniussi C, Zanardini R, et al. 5-HTTLPR and BDNF Val66Met polymorphisms and response to rTMS treatment in drug resistant depression. Neurosci Lett 2008; 437:130-134. Boggio PS, Rigonatti SP, Ribeiro RB, et al. A randomized, double-blind clinical trial on the efficacy of cortical direct current stimulation for the treatment of major depression. Int J Neuropsychopharmacol 2008;11:249-54. Boggio PS, Sultani N, Fecteau S, et al. Prefrontal cortex modulation using transcranial DC stimulation reduces alcohol craving: a double-blind, sham-controlled study. Drug Alcohol Depend. 2008;92:55-60. Boutros NN, Gueorguieva R, Hoffman RE, Oren DA, Feingold A, Berman RM. Lack of a therapeutic effect of a 2-week sub-threshold transcranial magnetic stimulation course for treatment-resistant depression. Psychiatry Res 2002; 113:245-254. Bretlau LG, Lunde M, Lindberg L, Undn M, Dissing S, Bech P. Repetitive transcranial magnetic stimulation (rTMS) in combination with escitalopram in patients with treatment-resistant major depression: a double-blind, randomised, sham-controlled trial. Pharmacopsychiatry 2008; 41:41-47. Brunelin J, Poulet E, Bediou B, et al. Low frequency repetitive transcranial magnetic stimulation improves source monitoring deficit in hallucinating patients with schizophrenia. Schizophr Res 2006; 81:41-45. Camprodon JA, Martinez-Raga J, Alonso-Alonso M, Shih MC, Pascual-Leone A. One session of high frequency repetitive transcranial magnetic stimulation (rTMS) to the right prefrontal cortex transiently reduces cocaine craving. Drug Alcohol Depend 2007; 86:9194. Chibbaro G, Daniele M, Alagona G, et al. Repetitive transcranial magnetic stimulation in schizophrenic patients reporting auditory hallucinations. Neurosci Lett 2005; 383:54-57. Cohen E, Bernardo M, Masana J, et al. Repetitive transcranial magnetic stimulation in the treatment of chronic negative schizophrenia: a pilot study. J Neurol Neurosurg Psychiatry 1999; 67:129-130. Cohen H, Kaplan Z, Kotler M, Kouperman I, Moisa R, Grisaru N. Repetitive transcranial magnetic stimulation of the right dorsolateral prefrontal cortex in posttraumatic stress disorder: a double-blind, placebo-controlled study. Am J Psychiatry 2004; 161:515-524. Cordes J, Arends M, Mobascher A, et al. Potential clinical targets of repetitive transcranial magnetic stimulation treatment in schizophrenia. Neuropsychobiology 2006; 54:87-99. DAlfonso AA, Aleman A, Kessels RP, et al. Transcranial magnetic stimulation of left auditory cortex in patients with schizophrenia: effects on hallucinations and neurocognition. J Neuropsychiatry Clin Neurosci 2002; 14:77-79. Dolberg OT, Schreiber S, Grunhaus L. Transcranial magnetic stimulation-induced switch into mania: a report of two cases. Biol Psychiatry 2001; 49: 468-470 Dougherty DD, weiss, AP, Cosgrove GR, et al. Cerebral metabolic correlates as potential predictors of response to anterior cingulotomy for treatment of major depression. J Neurosurg 2003; 99: 1010-1017 Eichhammer P, Johann M, Kharraz A, et al. High-frequency repetitive transcranial magnetic stimulation decreases cigarette smoking. J Clin Psychiatry 2003; 64:951-953.
301
Eschweiler GW, Wegerer C, Schlotter W, et al. Left prefrontal activation predicts therapeutic effects of repetitive transcranial magnetic stimulation (rTMS) in major depression. Psychiatry Res 2000; 99:161-172. Feinsod M, Kreinin B, Chistyakov A, Klein E. Preliminary evidence for a beneficial effect of low-frequency, repetitive transcranial magnetic stimulation in patients with major depression and schizophrenia. Depress Anxiety 1998; 7:65-68. Fitzgerald PB, Benitez J, Daskalakis JZ, et al. A double-blind sham-controlled trial of repetitive transcranial magnetic stimulation in the treatment of refractory auditory hallucinations. J Clin Psychopharmacol 2005; 25:358-362. Fitzgerald PB, Brown TL, Marston NA, Daskalakis ZJ, De Castella A, Kulkarni J. Transcranial magnetic stimulation in the treatment of depression: a double-blind, placebo-controlled trial. Arch Gen Psychiatry 2003; 60:1002-1008. Fitzgerald PB, Hoy K, McQueen S, et al. Priming stimulation enhances the effectiveness of low-frequency right prefrontal cortex transcranial magnetic stimulation in major depression. J Clin Psychopharmacol 2008; 28:52-58. Fregni F, Liguori P, Fecteau S, Nitsche MA, Pascual-Leone A, Boggio PS. Cortical stimulation of the prefrontal cortex with transcranial direct current stimulation reduces cue-provoked smoking craving: a randomized, sham-controlled study. J Clin Psychiatry. 2008a;69:32-40. Fregni F, Orsati F, Pedrosa W, et al. Transcranial direct current stimulation of the prefrontal cortex modulates the desire for specific foods. Appetite. 2008b;51:34-41 Garcia-Toro M, Mayol A, Arnillas H, et al. Modest adjunctive benefit with transcranial magnetic stimulation in medication-resistant depression. J Affect Disord 2001; 64:271275. Garcia-Toro M, Salva Coll J, Crespi Font M, et al. [Panic disorder and transcranial magnetic stimulation] Acta Esp Psiquiatr 2002; 30: 221-224 Geller V, Grisaru N, Abarbanel JM, Lemberg T, Belmaker RH. Slow magnetic stimulation of prefrontal cortex in depression and schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 1997; 21:105-110. George MS, Nahas Z, Molloy M, et al. A controlled trial of daily left prefrontal cortex TMS for treating depression. Biol Psychiatry 2000; 48:962-970. George MS, Wassermann EM, Kimbrell TA, et al. Mood improvement following daily left prefrontal repetitive transcranial magnetic stimulation in patients with depression: a placebo-controlled crossover trial. Am J Psychiatry 1997; 154:1752-1756. Greenberg BD, George MS, Martin JD, et al. Effect of prefrontal repetitive transcranial magnetic stimulation in obsessive-compulsive disorder: a preliminary study. Am J Psychiatry 1997; 154:867-869. Grisaru N, Chudakov B, Yaroslavski Y, Belmaker RH. TMS in mania: a controlled study. Am J Psychiatry 1998; 154: 867-869 Grisaru N, Amir M, Cohen H, Kaplan Z. Effect of transcranial magnetic stimulation in posttraumatic stress disorder: a preliminary study. Biol Psychiatry 1998; 44:52-55. Gross M, Nakamura L, Pascual-Leone A, Fregni F. Has repetitive transcranial magnetic stimulation (rTMS) treatment for depression improved? A systematic review and meta-
302
analysis comparing the recent vs. the earlier rTMS studies. Acta Psychiatr Scand 2007; 116:165-173. Guaiana G, Mortimer AR, Robertson C. efficacy of transcranial magnetic stimulation in panic disorder: case report. Aust N Z J Psychiatry 2005; 39: 1047 Haag C, Padberg F, Moller HJ. [Transcranial magnetic stimulation. A diagnostic means from neurology as therapy in psychiatry?]. Nervenarzt 1997; 68:274-278. Hajak G, Marienhagen J, Langguth B, Werner S, Binder H, Eichhammer P. High-frequency repetitive transcranial magnetic stimulation in schizophrenia: a combined treatment and neuroimaging study. Psychol Med 2004; 34:1157-1163. Hausmann A, Kemmler G, Walpoth M, et al. No benefit from repetitive trancranial magnetic stimulation in depression: a prospective, single centre, randomised, double blind, sham controlled add on trial. J Neurol Neurosurg Psych 2004; 75: 320-322 Herwig U, Schonfeldt-Lecuona C, Wunderlich AP, et al.The navigation of transcranial magnetic stimulation. Psychiatry Res 2001; 108: 123-131 Herwig U, Fallgatter AJ, Hoppner J, et al. Antidepressant effects of augmentative transcranial magnetic stimulation: randomised multicentre trial. Br J Psychiatry 2007; 191:441-448. Hoffman RE, Boutros NN, Hu S, Berman RM, Krystal JH, Charney DS. Transcranial magnetic stimulation and auditory hallucinations in schizophrenia. Lancet 2000; 355:1073-1075. Hoffman RE, Gueorguieva R, Hawkins KA, et al. Temporoparietal transcranial magnetic stimulation for auditory hallucinations: safety, efficacy and moderators in a fifty patient sample. Biol Psychiatry 2005; 58:97-104. Hoffman RE, Hawkins KA, Gueorguieva R, et al. Transcranial magnetic stimulation of left temporoparietal cortex and medication-resistant auditory hallucinations. Arch Gen Psychiatry 2003; 60:49-56. Holi MM, Eronen M, Toivonen K, Toivonen P, Marttunen M, Naukkarinen H. Left prefrontal repetitive transcranial magnetic stimulation in schizophrenia. Schizophr Bull 2004; 30:429-434. Holtzheimer PE, 3rd, Russo J, Claypoole KH, Roy-Byrne P, Avery DH. Shorter duration of depressive episode may predict response to repetitive transcranial magnetic stimulation. Depress Anxiety 2004; 19:24-30. Hoppner J, Schulz M, Irmisch G, Mau R, Schlafke D, Richter J. Antidepressant efficacy of two different rTMS procedures. High frequency over left versus low frequency over right prefrontal cortex compared with sham stimulation. Eur Arch Psychiatry Clin Neurosci 2003; 253:103-109. Januel D, Dumortier G, Verdon CM et al. A double-blind sham controlled study of right prefrontal repetitive transcranial magnetic stimulation (rTMS): therapeutic and cognitive effect in medication free unipolar depression during 4 weeks. Prog Neuropsychopharmacol Biol Psychiatry 2006;30:126130. Jin Y, Potkin SG, Kemp AS, Huerta ST, Alva G, Thai TM, et al (2006): Therapeutic effects of individualized alpha frequency transcranial magnetic stimulation (alphaTMS) on the negative symptoms of schizophrenia. Schizophr Bull 32:556-561.
303
Jorge RE, Robinson RG, Tateno A, et al. Repetitive transcranial magnetic stimulation as treatment of poststroke depression: a preliminary study. Biol Psychiatry 2004; 55:398405. Kimbrell TA, Little JT, Dunn RT, et al. Frequency dependence of antidepressant response to left prefrontal repetitive transcranial magnetic stimulation (rTMS) as a function of baseline cerebral glucose metabolism. Biol Psychiatry 1999; 46:1603-1613. Klein E, Kolsky Y, Puyerovsky M, Koren D, Chistyakov A, Feinsod M. Right prefrontal slow repetitive transcranial magnetic stimulation in schizophrenia: a double-blind shamcontrolled pilot study. Biol Psychiatry 1999; 46:1451-1454. Klein E, Kreinin I, Chistyakov A, et al. Therapeutic efficacy of right prefrontal slow repetitive transcranial magnetic stimulation in major depression: a double-blind controlled study. Arch Gen Psychiatry 1999; 56:315-320. Koerselman F, Laman DM, van Duijn H, van Duijn MA, Willems MA. A 3-month, followup, randomized, placebo-controlled study of repetitive transcranial magnetic stimulation in depression. J Clin Psychiatry 2004; 65:1323-1328. Kozel FA, Nahas Z., deBrux C, et al. How the distance from coil to cortex relates to age, motor threshold and possibly the antidepressant response to repetitive transcranial magnetic stimulation. J Neuropsychiatry Clin Neurosci 2000; 12: 376-384 Lee SH, Kim W, Chung YC, et al. A double blind study showing that two weeks of daily repetitive TMS over the left or right temporoparietal cortex reduces symptoms in patients with schizophrenia who are having treatment-refractory auditory hallucinations. Neurosci Lett 2005; 376:177-181. Loo C, Mitchell P, Sachdev P, McDarmont B, Parker G, Gandevia S. Double-blind controlled investigation of transcranial magnetic stimulation for the treatment of resistant major depression. Am J Psychiatry 1999; 156:946-948. Loo CK, Mitchell PB, Croker VM, et al. Double-blind controlled investigation of bilateral prefrontal transcranial magnetic stimulation for the treatment of resistant major depression. Psychol Med 2003; 33: 33-40 Manes F, Jorge R, Morcuende M, Yamada T, Paradiso S, Robinson RG. A controlled study of repetitive transcranial magnetic stimulation as a treatment of depression in the elderly. Int Psychogeriatr 2001; 13:225-231. Mantovani A, Lisanby SH, Pieraccini F, Ulivelli M, Castrogiovanni P, Rossi S. Repetitive transcranial magnetic stimulation (rTMS) in the treatment of obsessive-compulsive disorder (OCD) and Tourette's syndrome (TS). Int J Neuropsychopharmacol 2006; 9:95100. Martin JL, Barbanoj MJ, Perez V, Sacristan M. Transcranial magnetic stimulation for the treatment of obsessive-compulsive disorder. Cochrane Database Syst Rev:CD003387, 2003 McCann UD, Kimbrell TA, Morgan CM, et al. Repetitive transcranial magnetic stimulation for posttraumatic stress disorder. Arch Gen Psychiatry 1998; 55:276-279. McIntosh AM, Semple D, Tasker K, et al. Transcranial magnetic stimulation for auditory hallucinations in schizophrenia. Psychiatry Res 2004; 127:9-17.
304
Miniussi C, Bonato C, Bignotti S, et al. Repetitive transcranial magnetic stimulation (rTMS) at high and low frequency: an efficacious therapy for major drug-resistant depression? Clin Neurophysiol 2005; 116:1062-1071. Mogg A, Pluck G, Eranti SV, et al. A randomized controlled trial with 4-month follow-up of adjunctive repetitive transcranial magnetic stimulation of the left prefrontal cortex for depression. Psychol Med 2008; 38:323-333. Moncrieff J, Wessely S, Hardy R. Active placebos versus antidepressants for depression. Cochrane Database Syst Rev:CD003012, 2004 Mosimann UP, Schmitt W, Greenberg BD, et al. Repetitive transcranial magnetic stimulation: a putative add-on treatment for major depression in elderly patients. Psychiatry Res 2004; 126:123-133. Nahas Z, Kozel FA, Li X, Anderson B, George MS. Left prefrontal transcranial magnetic stimulation (TMS) treatment of depression in bipolar affective disorder: a pilot study of acute safety and efficacy. Bipolar Disord 2003; 5:40-47. Nahas Z, Molloy M, Risch SC, George MS. Transcranial magnetic stimulation in Neuuropsychiatry. American Psychiatric Press, 2000, pp 237-252. Novak T, Horacek J, Mohr P, et al. The double-blind sham-controlled study of highfrequency rTMS (20 Hz) for negative symptoms in schizophrenia: negative results. Neuro Endocrinol Lett 2006; 27:209-213. O'Reardon JP, Blumner KH, Peshek AD, Pradilla RR, Pimiento PC. Long-term maintenance therapy for major depressive disorder with rTMS. J Clin Psychiatry 2005;66:15241528. O'Reardon JP, Solvason HB, Janicak PG, et al. Efficacy and safety of transcranial magnetic stimulation in the acute treatment of major depression: a multisite randomized controlled trial. Biol Psychiatry 2007; 62:1208-1216. Padberg F, Zwanzger P, Keck ME, et al. Repetitive transcranial magnetic stimulation (rTMS) in major depression: relation between efficacy and stimulation intensity. Neuropsychopharmacology 2002; 27:638-645. Padberg F, Zwanzger P, Thoma H, et al. Repetitive transcranial magnetic stimulation (rTMS) in pharmacotherapy-refractory major depression: comparative study of fast, slow and sham rTMS. Psychiatry Res 1999; 88:163-171. Poulet E, Brunelin J, Bediou B, et al. Slow transcranial magnetic stimulation can rapidly reduce resistant auditory hallucinations in schizophrenia. Biol Psychiatry 2005; 57:188191. Poulet E, Brunelin J, Boeuve C, et al. Repetitive transcranial magnetic stimulation does not potentiate antidepressant treatment. Eur Psychiatry 2004; 19:382-383. Prasko J, Paskova B, Zalesky R, et al. The effect of repetitive transcranial magnetic stimulation (rTMS) on symptoms in obsessive compulsive disorder. A randomized, double blind, sham controlled study. Neuro- Endocrinol Lett 2006; 27:327-332. Rollnik JD, Huber TJ, Mogk H, et al. High frequency repetitive transcranial magnetic stimulation (rTMS) of the dorsolateral prefrontal cortex in schizophrenic patients. Neuroreport 2000; 11:4013-4015. Rosenberg PB, Mehndiratta RB, Mehndiratta YP, Wamer A, Rosse RB, Balish M. Repetitive transcranial magnetic stimulation treatment of comorbid posttraumatic stress disorder and major depression. J Neuropsychiatry Clin Neurosci 2002; 14:270-276.
305
Rossi S, Bartalini S, Ulivelli M, et al. Hypofunctioning of sensory gating mechanisms in patients with obsessive-compulsive disorder. Biol Psychiatry 2005; 57:16-20. Rossini D, Lucca A, Zanardi R, Magri L, Smeraldi E. Transcranial magnetic stimulation in treatment-resistant depressed patients: a double-blind, placebo-controlled trial. Psychiatry Res 2005; 137:1-10. Rumi DO, Gattaz WF, Rigonatti SP, et al. Transcranial magnetic stimulation accelerates the antidepressant effect of amitriptyline in severe depression: a double-blind placebocontrolled study. Biol Psychiatry 2005; 57:162-166. Saba G, Verdon CM, Kalalou K, et al. Transcranial magnetic stimulation in the treatment of schizophrenic symptoms: a double blind sham controlled study. J Psychiatr Res 2006; 40:147-152. Sachdev P, Loo C, Mitchell P, Malhi G. Transcranial magnetic stimulation for the deficit syndrome of schizophrenia: a pilot investigation. Psychiatry Clin Neurosci 2005; 59:354357. Sachdev PS, McBride R, Loo CK, Mitchell PB, Malhi GS, Croker VM. Right versus left prefrontal transcranial magnetic stimulation for obsessive-compulsive disorder: a preliminary investigation. J Clin Psychiatry 2001; 62:981-984. Sakkas P, Psarros C, Papadimitriou GN, Theleritis CG, Soldatos CR. Repetitive transcranial magnetic stimulation (rTMS) in a patient suffering from comorbid depression and panic disorder following acute myocardial infarction. Progr Neuropsychopharmacol Biol Psychiatry 2006; 30: 960-962 Schonfeldt-Lecuona C, Gron G, Walter H, et al. Stereotaxic rTMS for the treatment of auditory hallucinations in schizophrenia. Neuroreport 2004; 15:1669-1673. Schutter DJ. Antidepressant efficacy of high-frequency transcranial magnetic stimulation over the left dorsolateral prefrontal cortex in double-blind sham-controlled designs: a meta-analysis. Psychol Med 2008; 1-11. Speer AM, Kimbrell TA, Wassermann EM, et al. Opposite effects of high and low frequency rTMS on regional brain activity in depressed patients. Biol. Psychiatry 2000; 48, 1133 1141 Strafella AP, Paus T, Fraraccio M, Dagher A. Striatal dopamine release induced by repetitive transcranial magnetic stimulation of the human motor cortex. Brain 2003;126:26092615. Szuba MP, O'Reardon JP, Rai AS, et al. Acute mood and thyroid stimulating hormone effects of transcranial magnetic stimulation in major depression. Biol Psychiatry 2001, 50:22-27. Uher R, Yoganathan D, Mogg A, et al. Effect of left prefrontal repetitive transcranial magnetic stimulation on food craving. Biol Psychiatry 2005; 58:840-842. Zwanzger p, Minov C, Ella R, et al. Transcranial magnetic stimulation for panic. Am J Psychiatru 2002; 159: 315-316 Zwanzger P, eser D, Voelkel N, et al. Effects of repetitive transcranial magnetic stimulation (rTMS) on panic attacks induced by cholecystokinin-Tetrapeptide (CCK-4). Int J Neuropsychopharmacol 2007; 10: 285-289
EPILEPSY
Chapter 18
CORTICAL STIMULATION FOR MEDICALLY REFRACTORY EPILEPSY

David J. Mogul1,2,, Sergio Canavero3 and Ananda S. Fine1
University of Illinois, Chicago, IL, USA; Illinois Institute of Technology, Chicago, IL, USA 3 Turin Advanced Neuromodulation Group, Turin, Italy.
2 1
INTRODUCTION
Epilepsy is categorized as any of various disorders marked by repetitive aberrant electrical activity in the central nervous system and typically manifested by convulsive attacks known as seizures. It is estimated that approximately 0.5-1% of the worlds population (over 60 million people) may be afflicted with epilepsy, making it among the most common of neurological disorders. The hallmark of epilepsy is recurrent seizures. Recurrent seizures may occur as a result of a large number of causes and the underlying mechanisms frequently are not fully understood. If seizures cannot be controlled, the patient may experience major disruptions in family, social, educational and vocational activities that can have profound impacts on their quality of life. The mainstay of treatment is chronic medication based on modulation of cortical inhibition/excitation balance to prevent seizures. Anticonvulsant drugs help about twothirds of epileptic patients achieve effective seizure control, whereas the remaining onethird are refractory. About 10% of epileptic casesseveral million peoplecan be classified as medically intractable, unresponsive to drug treatments and without a defined epileptic focus amenable to surgical removal. Unfortunately, many patients develop a tolerance to the anticonvulsant effects causing a marked decrease in drug efficacy. In addition, these drugs frequently have many concomitant side effects such as dizziness, drowsiness, impaired vision, headache, mood change, rash and weight gain. For some patients, surgery may be another option by having the focus generating a partial seizure electrically mapped and surgically removed. While surgery is successful in
Correspondence concerning this article should be addressed to: Dr. David J. Mogul, e-mail: mogul@iit.edu.
310
David J. Mogul, Sergio Canavero and Ananda S. Fine

preventing seizures in about 8% of the total epileptic patients (Tellez-Zenteno et al 2005), surgical complications as well as neurological deficits such as memory loss or cognitive impairment are seen. Also, the excision of an epileptogenic focus becomes impractical or inefficacious in patients with multiple foci, generalized seizures or foci located too close to eloquent tissue (such as language centers). Finally, surgery techniques are anatomically irreversible. Such intractable epilepsy, both resistant to drug treatment and unsuitable for surgery, is a significant public health problem so that other alternative therapeutic approaches are needed. The capability of an applied electric field to influence the excitability of a neuron has been recognized as a potential treatment for epilepsy for over 20 years (Durand 1986). Stimulation can have the advantages of reversibility and adjustability for maximizing efficacy. In addition, it has the potential to be used in patients who would not otherwise be thought of as candidates for surgery: no brain tissue is destroyed and the stimulator can theoretically be adjusted to achieve the best outcome. It can also be turned off or removed if adverse side effects occur. However, in order to maximize efficacy of neural stimulation, it is important to consider the underlying dynamics of the system we wish to control, in this case, epilepsy. Unlike cardiac pacemakers that regulate a rhythmic heartbeat, the brain is a much more complicated dynamical system with trillions of neurons firing in complicated and asynchronous patterns that require much more complex stimulation protocols to manipulate. The extent to which stimulation will activate or inhibit neurons at various distances from the stimulating electrode is uncertain. Most structures of interest do not show spherical geometry around a stimulating electrode. Effects of stimulation change with intensity, frequency and duration of stimulation, among other factors. Therefore, stimulation at a particular site may inhibit seizures with some settings and provoke seizures with others. In open loop control (i.e. without any feedback information), the stimulation is kept periodically on and off following preset programming regardless of the underlying brain state. However, the exact nature and timing of these cycles are sometimes very critical. Under normal conditions in which no pathological state is present, chronically stimulated neurons could easily get fatigued under such long-term activation resulting in alteration of normal brain function. Alternatively, constant stimulation could lead to either an alteration in synaptic efficacy in the affected region thereby changing network characteristics in possibly a deleterious way such as by kindling new seizure activity or potentially by decreasing stimulation efficacy local to the electrodes. Studies of invasive and noninvasive cortical stimulation will be reviewed in this chapter.
1-THERAPEUTIC SUBDURAL CORTICAL STIMULATION

Penfield and Jasper (1954), using direct stimulation of the brain and simultaneous recording, were able to map out motor, sensory and language function in awake patients: they observed that, in some cases, acute focal electrical stimulation of the exposed cortex resulted in a flattening of the local electrocortico-graphy (both normal rhythms and spontaneous epileptiform discharges). More recently, human experimental trials sought to use stimulation to abort seizures by means of subdural electrodes placed for the purpose of localizing the epileptic focus for
Cortical Stimulation for Medically Refractory Epilepsy
311
surgical excision; the clinical outcome pursued was a decrease in number of afterdischarges (AD) or an increase in AD threshold. An AD is any abnormal electrical activity (i.e. differing from background activity), usually repetitive, that occurs as a result of stimulation of the cortex (Blume et al 2004). These ADs can vary significantly between patients and even within a patient, depending on the area stimulated. These may be considered electrographic seizures; i.e. seizures that may or may not mimic the electrographic behavior of the patients typical seizures and which are not accompanied by clinical signs or symptoms. ADs are often elicited during the course of localizing stimulation which can interfere with proper cortical mapping and potentially induce seizures; thus a secondary protocol to decrease this physician-induced interictal epileptiform activity would be beneficial. However, stimulation-induced clinical seizures may occur without observation of ADs. Yamamoto et al (2002) investigated the effect of low-frequency cortical stimulation on the number of interictal epileptiform discharges in a patient with intractable mesial temporal lobe epilepsy. This group used a 0.9 Hz biphasic square-wave stimulus with 0.3 msec duration for a total of 250 seconds of stimulation at current intensities of 0.5, 2.0, and 7.5 mA. In this study it was found that interictal epileptiform discharges decreased at a stimulus intensity of 0.3 mA with a maximal decrease in discharges after the third stimulus session. However, after the fourth and fifth sessions, the number of discharges actually began to increase, but not to pre-stimulus levels. Once stimulation ceased, however, discharge numbers returned to pre-stimulus levels. Stimulation at current densities higher than 0.5 mA elicited an aura in this patient and thus were not further studied. This same group (Kinoshita et al 2004) attempted both low (0.9 Hz) and high (50 Hz) frequency stimulation at 1.0 and 7.0 mA intensities of 0.3 msec duration in a patient with medically refractory partial seizures of neocortical origin. This study measured spike frequency and electrocorticogram (ECoG) power spectrum changes due to stimulation at the electrodes showing focal seizure activity, as well as distal electrodes using pre-stimulation ECoG activity as baseline comparison for activity and spectral analyses. 50 Hz stimulation at the epileptic focus resulted in a 48.3% reduction in spike frequency which was significant within 10 minutes of stimulation and between 15-20 minutes of stimulation. Furthermore, these stimulations resulted in suppression of fast EEG activity and a significant decrease in ECoG power in the 22-26 Hz range that lasted up to 10 minutes after stimulation. This decrease in power was only observed in the area directly stimulated and not adjacent cortical tissue. Low frequency stimulation resulted in a 23.9% reduction in spiking activity which was significant 10-20 minutes after stimulation. ECoG power was not significantly affected by low frequency stimulation. When stimulation was attempted at electrodes distal to the epileptic focus, no change in spiking activity was observed. However, at a frequency of 50 Hz, ECoG power in the 18 28 Hz range was significantly decreased within 15 minutes of stimulation. It is important to note that in this study, beneficial effects of high frequency stimulation did not occur immediately after stimulation, as seen following low-frequency stimulation, but instead took several minutes. Also, high-frequency stimulation effects were long-lasting (up to 20 minutes after stimulation). After these initial, exploratory studies, this group went on to further study low (0.9 Hz, bipolar stimulation for 15 minutes) and high (50 Hz) frequency stimulation (50 Hz bipolar
312
alternating square pulse of 0.3 ms duration for high frequency stimulation, starting with 1 mA for 1 second and increased up to 15 mA for 5 seconds; total stimulus duration shorter than 224 seconds) at the electrodes over the seizure focus in eight patients with medically refractory focal epilepsy (Kinoshita et al 2005, Yamamoto et al 2006). In the earlier study, ECoG power and spike frequency were measured before and after stimulation, while in the later study interictal spiking and seizure frequency were measured. In these two studies, a spike was defined as a clear transient sharp wave with 0-200 msec duration and > 200 V amplitude, with a slow-wave following. These studies found that interictal spiking activity was significantly decreased by low (18.5% mean reduction) and high (-24.7% mean reduction) frequency stimulation, with low frequency effects seen within 5-10 minutes after stimulation and high-frequency effects within 10-15 minutes (up to 20 minutes after stimulation). ECoG power decreased due to 50 Hz stimulation of both the epileptic and non-epileptic tissue; this decrease persisted up to 20 minutes after stimulation. 0.9 Hz stimulation reduced ECoG power only when the focal area was stimulated and this effect was transient. Interestingly, this decrease in power at the stimulated site was positively correlated with the total delivered charge: the more charge delivered at the electrode, the larger the reduction in ECoG power while the power change at distal electrodes was negatively correlated with total charge delivered. With repeated lowfrequency stimulation at the focal electrode at high current intensity (15 mA) and 250 second blocks of stimulation, these authors found a decrease in number of seizures after the third stimulation session in one patient. This decrease was not observed at a stimulus intensity of 0.5 mA (in fact, these stimulus parameters seemed to elicit more seizures in this patient). They concluded that 50 Hz stimulation, having a quick effect (0-5 minutes), would be indicated for fast suppression of ongoing seizures, whereas a gradual reduction of spikes which occurred at 10 minute after low frequency stimulation suggested its usefulness in fine modulation of partial seizures. Schrader et al (2006) attempted to abort a case of refractory status epilepticus in a 26year old female patient who developed five independent seizure foci (both depth and neocortical) following several days of febrile illness. Control of seizures in this patient required 2-4 antiepileptic drugs as well as continuous delivery of two anesthetics. Due to the poor prognosis, cortical stimulation was initiated. In this case, two to four 30 minute sessions (each separated by 10 15 minutes) of low frequency (0.5 Hz, 500 s, 16 mA) stimulation was applied to the ictal onset area through subdural electrodes for seven days. After one day of stimulation, one antiepileptic drug was removed and one was reduced without seizure recurrence. In addition, stimulation led to decreased frequency and cortical distribution of interictal epileptiform activity. Although this stimulation allowed for a reduction of seizure medication and control of some of the epileptic foci, the patient continued to seize from other areas and was thus submitted to resective surgery. After surgery, the patient continued to seize and required higher doses of antiepileptic drugs and unfortunately never again regained consciousness. Despite the poor outcome in this patient, this study demonstrated that this stimulation protocol can at least diminish the effects of seizure activity. The studies outlined thus far in this section were uncontrolled studies aimed at simply changing the frequency or occurrence of spontaneous epileptiform activity. A slightly different approach was taken by a US group (Lesser et al 1999). These investigators actually initiated ADs with electrical
313
stimulation and used a second stimulation protocol to attempt to abort them. In this study, 19 patients were implanted with subdural electrodes as part of presurgical evaluation for resective surgery. Localizing stimulation parameters consisted of 0.3 msec, 50 Hz biphasic square waves at up to 15.0 mA intensity applied for 4 5 seconds. When this stimulation led to ADs, a stimulus burst was applied in order to terminate them. The stimulus burst parameters were the same as those used for localizing stimulation, except that the burst train lasted only 0.3 2.0 seconds (median 0.5 secs). This short burst of stimulation led to a decrease in AD number within 2 seconds of stimulation; however the effect was transient. In contrast to the other studies described, ADs were divided into three types:1) rhythmic epileptiform activity (semisinusoidal), 2) rapid spiking with no rhythmicity, and 3) discrete discharges (spikes or polyspikes). In general, the brief stimulation burst was most successful at stopping rhythmic ADs; longer durations (1.5 2.0 seconds) of pulse trains were still more effective. Importantly, if there were interictal epileptiform discharges recorded near the stimulation electrode, the brief burst of stimulation was less effective. Those electrodes that recorded interictal discharges tended to also display longer duration ADs which were less effectively aborted by the brief stimulus burst. Motamedi et al (2002) continued this approach of delivering a brief burst of stimulation to abort ADs in seven patients implanted with subdural electrodes. This study found that this stimulation protocol was 8.6 times more likely to stop ADs if delivered at the electrodes recording the strongest AD signal. In addition, if this stimulation was delivered within 4.5 seconds of AD onset, the stimulus was more effective at terminating ADs (effective in 67% of the trials if delivered within 4.5 secs versus 46% at 4.5 secs or more). Interestingly, the phase of the AD waveform was critical to the success of stimulation for AD termination. If stimulation was delivered during the negative phase of the AD waveform, it was almost twice as likely to stop the AD than if it were delivered at any other time. In other words, the efficacy was higher during the depolarization phase of the waveform (as the depolarized region is likely to be smaller and to be synchronous throughout), and much less during the refractory phase between individual discharges: as the active epileptiform region enlarges, some sites will be depolarizing, while others are entering a refractory period, making control less effective. Furthermore, if the ADs were more rhythmic in pattern, the brief burst was most successful. Lastly, the brief bursts of pulse stimulation was more successful when ADs did not begin immediately after the end of the initial localizing stimulation . These authors entertained the possibility that an implanted closed loop device could both detect and abort epileptiform activity (see chapter 19). Shown below is a summary of the stimulation parameters utilized in the above reviewed studies (Table 1).
2-NONINVASIVE CORTICAL STIMULATION

Investigators have attempted to use low-frequency rTMS to treat seizure disorders and other manifestations of cortical hyperirritability, based on its ability to reduce the excitability of the motor cortex. Noninvasive stimulation may provide an attractive alternative for
314
medication resistant disorders, or for individuals who are either unable (i.e. epileptic focus in the eloquent cortex) or wish not to have surgery. Table 1. Summary of stimulation parameters
Authors Peak Curren t (mA) 15 Pulse Duration (msec) 1, 0.2 Frequency (Hz) Stimulus Duration (sec) 2, 5 Current Density/Phase (/cm2/ph) NR Outcome Measure
Motamedi et al 2007 Szelenyi, et al 2007 Schrader et al 2006 Yamamoto et al 2006
50, 100
129
40
0.19 - 0.39
256 - 512
train of 5 - 7 stimuli 1800 250
1130
1 4
16 15
0.5 0.3
0.5 0.9
14.2 - 28.3 NR
Kinoshita et al 2005
15
0.3
50
NR
Kinoshita et al 2004
0.3
50
NR
Motamedi et al 2002 Yamamoto et al 2002
17.5
0.3
50
0.3 - 2
NR
0.5, 2, 7.5
0.3
0.9
250
NR
Lesser et al 1999 Gordon et al 1990
17
17.5
0.3
50
0.3 - 2
NR
13.6, 15.0
0.3
50
2 - 124
52 - 57
Control of stimulationinduced ADs Incidence of stimulationinduced seizures Suppression of status epilepticus Control of interictal epileptiform activity Effect of stimulation on interictal spike frequency and ECoG power Effect of stimulation on interictal spike frequency and ECoG power Control of stimulationinduced ADs Effect of stimulation on number of interictal epileptiform discharges Control of stimulationinduced ADs Histopathologica l Damage
NR: not reported; AD: afterdischarge; N: number of patients.
In an open trial in 9 medicated patients with frequent and intractable seizures, Tergau et al (1999) found that 5 days of 1000 pulses of 0.3 Hz rTMS delivered at 100% of the MEP threshold appeared to reduce self-rated seizure frequency that lasted for weeks. Menkes and Gruenthal (2000) reported a single patient with a cortical dysplasia which was the intended target of 0.5 Hz rTMS at 95% of MT, delivered with a large round coil twice a week for 4 weeks. During the treatment period, there was a 70% reduction in the frequency of seizures
315
and a 47% reduction during the first month of treatment, with gradual relapse after 2 months. Encouraged by pilot results (Wedegaertner et al 1997), Wassermann (unpublished observations 2001) treated several cortical myoclonus patients with 10 days of 1 Hz rTMS over MI and occipital sham stimulation at just above the MEP threshold, without significant beneficial effects. Theodore et al (2002) performed a randomized controlled trial in 24 patients and found that active 1 Hz rTMS (900 pulses, 120% MT, bid for 1 week) applied to the epileptogenic focus (as indexed by EEG) did not significantly alter the number of seizures in epileptic subjects as compared to a sham control group, although there was a trend toward improvement (p = 0.06). Recruited subjects also had epileptogenic zones of neocortical origin, but located in deep structures, such as the mesial temporal area. Indeed, patients with neocortical foci had a greater reduction in seizure frequency as compared to those with mesial temporal foci. The magnitude of the field induced by TMS drops as a cubic function of the vertical distance between a surface coil and target neural structures; the in-plane field fall-off is even greater. This suboptimal stimulation for some subjects may have caused the overall study to not reach significance. A similar study was reported by an Italian group (Daniele et al 2003). Fregni et al (2006, 2007) performed randomized, sham-controlled trials using similar parameters as the study of Theodore et al., but recruited subjects whose epileptic foci were of neocortical origin (heterotopias, polymicrogyrias, dysplasias). Using 1 Hz stimulation for 20 min over a five-day period (1200 pulses, 70% max stimulator output), they found a significant decrease in the number of seizures between the active rTMS group and those receiving sham treatment. This effect was apparent both in the days following, as well as more than two months after, treatment, thus showing a long-lasting modulatory effect. Kinoshita et al (2005b) treated 7 patients with frontal lobe epilepsy with 0.9 Hz rTMS (two sets of 15 min stimulation daily, for 5 sessions in a week at 90% MT). The frequency of all seizure types was reduced, in particular complex partial seizures (- 35.9%, up to - 88.5% in a single patient), although this did not reach statistical significance. Joo et al (2007) enrolled 35 patients with localization-related epilepsy who had experienced at least one complex partial seizure or a secondarily generalized seizure per week on a constant anti-epileptic drug regimen over an 8-week period. rTMS was administered at 0.5Hz for 5 consecutive days at 100% of rMT (resting motor threshold). Patients were divided into a focal stimulation group with a localized epileptic focus, or a nonfocal stimulation group with a non-localized or multifocal epileptic focus. These two groups were then randomly subdivided into four subgroups depending on the total number of stimulations administered, i.e., 3000 pulses and 1500 pulses subgroups. Weekly seizure frequencies were determined for 8 weeks before and after rTMS. To compare the number of interictal spikes before and after rTMS, EEG was recorded twice before (1st day) and after rTMS (5th day). Mean weekly seizure frequency was non-significantly decreased after rTMS (8.4-->6.8/week, -13.9%). Longer stimulation subgroups (3000 pulses, -23.0%) tended to have fewer seizures than shorter stimulation subgroups (1500 pulses, -3.0%), without statistical significance. TMS stimulation site and structural brain lesions did not influence seizure outcome. However, interictal spikes significantly decreased (-54.9%, P=0.012) after rTMS and they totally disappeared in 6 patients (17.1%, 6/35).
316
All in all, these preliminary results are promising, but are clearly not conclusive. Online EEG recording-guided rTMS, in which the pulses of TMS could be precisely coordinated with the epileptic spike for maximal efficacy and for prevention of seizure, could improve results.
3-CHRONIC INVASIVE CORTICAL STIMULATION

The epileptogenic zone may be defined as: the minimum amount of cortex that must be resected (inactivated or completely disconnected) to achieve seizure freedom. Five zones have been described as those most pertinent to identify during pre-surgical evaluation: 1) irritative zone the area from which spontaneous interictal spikes originate, 2) seizure onset zone the area from which spontaneous seizures arise, 3) symptomatogenic zone the area from which seizure symptoms and/or signs may be elicited via stimulation, 4) epileptogenic lesion an area which is either grossly damaged (due to cortical dysplasia or tumor, for example) or is secondarily excitable due to nearby lesioned tissue, and 5) function deficit zone an area of cortex that functions abnormally interictally (Luders et al 2006). The extent of the epileptogenic zone may or may not coincide with the zone of ictal onset; in most cases, the epileptogenic zone is larger than that of the ictal onset. Furthermore, the epileptogenic zone is often dynamic; that is, it may migrate over the course of the patients disease process or it may remain static. Obviously, a static epileptogenic zone will be easier to control via focal cortical stimulation. In deciding the goal for therapeutic electrical stimulation of the cortex, the main consideration is to determine what facet of the epileptic network is to be affected by stimulation. While most of the studies outlined in section 1 pertain exclusively to the control of ADs, it may also be useful for future studies to address the possibility of controlling other regions in the epileptogenic zone. In addition, the dynamic nature of epileptogenic zones should be kept in mind as the primary area for seizure control may migrate over the course of the patients disease progress. Finally, contacts should be less than 2 cm apart for effective epileptic control (Gwinn et al 2008). In light of these considerations, invasive cortical stimulation becomes an option in select cases. There is only one report in the literature (Elisevich et al 2006). These authors implanted a strip electrode (Resume quadripolar lead) subdurally along the axis of the dorsolateral convexity of the precentral gyrus (MI) in a man with intractable postencephalitic epilepsy with ictal onset in MI. For the five years of follow-up, cortical stimulation (2.1 mA, PW 450 s, 50 Hz, ON 3 min/OFF10 min), with minimal alterations, successfully eliminated the jacksonian march and secondary generalization and reduced seizure frequency and intensity, with an immediate postictal return of motor function. Over time, the seizure frequency subsided by more than 90%, without adverse effects. One can speculate that closed-loop MCS would be even more effective. Modeling of temporal lobe epileptogenicity suggests that a small redistribution of cerebral blood flow from non-epileptic to epileptic cortex should produce substantial reduction in temporal lobe seizure frequency in association with prolongation of interhemispheric prolongation time (Weinand 2000). Blood flow perturbations can be easily achieved with cortical stimulation and as such may put this theory to test.
317
CONCLUSION
Considerably more clinical experience has been obtained for treating epilepsy with deepbrain rather than cortical stimulation, even though the latter is easier to accomplish since electrode placement on rostral brain structures is less complicated surgically. However, data suggest that, for patients with focal lesions, the dysfunctional brain area should be targeted directly, whereas, in patients with discordant lesion location and epileptiform focus, the functional focus may be targeted first, but this suggestion requires validation. Patients with multifocal epilepsy or primary generalized epilepsy may be best treated by targeting a bihemispheric central area at the vertex using a large circular coil, as suggested by PascualLeone. A higher fraction of seizure foci in pediatric epileptic cases occurs in the neocortex compared to deeper structures, thus cortical stimulation may ultimately have a higher impact on the treatment of this younger patient cohort. Antiepileptic drugs can interfere with cortical stimulation and trials should test for the clinical effects of this interaction. The use of exogenous stimulation for the treatment of epilepsy is still in its infancy. Possible neuromodulatory effects of repetitive direct brain stimulation have been shown. EEG changes occur even if the current is not directly over the epileptogenic cortex. However, the significant number of patients with seizures that remain refractory to conventional therapeutics strongly suggests that stimulation of the cortex will continue to see progressively greater interest both in the research laboratory and in the clinic.
REFERENCES
Blume WT, Jones DC & Pathak P. Properties of after-discharges from cortical electrical stimulation in focal epilepsies. Clin Neurophysiol 2004; 115, 982-989. Daniele O, Brighina F, Piazza A, Giglia G, Scalia S, Fierro B. Low-frequency transcranial magnetic stimulation in patients with cortical dysplasia- a preliminary study. J Neurol 2003; 250: 761-762. Durand D.Electrical stimulation can inhibit synchronized neuronal activity. Brain Res 1986; 382: 139-144. Elisevich K, Jenrow K, Schuh L, Smith B. Long-term electrical stimulation-induced inhibition of partial epilepsy. Case report. J Neurosurg 2006; 105: 894-897. Fregni F, Otachi PT, DoValle A, et al.A randomized clinical trial of repetitive transcranial magnetic stimulation in patients with refractory epilepsy. Ann Neurol 2006; 60: 447-455. Fregni F, Otachi P, Valle AC, Rigonatti SP, Pascual-Leone A, Valente KD. A randomized and sham-controlled study of rTMS in patients with cortical malformations and refractory epilepsy. Stereotact Funct Neurosurg 2007; 85: 44. Gwinn RP, Spencer DD, Spencer SS, et al. Local spatial effect of 50Hz cortical stimulation in humans. Epilepsia 2008 (in press). Joo EY, Han SJ, Chung SH, Cho JW, Seo DW, Hong SB. Antiepileptic effects of lowfrequency repetitive transcranial magnetic stimulation by different stimulation durations and locations. Clin Neurophysiol 2007; 118: 702-708.
318
Kinoshita M, Ikeda A, Matsumoto R, et al. Electric stimulation on human cortex suppresses fast cortical activity and epileptic spikes. Epilepsia 2004; 45, 787-791. Kinoshita M, Ikeda A, Matsuhashi M, et al. Electric cortical stimulation suppresses epileptic and background activities in neocortical epilepsy and mesial temporal lobe epilepsy. Clin Neurophysiol 2005; 116: 1291-1299. Kinoshita M, Ikeda A, Begum T, Yamamoto J, Hitomi T & Shibasaki H. Low-frequency repetitive transcranial magnetic stimulation for seizure suppression in patients with extratemporal lobe epilepsy-a pilot study. Seizure 2005b; 14, 387-392. Lesser RP, Kim SH, Beyderman L, et al. Brief bursts of pulse stimulation terminate afterdischarges caused by cortical stimulation. Neurology 1999; 53, 2073-2081. Luders H, Najm, I, Nair, D, Widdess-Walsh, P, Bingman, W. The epileptogenic zone: general principles. Epileptic Disorders 2006; 8, S1-S9. Menkes D.L., Gruenthal M. Slow-frequency repetitive transcranial magnetic stimulation in a patient with focal cortical dysplasia. Epilepsia 2000; 41: 240-242. Motamedi GK, Lesser RP, Miglioretti DL, et al. Optimizing parameters for terminating cortical afterdischarges with pulse stimulation. Epilepsia 2002; 43, 836-846 (Erratum in: Epilepsia 2002;43:1441). Penfield W, Jasper, H. Epilepsy and the Functional Anatomy of the Human Brain. Little and Brown, Boston, MA, 1954. Schrader LM, Stern JM, Wilson CL, et al. Low frequency electrical stimulation through subdural electrodes in a case of refractory status epilepticus. Clin Neurophysiol 2006; 117, 781-788. Tergau F, Naumann U, Paulus W, Steinhoff BJ. Low-frequency repetitive transcranial magnetic stimulation improves intractable epilepsy. Lancet, 1999; 353: 2209. Tellez-Zenteno JF, Dhar R & Wiebe S. Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-analysis. Brain 2005; 128, 1188-1198. Theodore WH, Hunter K, Chen R, et al. Transcranial magnetic stimulation for the treatment of seizures: a controlled study. Neurology 2002; 59: 560-562. Wedegaertner F.R., Garvey M.A., Cohen L.G., et al. Low-frequency repetitive transcranial magnetic stimulation can reduce action myoclonus. Neurology, 1997; 48, A119. Yamamoto J, Ikeda A, Satow T, et al. Low-frequency electric cortical stimulation has an inhibitory effect on epileptic focus in mesial temporal lobe epilepsy. Epilepsia 2002; 43, 491-495. Yamamoto J, Ikeda A, Kinoshita M, et al. Low-frequency electric cortical stimulation decreases interictal and ictal activity in human epilepsy. Seizure 2006; 15, 520-527.
Chapter 19
CLOSED-LOOP STIMULATION FOR CONTROL OF FOCAL EPILEPSY

J. Smith1,, K. Fountas2, M. Murro3, Y. Park3, P. Jenkins1, M. Morrell4, E. Esteller4 and S. Canavero5
1
Department of Neurosurgery, Medical College of Georgia, Augusta, Georgia, USA; 2 Department of Neurosurgery, University of Thessaly School of Medicine, Larisa, Greece; 3 Department of Neurology, Medical College of Georgia, Augusta, Georgia, USA; 4 Neuropace, Mountainview, CA, USA; 5 Turin Advanced Neuromodulation Group, Turin, Italy.
INTRODUCTION
Neurologists have known for a long time that some patients can predict their own seizures well in advance, and certain subsets of patients can do this quite reliably. These 'preictal' changes are often undetectable on EEG either because they consist of relatively small, intermittent changes in the EEG signal, or because they occur beyond the frequency or spatial resolution of the EEG systems currently used in clinical practice. However, these findings prompted research into seizure detection algorithms, now under development for more than 10 years (Stacey and Litt 2008, Hughes 2008). As of the time of this writing, prospective seizure prediction has not yet been convincingly demonstrated, but definitive evidence of statistically significant seizure prediction may be imminent (Stacey and Litt 2008). The theoretical benefit is that, if seizure generation can be identified long before it is manifested clinically, it is likely that the process is more spatially confined, and might be more amenable to abortive therapy. Over the past 10 years, many strategies for analyzing and predicting seizures have been evaluated, including many nonlinear and chaos measurements, wavelet decompositions, machine learning, and other methods (Hughes 2008). The results have
Correspondence concerning this article should be addressed to: Dr. J. Smith, MD, e-mail: jsmith@mail.mcg.edu.
320
J. Smith, K. Fountas, M. Murro et al.

been somewhat inconsistent, and to date no method has been successful in prospective tests. The need to implement such algorithms is strictly associated with the development of so-called closed loop devices for seizure termination. These devices sense an oncoming seizure and release an electric jolt that terminates the fit. Feedback enables real-time correction if the intervention is insufficient and this may lead to a reduction in the overall treatment dose, thereby reducing adverse effects and system wear. Earlier still interventions might even prevent clinical seizures from occurring, rather than trying to abort a de facto seizure. To date, this strategy has not been implemented in any of the devices under investigation. One dilemma in seizure prediction is how to set the threshold for false positives. Current opinion states that a high sensitivity is preferable in order to ensure that no seizures are missed, even if stimulations are triggered for false positives. The rationale behind this approach is that responsive stimulations are selected to be harmless, i.e. they are below the threshold at which tissue injury is induced and the total treatment dose administered is lower than that for open-loop devices. Thus, the essence of a closed-loop device is to deliver less-frequentbut hopefully more-effectiveinterventions than open-loop devices. Given the heterogeneity of causes of epilepsy, many investigators feel that it is likely that seizure detection and prediction methods will be improved if they are tuned to each individual patient. In fact, both detection and prediction of seizures as defined in an individual patient are difficult to generalize to all patients. It is important to establish what recording bandwidth will be necessary and sufficient for accurate seizure detection and prediction, and on what spatial scale implantable devices need to operate (Stacey and Litt 2008). In addition, pulse trains, as currently delivered by electrical devices, are simple to implement, but are not necessarily the best stimulation method. Recently, the first and only- responsive, closed loop device for epilepsy control has been submitted to a controlled trial in view of FDA approval, the NeuroPace Responsive NeuroStimulator RNS device (NeuroPace, Inc., Mountain View, CA), which incorporates algorithms which are manually trained to individual patient patterns and updated as necessary at subsequent doctor visits. In contrast to open-loop stimulation, such as vagal nerve stimulation (VNS), which makes seizure-free less than 2% of the patients at 1 year, closed-loop or responsive stimulation aims to suppress epileptiform activity by delivering stimulation directly in response to the electrographic activity. In this chapter, we will review responsive stimulation therapy for epilepsy and describe surgical implantation and preliminary results of the NeuroPace system.
RESPONSIVE STIMULATION THERAPY FOR EPILEPSY

Early experiments to demonstrate the efficacy of responsive stimulation in humans were acute trials of stimulation performed on patients already undergoing intracranial mapping to localize seizure onset prior to epilepsy surgery (see chapter 18). Peters et al. (2001) delivered closed-loop stimulation in response to seizure detection. The system includes a Neoped 4000 EEG collection system (Nicolet BMSI, Madison, WI), two personal computers, custom software for electroencephalography (EEG) analysis using Visual C++ (Version 6.0; Microsoft, Seattle, WA) and MATLAB (Version 5.0; Mathworks, Natick, MA), two Grass S12 stimulators (Grass-Telefactor, West Warwick, RI), and other custom-built units to interface the components. A specific detection algorithm for this system
Closed-loop Stimulation for Control of Focal Epilepsy
321
using a wavelet transform and median filter was developed to detect events early and in real time. This algorithm resulted in earlier detection and shorter detection latencies compared to commercially available detection algorithms, which at the time were designed to detect the body and not the onset of an event. The median detection latency was 3.6 s for the generic algorithm. By adapting the filter to the individual subjects seizures, the detection latency was further reduced. In the five subjects with the greatest median detection delay, the adaptive approach reduced the median delay from 12.3 to 4.8 s. In almost all cases in which the electrographic onset preceded the clinical onset, the detection occurred several tens of seconds before the clinical onset. The system was evaluated in a trial of eight patients undergoing intracranial monitoring (Osorio et al 2005). The authors applied high frequency stimulation (= />100 Hz) with a pulse width of 100 s/phase, inside the safety limit of 30 microC/cm2 per phase (Agnew and McCreery 1990). High frequency stimulation for the local closed loop group was 1 second (with one redelivery possible within 1 minute and no more than five re-stimulations per detection cluster). The duration was 30 seconds in the remote closed loop group. For the local group mean frequency was 251 Hz, mean intensity 5 mA and mean duration 1 second, whereas for the remote group, these were 151 Hz, 5.4 mA and about 22 seconds. In four patients, responsive stimulation was delivered directly into the epileptogenic zone; in the other four patients, responsive stimulation was delivered into the anterior thalamic nucleus. This study demonstrated that automated responsive delivery of stimulation in close temporal proximity to the seizure onset was practicable. The study also provided preliminary evidence that high-frequency responsive stimulation could be efficacious, with three of four responders (>50% seizure reduction) among patients who received stimulation to the epileptogenic zone. Effects were immediate and outlasted the stimulation period. Kossoff et al (2004) studied the efficacy of an external responsive stimulator (NeuroPace) with the intent to investigate whether an implantable device with similar capabilities (internal responsive neurostimulator) could reduce or abort spontaneous seizures. This open trial assessed clinical and electrographic effects of cortical stimulation applied to persons hospitalized and implanted with intracranial electrodes for purposes of localization. Stimulation was provided by an external device that detected electrographic seizures and applied preprogrammed stimulation. The detection tools were designed to require low computational power, so that they could be implemented in an implantable device. The detection tools were also highly configurable, to provide early and specific detections individualized for each patient. When detection occurred, the device could deliver responsive therapy consisting of biphasic charge balanced electrical pulses (frequency 1-200 Hz, intensity: 0.5-12 mA, pulse width: 40-1000 s, combinations of the 8 connected contacts, burst duration no more than 5 seconds and 1-5 bursts). After a pulse train was delivered, a redetection algorithm determined whether the epileptiform activity was still present: if so, up to 4 additional bursts could be delivered per episode. The study reported on four patients who were treated with responsive stimulation for up to 68 h (range, 668 h). In all patients, responsive stimulation appeared to be safe and well tolerated. Two patients reported brief transient side effects, such as sensations in the tongue, facial tingling, and visual flashes; however, the sensations stopped after stimulation parameters were changed. Although the study was not designed to assess efficacy, stimulation appeared to reduce the number of
322
clinical seizures in both temporal and extratemporal locations, and electrographic seizures were altered and suppressed during the period when the patients received responsive stimulation. In one patient, stimulation also appeared to improve the baseline electrocorticography. Murro et al (2003) found that, of 27 patients studied with the external neurostimulator system, 11 (41%) exhibited a positive EEG effect on electrographic seizure activity and no serious adverse events related to the external neurostimulator system occurred. In sum, automated contingent delivery of high frequency electrical stimulation in close temporal proximity to seizure onset is practical in real time and reliable and may be efficacious and safe (see also Nair et al 2004). A carryover effect may be noted. Direct stimulation of the epileptic zone may (or may not) be superior to indirect stimulation. Early detection and accurate lead placement may be key to successful responsive stimulation. The optimal stimulation frequency for seizure termination has not been established and could vary depending on the anatomic and pathologic areas treated. The risk of activating seizure foci, as observed in TMS studies (Hufnagel and Elger 1991, Dhuna et al 1991) is possibly very low. Stimulation parameters for creating kindled seizures differ from the high-frequency, short-duration stimulation delivered to human patients. In fact, this type of stimulation may provide what is called kindling antagonism, whereby stimulation in one area can inhibit the development of fully kindled seizures from another site (Bertram 2007).
THE NEUROPACE SYSTEM

1-General Description Experience with these external responsive neurostimulators led to the development of the first implantable responsive neurostimulator for epilepsy, the NeuroPace RNS system. This system is capable of performing real-time seizure detection and delivering responsive electrical stimulation. The implantable components of the system include a cranially implanted neurostimulator and intracranial depth and strip leads. The neurostimulator is a battery-powered, microprocessor-controlled device that continuously monitors electrographic activity from the cortical depth and strip leads and delivers programmable electrical stimulation in response to detected events to two distinct epileptogenic zones independently. Due to techonological constraints, the sensed electrographic data are not continuously recorded; however, the neurostimulator can store segments of the electrographic data for review by the physician. The neurostimulator system has a 32-minute ECoG memory buffer. The number of ECoGs stored depends on the number of recording channels and the recording time selected. Typically, 2 bipolar recording channels are selected with a 60-second pretrigger and 30-second posttrigger duration, which allows 9 ECoGs to be stored. Any additional ECoGs overwrite the previous recordings. ECOG storage can be triggered by any of several electrographic events, including seizure onset. An important feature is the use of a personalized 'training period', in which the device is individually tuned to the patient after it has recorded seizures.
323
The RNS neurostimulator is designed to match typical skull thickness and curvature, and is intended for implant in a ferrule, or socket, placed in a craniectomy. Up to two leads, each containing four electrodes can be connected to the neurostimulator; each may be a depth lead or cortical (subdural) strip lead, and each has four electrode contacts, which are used for sensing and providing stimulation. To provide early seizure detection and delivery of focal electrical stimulation, leads are positioned as close as possible to the seizure focus or foci. The external components of the system include a physician programmer, a patient data transmitter and a telemetry wand. The programmer is used by the physician to program detection and stimulation settings and retrieve stored information (e.g., electrographic activity) from the neurostimulator. The data transmitter is provided to the patient to allow uploading and remote monitoring of the device between clinic visits. The telemetry wand allows wireless communication between the neurostimulator and the programmer (or data transmitter). The device data and electrographic data are uploaded via the Internet to a central patient data management system and may be reviewed by physicians using a secure Web browser. Electrographic data storage can be triggered by detection, responsive stimulation, scheduled time of day, magnet (used by the patient to indicate a clinical event), or other events as programmed by the physician. These data allow physicians to assess detection sensitivity and effects of stimulation. The detection algorithms implemented in the RNS system are designed to be computationally efficient and are highly optimized to perform real-time seizure detection within the constraints of currently available implantable technology, such as limited power and processing capabilities. Three detection tools are provided: area, line-length, and halfwave. The detection tools are highly configurable and can be adjusted by the physician to optimize the sensitivity and specificity trade-off for each individual patient. Up to two independent detectors can be programmed for any two sensing channels. The half-wave tool (Gotman 1982) is used to detect spikes and rhythmic activity occurring in specific frequency ranges. The half-wave tool segments the electrographic signal at local minima and maxima, resulting in half-waves, the amplitude and duration of which are representative of the amplitude and frequency components of the EEG. Half-waves that exceed a physician-programmed amplitude and duration are counted; the number of these half-waves occurring within a given window length must exceed a certain threshold for detection to occur. Detection parameters may be selected to adjust the sensitivity, specificity, and latency of the detection. For example, a detector may be configured for early detection (within a fraction of a second) with as few as two qualified half-waves, whereas a more specific detector may require several qualified half-waves that occur over a few seconds. The line-length algorithm (Esteller et al 2001, DAlessandro et al 2003) is used to identify changes in both amplitude and frequency, but is more commonly used to detect high frequency, low amplitude activity that diverges in amplitude little from the baseline electrographic amplitude, but has significant summed line length. The line length is defined as the average of absolute sample-to-sample differences within a window. A short-term sliding window average (128 ms to 4 s) is compared to a long-term sliding window average (4 s to 16 min). Detection occurs when the short-term measurement exceeds an absolute or relative threshold, compared with the long-term measurement. A negative threshold can also
324
be used to detect decreases in line length, which may represent a period of electrodecrement or decreased frequency. The area feature, similar to an energy or power measurement (Litt and Echauz 2002), is used to identify changes in overall signal energy without regard for frequency. It is more commonly used for slower rhythmic higher amplitude epileptiform activity with large integrated areas. Area is defined as the average absolute area-under-the-curve within a window. As with line length, a short-term window average is compared to a long-term background window average, and detection occurs when a positive or negative threshold is exceeded. The area and line-length detection algorithms are efficient (requiring low computational power) and can be configured to detect events within a fraction of a second or to detect more subtle changes in amplitude, frequency, or power that occur over several seconds. The neurostimulator delivers current-controlled, charge-balanced biphasic pulses; it can be programmed by the physician to deliver stimulation frequencies ranging from 1 to 333 Hz, current amplitudes from 1 to 12 mA, and pulse-widths from 40 to 1000 s. Any of the electrode contacts or the pulse generator housing may be programmed as anode or cathode. The stimulation montage can be configured to deliver current between any combination of electrodes and the neurostimulator case. Up to five individually configured sequential therapies of electrical stimulation may be programmed, where each therapy is composed of two independently configurable bursts. The RNS will attempt to redetect the epileptiform activity after each therapy is delivered. If the epileptiform activity is still detected, the next (sequential) therapy will be delivered. If the epileptiform activity is no longer detected, the remaining therapies will not be delivered and the episode ends. The therapy sequence will refresh with the detection of each new episode. Stimulation parameters are determined empirically, while respecting safety limits for charge density. The parameters of each therapy and each burst may be the same or varied. The RNS system has controls in place at the programmer and neurostimulator, so that current densities remain below a conservative limit of 25 C/cm2 per phase. Furthermore, intermittent (more acute-like) stimulation may pose fewer risks to neural tissue than does continuous stimulation (Agnew and McCreery 1990, Gordon et al 1990). The stimulation configuration is also determined empirically. Options include providing bipolar stimulation across electrode pairs or stimulating across all eight electrodes to the case of the neurostimulator. Whether stimulation is delivered to a few or many electrodes depends to some extent on the area of onset.
2-Candidates Candidates for responsive neurostimulation implantations have histories and EEG findings consistent with drug-resistant focal epilepsy. They are neither candidates for resective surgery due to the location of the focus in functional cortex, nor desire it. All candidates are required to have an average of 4 disabling seizures per month over a 3-month period before final consideration for implantation.
Closed-loop Stimulation for Control of Focal Epilepsy 3-System Components
325
Pulse Generator (IPG). The IPG is a hermetically sealed neurostimulator containing electronics, battery, telemetry coil, and connector hardware that accommodates one or two 4-contact electrodes. The IPG continuously analyzes the patient's electrocorticogram (EcoG) and triggers electrical stimulation when it detects specific ECoG characteristics programmed by the clinician as indicative of interictal or ictal epileptiform activity. The IPG then stores diagnostic information that details detections and stimulations, including multichannel stored ECoGs. The IPG has the following dimensions: 41 mm wide, 60 mm long and 7 mm thick (weight 19.5 g, volume 10.5 cm3). The IPG is curved in shape to facilitate cranial implantation. It is positioned extradurally in a tailored cranial defect and held in place with a ferrule (Figure. 1: craniectomy with the implanted ferrule; Figures 1b: IPG in place).
Figure 1a. Rectangular craniectomy with implanted ferrule. Arrow points to the burr hole which will accommodate the fixation tab used to secure the IPG in the ferrule (Figure 1b).
Figure 1b. Ferrule secured in position with self tapping screws in the four tangs at the corners of the ferrule. Note the fixation tab (arrow) which is accommodated by the burr hole. Note single depth electrode (short arrow) inserted in connector port
326
Electrodes. Depth electrodes are quadripolar and designed for stereotactic implantation. They are available with 3.5mm and 10mm intercontact spacings, with lengths of 30 cm and 44 cm. The individual contacts are 1.1 mm in diameter and 2.3 mm in length. Subdural strip electrodes are quadripolar with 4 mm diameter circular electrodes and intercontact spacings of 10 mm. They are available in 15 cm and 25 cm lengths. All electrodes are composed of platinum (90%) and iridium (10%). Programmer - The programmer is a laptop computer with specialized software and a telemetry wand that communicates with the IPG. The programmer downloads diagnostic and ECoG data from the IPG. In addition, the programmer has an electrophysiology test stimulation mode, which allows real-time stimulation with simultaneous ECoG viewing. Figure 2 shows an example of an epileptiform discharge that was successfully aborted before it developed into an electrographic seizure.
Figure 2. Successfully aborted electrographic seizure. Left arrows point to the onset of seizure detection (DO) and the right arrow points to the therapeutic stimulation (Tr). Lower tracing is a zoomed-in view of the bracketed portion of the upper tracing.
4-Technique of Implantation Currently, we utilize invasive monitoring in all cases to determine seizure focus localization and, therefore, optimization of the location and type of closed loop electrodes used. For example, patients with a seizure focus in the left lateral temporal neocortex would receive two 1x4 subdural electrodes (Figure 3a). If the seizure focus were bilateral hippocampal, right and left 1x4 (10 mm contact spacing) depth electrodes would be implanted (Figure 3b). If the seizure focus were left hippocampal in a patient in whom Wada memory testing revealed absence of contralateral memory support, a 1x4 hippocampal depth electrode and an ipsilateral anterior subtemporal 1x4 subdural electrode would be implanted
327
(Figure 3c). In general, the type and location of electrode implantation will be tailored to the individual patient. Implantation site for the IPG will be determined by the implantation site of the two electrodes. If depth electrodes alone are used, a large skin flap will be needed to expose enough skull to incorporate the IPG and its ferrule as well as the two 14 mm diameter burr holes and their burr hole rings and covers (Figure 3b). If there was a previous craniotomy performed for invasive monitoring, this area may be re-exposed for implantation of both the subdural electrodes and the IPG and ferrule (Figure 3a). If a hippocampal depth electrode is implanted through an occipital entry in a case requiring repeat temporal craniotomy or repeat burr hole for implantation of a subdural electrode (Figure 3c), then the tail of the depth electrode will need to be tunneled into the area of cranial exposure. Passing the tail of the depth electrode through the appropriate length of silastic tubing provided in the depth electrode kit will protect the electrode tail when the old or new craniotomy incision is subsequently made for subdural strip electrode and IPG placement.
Figure 3a. Example of RNS with 2 4-contact subdural strip electrodes implanted over the left superior and middle temporal gyri. The three arrows (left to right) point to the connector port, microchip, and battery of the IPG.
Figure 3b. Example of RNS with bilateral 4-contact hippocampal depth electrodes implanted in a case with bilateral independent hippocampal foci.
328
Figure 3c. Example of RNS with one 4-contact depth and one 4-contact subdural strip electrode implanted in a case with left hippocampal seizure onset and absence of contralateral memory support as determined by previous Wada test.
Figure 4a. The entry point of the one inch, 25 gauge needle (just superior to the lateral aspect of the right eyebrow) is used for anterior scalp field block.
The patient is appropriately positioned on the operating table by attaching the base-ring to a Mayfield holder. The craniotomy incision is marked off with gentian violet and the surgical area is prepped and draped in a standard sterile fashion. The previously marked-off incision is now infiltrated with local anesthetic and a horseshoe-shaped craniotomy flap is turned. A sterile ferrule template is provided than can be onlayed over the exposed skull to find an area where the convex contour of the template best fits that of the skull (diminishing
329
the potential risk of bone erosion). We use the monopolar cutting current to mark off this rectangular area. A single burr hole is placed at either end of the rectangle and a Penfield 3 dissector is used to separate the dura from the skull prior to using the craniotome to cut out the rectangle of bone. We prefer a full thickness craniectomy in order to minimize protrusion of the IPG. Note that one burr hole of the craniectomy should be off center to accommodate the fixation tab that secures the IPG in the ferrule (Figures 1a and b). We place depth electrodes using a frame-based stereotaxic system and carry out depth electrode implant planning with a commercially available computer workstation. Frame application is carried out using i.v. sedation consisting of Alfentanyl and Propofol. A field block of the scalp is performed using approximately 40 cc of a 1:1 mixture of 0.25% Marcaine and 0.5% Lidocaine (figures 4a and 4b). Carbon fiber posts and aluminum pins are used to limit metal artifacts on the CT scan, which follows the MRI study. A contrast-enhanced, T-1 weighted, volume acquisition MRI is performed consisting of 1.3 mm axial slices with 0 mm slice gap. The MRI is followed with non-contrasted axial CT with 3 mm axial slices and 0 mm slice gap. The images are transferred over the local area network to the computer workstation which is used to perform image fusion of the MRI study to the CT study. After depth electrode trajectories are generated (figure 5a), a probe view algorithm is used to assess (1) the proximity of any cortical vessels to the depth electrode entry and (2) the proximity of the trajectory to the ventricular system or any subependymal veins (figure 5b). Any appropriate adjustments to trajectories are made before actual implantation. The stereotactic arc system is attached to the base ring and a drill guide tube is then advanced through the incision down to the skull and antibiotic irrigation is flushed through the tube. The appropriate burr hole is outlined on the skull and a high-speed air-driven drill is used. The underlying dura is opened in a linear fashion in order to facilitate subsequent passage of the guide block. At this point, a 2.1 mm inner-diameter guide block is introduced and the dura and pia are cauterized with a monopolar electrode. The diameter of the burr hole must exactly match the diameter of the securing device for the implanted depth lead. Next a 14-gauge depth electrode cannula is passed through the same guide block to the target point. Intraoperative fluoroscopy verifies proper placement. The cannula and guide tube are then withdrawn. A cranial base and cap device are then implanted into the burr hole and secured by using the two self-tapping screws provided. The surgeon must align the exit groove on the base in the posterolateral direction, in the same direction that the subcutaneous portion of the lead would later be directed. An insertion tool is then passed through a large diameter guide block and inserted into the slot and the securing device. The insertion tool and guide block are then removed. The depth lead is then carefully inserted into the target point. After fluoroscopy confirmation, the stylet of the lead is removed and the distal shaft of the implanted lead is secured into the securing device. Postimplant intraoperative fluoroscopy is performed after locking the depth electrodes in place in order to assure proper positioning (figure 6). If the subdural strip lead is to be placed, the dura is opened linearly and the appropriate cortical lead is inserted through the dural opening under fluoroscopy. The distal shaft of the implanted lead is safely attached to the securing device. At this point, the provided ferrule is placed on the exposed bone and the desired bony defect is outlined. The outlined bone is drilled out by high-speed air-driven drill, bony edges are smoothened, irrigated and waxed. The provided ferrule is then implanted and secured to the adjacent bone at four points with the provided self-tapping
330
mini-screws. The IPG is connected to the distal end of the already implanted lead or leads and then is safely secured in the implanted ferrule. To summarize, burr holes for depth electrodes are made with a 14 mm diameter bit so that the manufacturer supplied burr hole rings, which are secured with self-tapping screws, will exactly fit. Subdural electrodes are secured with purse string 4-0 silk suture at the dural incision exit site. If the depth electrode implant is bilateral, the entire procedure is done with the patient in the stereotaxic frame using local anesthesia and i.v. Propofol and Alfentanyl. If the implant is unilateral, we tunnel the tail of the depth electrode to the planned craniotomy site, close the burr hole incision, and remove the frame. The area of the subsequent craniotomy incision will determine the area of skull to be clipped (avoid razors to minimize skin microtrauma and possible infection). The patient is intubated and ventilated, but the procedure may also be carried out under neuroleptoanalgesia along with field blocks of the scalp. Control of blood pressure minimizes potential intraoperative hemorrhage. The operative blood loss is generally minimal (<100 ml). At the end of the procedure, all components of the RNS neurostimulator system are connected, the skin flap is positioned for closure, and a telemetry wand (covered with a
Figure 4b. The entry point for the one inch, 25 gauge needle (half way between the root of the mastoid and the inion) used to block the posterior scalp.
331
Figure 5a. The right and left hippocampal depth electrode trajectories are displayed. The view in the lower left corner shows that the trajectory did not breach the ventricular system.
Figure 5b. The left hippocampal depth electrode trajectory is displayed. The view in the lower left corner shows the cortical entry point (arrow) which is a safe distance from adjacent cortical veins.
332
Figure 6. Postimplant fluoroscopy of a 4-contact hippocampal electrode (the deepest contact is directly superior to the sella turcica).
sterile camera drape) is placed over the IPG so that recording of electrode contact impedances as well as intraoperative electrocorticography (ECoG) can be performed. If impedances and ECoG recordings are satisfactory, skin closure is completed. The surgical wound is irrigated with an antibiotic solution and then closed in anatomical layers and the wound is covered with a sterile dressing. Perioperative i.v. antibiotic therapy is begun the morning of the procedure and continued 24 hours post-operatively. On the third postoperative day, the first interrogation-detection session is performed.
5-Initial Results In a feasibility study of 65 patients, the NeuroPace RNS system demonstrated excellent safety and tolerability, as well as preliminary evidence of efficacy. Based on an interim analysis of 39 subjects, there were no serious device-related unanticipated adverse events. Stimulation-related symptoms in several subjects were addressed by adjusting the stimulation settings. No subjects experienced sustained seizure types of greater severity or convulsive status epilepticus. There was preliminary evidence for efficacy in the initial 24 subjects with complete data. The responder rate (50% reduction in seizures) was 36% for complex partial seizures, 50% for generalized tonic-clonic seizures and 36% for total disabling seizures. The median percentage reductions in seizure frequency were respectively 28%, 50% and 30% (Smith et al 2007). Experience at the University of Georgia in 9 patients shows a 67%
333
responder rate with a mean reduction of 77% (median 75%) (Smith et al 2007). Time to IPG replacement due to battery depletion has been approximately 1-2 years. A randomized, double-blinded, multicenter, sham-controlled clinical trial is currently underway ( Clinicaltrials.gov NCT00264810 ).
CONCLUSIONS
External responsive neurostimulation studies have shown that closed-loop stimulation can significantly effect duration of spontaneously occurring electrographic seizure activity (Peters et al 2001, Murro et al 2002, 2003). Early seizure detection and stimulation of multiple contacts in the immediate vicinity of the seizure focus have improved the frequency of suppression of epileptiform activity. Observations on the currently described closed loop neurostimulator system also support the ability of this automated seizure detection/therapeutic stimulation device to positively influence electrographic seizure activity.
REFERENCES
Agnew WF, McCreery DB (Eds) Neural prostheses: fundamental studies. Englewood Cliff NJ: Prentice-Hall, 1990 Bertram E. The relevance of kindling for human epilepsy. Epilepsia 2007; 48 Suppl 2: 65 74 Dhuna A, Gates J, Pascual-Leone A. Transcranial magnetic stimulation in patients with epilepsy. Neurology 1991; 41: 1067-1071 Esteller R, Echauz J, Tcheng T, Litt B, Pless B. Line length: an efficient feature for seizure onset detection. Engineering in Medicine and Biology, 2001, EMBS 01, Proc 23rd Annu Int Conf IEEE, 2001; 2: 17071710 DAlessandro M., Esteller R., Vachtsevanos G., Hinson A., Echauz J., Litt B., Epileptic seizure prediction using hybrid feature selection over multiple intracranial EEG electrode contacts: a report of four patients, [Erratum in: IEEE Trans Biomed Eng 2003;50:1041] IEEE Trans Biomed Eng 2003; 50: 603615 Gordon B, Lesser R, Rance N et al. Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation. EEG Clin Neurophysiol 1990; 75: 371-377 Gotman J. Automatic recognition of epileptic seizures in the EEG. Electroencephalogr Clin Neurophysiol. 1982; 54: 530-540 Hufnagel A, Elger CE. Responses of the epileptic focus to transcranial magnetic stimulation. EEG Clin Neurophysiol Suppl. 1991; 43: 86-99 Hughes JR. Progress in predicting seizure episodes with nonlinear methods. Epilepsy & Behavior 2008; 12: 128-135 Kossoff E.H., Ritzl E.K., Politsky J.M., et al. Effect of an external responsive neurostimulator on seizures and electrographic discharges during subdural electrode monitoring. Epilepsia 2004; 45: 15601567
334
Litt B, Echauz J. Prediction of epileptic seizures. Lancet 2002; 1: 22-30 Motamedi GK, Lesser RP, Miglioretti DL, et al. Optimizing parameters for terminating cortical afterdischarges with pulse stimulation. Epilepsia 2002; 43:836-846. Murro AM, Park YD, Bergey GK, et al. Multicenter study of acute responsive stimulation in patients with intractable epilepsy. Epilepsia 2003;44 (Suppl 9):326 Nair DR, Matsumoto R, Lders HO, Burgess R, Bingaman W. Direct cortical electrical stimulation in the treatment of epilepsy. In: Lders HO (ed) Deep brain stimulation and epilepsy. London, Martin Dunitz, 2004 Osorio I, Frei MG, Sunderam S, et al. Automated seizure abatement in humans using electrical stimulation. Ann Neurol 2005; 57: 258-268 Peters TE, Bhavaraju NC, Frei MG, Osorio I. Network system for automated seizure detection and contingent delivery of therapy. J Clin Neurophysiol 2001;18:545-549 Smith JR, Murro AM, Park YD, et al. Follow-up results with a closed loop stimulation system in the control of focal epilepsy. Stereotact Funct Neurosurg 2007; 85: 43 Stacey WC, Litt B. Technology Insight: Neuroengineering and Epilepsy--Designing Devices for Seizure Control. Nat Clin Pract Neurol. 2008;4:190-201
TINNITUS
Chapter 20
TMS FOR REFRACTORY TINNITUS

B. Langguth1,, D. de Ridder2, P. Eichhammer1 and G. Hajak1
1
Department of Psychiatry and Psychotherapy, University of Regensburg, Germany; 2 BRAIN & Department of Neurosurgery, University of Antwerp, Belgium.
INTRODUCTION
Any lesion along the auditory tract, interfering with its ascending or descending pathways can generate tinnitus. In the western world, prolonged tinnitus requiring treatment affects 10% to 15% of the population (Heller 2003). Mild tinnitus, intermittent or constant (i.e. without auditory habituation) is more prevalent affecting 25% of the population in their twenties and up to 35% in their seventies (Sindhusake et al 2003). Severe tinnitus is age related, affecting 2% of the population in their twenties, 6% in their 50s, 10% in their 70s. Severe tinnitus leads to depression in approximately half of the patients, to insomnia in 40% and leads to an important decrease in the quality of life in 20% (Cronlein et al. 2007; Langguth et al. 2007a). The prevalence increases in noiseexposed workers to 24%. People presenting with left sided tinnitus complain more than people with right sided tinnitus (Hallberg and Erlandsson 1993), possibly explaining why tinnitus seems to be more predominant on the left. Tinnitus evaluation concerns 2 main items: tinnitus intensity, i.e. how loud the tinnitus is perceived, and the amount of distress and annoyance and its impact on daily living. Usually tinnitus intensity is estimated by application of a subjective Visual Analogue Score (VAS) for intensity and distress and a number of self-report tinnitus questionnaires. Objectively quantifying tinnitus is attempted by matching or masking methods, but its quantification lacks standardization. Moreover, there is a poor correlation between the intensity of the tinnitus as assessed by matching techniques and the degree of annoyance the tinnitus creates. Tinnitus is usually well tolerated if its intensity is below 3 on a VAS, but becomes very distressing once it is over 3/10 (Van de Heyning, unpublished results). The fact that tinnitus intensity is poorly related to the
Correspondence concerning this article should be addressed to: Dr. Berthold Langguth, MD, e-mail: Berthold.Langguth@medbo.de.
338
B. Langguth, d. de Ridder, P. Eichhammer et al.

perceived annoyance has resulted in the idea that emotional factors play an important role in the way tinnitus is perceived. Tinnitus is very difficult to treat (Dobie 1999). Auditory stimulation and cognitive behavioral treatment may improve habituation and coping strategies. However, more causally oriented treatment strategies are needed.
1. RATIONALE FOR THE APPLICATION OF RTMS IN TINNITUS

There is growing consensus that dysfunctional neuroplastic processes in the brain are involved in the genesis of chronic tinnitus (Moller 2001, 2003, 2007a,b; Eggermont 2007). In particular, chronic tinnitus could be an auditory phantom perception, the correlate of maladaptive attempts of the brain at reorganization due to distorted sensory input (Jastreboff 1990; Moller 2007a). Magnetoencepholo-graphy (MEG) demonstrated a shift in the tonotopic map of the auditory cortex contralaterally to the tinnitus (Muhlnickel et al 1998). Positron emission tomography (PET) showed asymmetry in the auditory cortices of tinnitus patients with higher levels of spontaneous neuronal activity on the left side, irrespective of tinnitus laterality (Arnold et al 1996; Kleinjung et al 2005, Langguth et al 2006b). Other imaging studies revealed additional changes in the middle temporal and temporoparietal regions as well as activation in frontal and limbic areas (Lockwood et al 1998; Giraud et al 1999; Mirz et al 2000; Plewnia et al 2007a). EEG and MEG studies in humans have demonstrated that tinnitus is associated with reduced alpha and increased gamma activity in the auditory cortex (Dohrmann et al 2007, Weisz et al 2007a,b): gamma band activity in the auditory cortex is known to correlate with the conscious perception of auditory signals. MEG data indicate that the amount of tinnitus-related emotional distress correlates with the degree of synchronicity between temporal and frontal areas (Weisz et al 2007a,b) and that the tinnitus-related cortical network changes over time with a more widespread distribution and less relevance of auditory areas in longer tinnitus duration (Schlee et al 2008). According to a speculation, thalamic deafferentiation from auditory input, e.g. due to hearing loss, may produce slow theta-frequency oscillations in thalamocortical ensembles due to changes in firing patterns of thalamic relay cells. Decreased cortical lateral inhibition could generate high-frequency activity in the gamma band, a possible correlate of the tinnitus percept (Weisz et al 2007a,b), and may induce cortical plastic reorganization with attendant alterations of the tonotopic map (Eggermont 2007). Neuroimaging data (Lockwood et al 1998, Giraud et al 1999, Mirz et al 2000, Plewnia et al 2007a) indicate that nonauditory brain areas are also involved in the pathophysiology of chronic tinnitus. Tinnitus distress is related to activation of the right amygdala, right anterior cingulate (ACC), right anterior insula and orbitofrontal cortex. This has recently been suggested by results derived from a very small pilot study on 5 patients with unilateral left sided narrow band noise tinnitus (Van der Loo, De Ridder D, in preparation). A t-test independent-sampled comparison between patients distressed or not by their tinnitus was performed on Low Resolution brain Electric Tomography (LORETA) EEGs. Results showed significant higher activation in the low beta band (13-18Hz) in a right-sided distress network consisting of the amygdala, anterior cingulate, anterior insula and orbitofrontal
TMS for Refractory Tinnitus
339
cortex extending into BA10. Notably, the tinnitus distress network and the pain matrix are identical. Table 1. Effects of single sessions of rTMS on tinnitus
Authors /date Stimulation site / coil positioning Various scalp positions according to 10-20 EEG system Frequency Intensity Pulses 10 Hz 120% MT 30 Number of patients 14 Study Design Outcome Notes
Plewnia et al., 2003
De Ridder et al., 2005
Auditory cortex contralateral to tinnitus site
1, 5, 10, 20 Hz 90% MT 200
114
Controlled study; control condition: stimulation of non-auditory cortical areas Controlled study; control condition: coil angulation
In 8 patients (58%) tinnitus suppression after left temporal/temporopa rietal stimulation In 60 patients (53 %), good or partial tinnitus suppression after active rTMS; in 33% suppression after sham rTMS
First study investigating rTMS for the suppression of tinnitus The large sample allowed to determine a correlation between tinnitus duration and most efficient stimulation frequency Only available study comparing rTMS with tDCS
Fregni et al., 2006
Left temporoparietal areas, according to 10-20 EEG system, anodal and cathodal tDCS of the same area
10 Hz 120% MT 30
Controlled study; control conditions: sham coil and active stimulation of mesial parietal cortex
Folmer et al., 2006
Left and right temporal cortex, according to 10-20 EEG system
10 Hz 100% MT 150
15
Controlled study; control condition: sham coil
In 3 patients (42%), tinnitus suppression after left temporoparietal stimulation, no effect for both control rTMS conditions; anodal tDCS resulted in tinnitus suppression in the same 3 patients, cathodal tDCS had no effect In 6 patients (40%) tinnitus suppression after active rTMS (in 4 after contralateral rTMS, in two after ipsilat. TMS); in 2 patients suppression after sham rTMS
340
B. Langguth, d. de Ridder, P. Eichhammer et al. Table 1. (Continued)
Authors /date
Londero et al., 2006
Stimulation site / coil positioning Auditory cortex as determined by fMRI,
Frequency Intensity Pulses 1, 10 Hz 120% MT 30
Number of patients 13
Study Design
Outcome
Notes
Controlled study; control condition:
De Ridder et al., 2007
Area of maximum tinnitus related PET activation (temporoparieta l cortex), neuronavigatio n Auditory cortex contralateral to tinnitus site
1 Hz 120% MT 300, 900, 1800
Controlled study; control condition: control position (occipital cortex) Controlled study; control condition: Coil angulation
in 5 patients (38%) tinnitus suppression after 1 Hz rTMS of the auditory cortex; no suppression after 10 Hz; in 2 patients suppression after stimulation of a control position In 6 patients (75%), tinnitus reduction after active rTMS; better suppression with more pulses
Lasting effects of single sessions of 1 Hz rTMS
The only study which systematically investigated the effect of pulse numbers
5, 10, 20 Hz tonic; 5, 10, 20 Hz burst 90%MT 200
46
14 placebonegative patients were analyzed: in those with narrow band/white noise tinnitus, burst TMS was more effective as compared to tonic TMS, whereas for pure tone tinnitus no difference was found between burst and tonic.
The first study which systematically investigated burst TMS and compared the results to tonic TMS
MT: Motor threshold.
2. CLINICAL EFFECTS OF RTMS IN TINNITUS

In recent years, single sessions of rTMS have been performed in order to transiently disrupt tinnitus perception for diagnostic and heuristic purposes (Table 1). Trains of highfrequency rTMS (10 20 Hz) have been administered in order to induce an immediate, shortlasting interruption of tinnitus perception. On the other hand, repeated sessions of lowfrequency (1 Hz) rTMS on several consecutive days (Table 2) have been administered for therapeutic purposes.
A- Studies Using Single Sessions of rTMS
341
Plewnia et al (2003) investigated 14 chronic tinnitus patients in order to find out whether focal stimulation of different cortical areas can transiently suppress tinnitus. High-frequency rTMS (10 Hz) was applied to twelve different positions over the scalp. A significant reduction of tinnitus was observed when stimulation was administered to the left temporoparietal cortex. This result suggests that the secondary auditory areas (Brodmanns area (BA) 42 and 22) are critically involved in tinnitus perception. In a large series of 114 patients with unilateral tinnitus, de Ridder et al (2005) investigated tinnitus suppression by different frequencies of rTMS (between 1 and 20 Hz). The coil was localized over the auditory cortex contralateral to the site of the tinnitus. The amount of tinnitus suppression correlated positively with stimulation frequency and negatively with tinnitus duration. Higher frequency was more effective than lower frequency on tinnitus of short duration and, vice versa: long-standing tinnitus was more difficult to suppress. Two studies (Fregni et al 2006; Folmer et al 2006) confirmed the findings of transient tinnitus reduction after high frequency stimulation of the left temporoparietal cortex. Both found similar response rates of patients (42% and 40%) compared to Plewnias (57%) and de Ridders (53%) studies. In addition, Fregni et al found that the same patients that had significant reduction in tinnitus after rTMS also showed good response to anodal transcranial direct current stimulation (dTCS), applied over the auditory cortex. Two recent studies combined TMS with different functional imaging techniques: Plewnia et al (2007a) investigated tinnitus patients in whom tinnitus could be suppressed by an intravenous bolus of lidocaine. By using [15O]H2O PET before and after lidocaine injection, they identified changes in neuronal activity in the left middle and inferior temporal (BA 37), in the right temporoparietal cortex (BA 39) and in the posterior cingulum. Then, single sessions of 5, 15 and 30 min low frequency (1 Hz) rTMS were performed in a sham controlled design with the coil localized over the brain areas where lidocaine exerted maximal effects. Tinnitus reduction occurred in 6 out of 8 subjects and lasted up to 30 min. Longer stimulation was more effective and patients with shorter tinnitus duration were more responsive. In another study (Londero et al 2006), fMRI was used for target detection and the effects of high- and low-frequency rTMS were compared in 13 patients. Auditory stimulation with music and language resulted in fMRI activation of the auditory cortex contralateral to the perceived tinnitus. High-frequency rTMS trains over this area resulted in one patient only obtaining reliable tinnitus suppression. On the other hand, after one long single train of low-frequency rTMS, the majority of investigated patients reported a reduction of their tinnitus. The duration of effects was highly variable, lasting up to 10 days. The clinical effect of single-session burst rTMS on tinnitus has been investigated (De Ridder et al 2007b,c). Burst stimulation has been shown to be more powerful in activating the cerebral cortex than tonic stimulation (Huang et al 2005) and bursts may activate neurons that are not activated by tonic stimulations (Moller 2006). The effects of both tonic and burst rTMS were investigated in 46 patients. Narrow band/white noise tinnitus was better suppressed with burst TMS in comparison to tonic TMS, whereas, for pure tone tinnitus, no difference was observed. Based on the hypothesis that white noise/narrow band tinnitus is related to hyperactivity in the non-tonotopic (extralemniscal) system and pure tone the result of increased activity in the tonotopic (lemniscal) system, it was suggested that burst stimulation modulates both the extralemniscal and lemniscal system, whereas tonic
342
stimulation only affects the lemniscal system. This is further evidence that different forms of tinnitus exist, with different pathophysiologies. De Ridder et al (2004, 2005, De Ridder et al 2006, 2007a) administered short trains of high-frequency rTMS as a screening method to select patients for surgical implantation of cortical electrodes. Short-term responders were considered as good candidates for permanent electrical stimulation of the auditory cortex. However, no study has yet determined the value of rTMS as a prognostic test.
Table 2. Effects of repeated sessions of rTMS in tinnitus patients

Authors / Date Stimulation site / coil positioning Frequency Intensity Sessions Pulses 1 Hz 110% MT 5 2000 Number of patients Study design Outcome
Kleinjung et al., 2005
Langguth et al., 2006
Area of maximal PET activation in the temporal cortex, neuronavigation Left auditory cortex, according to 1020 EEG system Area of maximum tinnitus-related PET activation (temporoparietal cortex), neuronavigation Left auditory cortex, neuronavigation al system
14
Sham-controlled, cross-over. Control condition: sham coil open
Significant reduction of tinnitus after active vs sham rTMS; lasting tinnitus reduction (6 months) Significant reduction of tinnitus until end of followup (3 months) Significant reduction of tinnitus after active vs control rTMS; no lasting effects
1 Hz 110% MT 10 2000 1 Hz 120% MT 10 1800
28
Sham-controlled, cross-over. Control condition: occipital cortex
1 Hz 110% MT 10 2000
45
open
Rossi et al., 2007
Left temporoparietal cortex, neuronavigation / 10-20 EEG system Area of maximal PET activation in the temporal cortex, neuronavigation al system
1 Hz 120% MT 5 1200
16
Smith et al., 2007
1 Hz 110% MT 5 1800
Sham-controlled, cross-over. Control condition: coil angulation + electrical stimulation of facial nerve Sham-controlled, cross-over; control condition: coil angulation
Significant tinnitus reduction after rTMS,lasting up furing follow-up period (3 months) responders were characterized by shorter tinnitus duration and less hearing impairment Significant reduction of tinnitus after active vs control rTMS; no lasting effects
Modest response to active treatment in 3 patients (75%)

Khedr et al., 2008 Left temporoparietal cortex, according to 1020 EEG system Left auditory cortex, neuronavigation 1 Hz, 10 Hz, 25 Hz 100% MT 10 1500 1 Hz, 6 Hz + 1 Hz 110% MT (90% MT for 6 Hz rTMS) 10 2000 0,5 Hz 100% MT 5 600 1 Hz, 20 Hz (DLPFC) + 1 Hz 110% MT 10 2000 56 Sham-controlled, parallel group design; control condition: occipital cortex Randomization between two active treatment conditions, parallel group design Open study
343
Significant reduction of tinnitus after all three active rTMS conditions vs control condition; tinnitus reduction during follow-up period (4 months) Significant improvement for both stimulation conditions, no difference between conditions, no lasting effects
Langguth et al., 2008
32
Lee et al., 2008
Left temporoparietal cortex Left auditory cortex, neuronavigatio; left dorsolateral prefrontal cortex
No significant reduction of tinnitus
32
two active treatment conditions, parallel group design
Directly after stimulation significant improvement for both stimulation conditions, at 3 month follow-up significantly better results for the combined frontal and temporal stimulation
MT: Motor threshold, p.: pulses.
B- Studies Using Repeated Sessions of rTMS Based on the success of 1Hz rTMS in treating other conditions which appear to be associated with increased cortical activity, repeated sessions of low frequency rTMS have been proposed as a treatment for disabling tinnitus (Eichhammer et al 2003, Langguth et al 2003). FDG-PET was performed to identify the site of maximum activation in the auditory cortex. Neuronavigated rTMS (110% motor threshold, 2000 pulses/day) was applied to three patients in a sham-controlled cross-over design for five consecutive days. Initial encouraging results were confirmed by a sham controlled TMS study on 14 patients using an identical study design (Kleinjung et al 2005). After active treatment, a significant decrease in the score on the tinnitus questionnaire was reported, whereas sham stimulation had no effect. At 6 months after treatment, 57% of patients reported sustained reduction in tinnitus. However, treatment results were characterized by high inter-individual variability. A further study focusing on possible response predictors (Kleinjung et al 2007) found a significant relationship between tinnitus duration and benefit from treatment. In agreement with other studies (De Ridder et al 2005, 2006, Plewnia et al 2007b), shorter tinnitus duration was associated with better treatment outcome. Normal hearing has been identified as an additional favorable predictor. Interestingly, both predictive factors have been demonstrated to be positive outcome predictors in other treatments for tinnitus as well (Moller et al 1993, Ryu et al 1998). Plewnia et al (2007b) investigated a small sample of six lidocaine responsive patients by using a sham-controlled cross-over design with 2x2 weeks of rTMS applied over the area of
344
maximum lidocaine-related activity change as determined by [15O]H2O PET. This sophisticated procedure resulted in stimulation of the temporoparietal cortex (BA22, BA39). There were significant beneficial effects after active stimulation; however, two weeks after the last session, these effects were no longer detectable. In a case study (Richter et al 2006), 1 Hz rTMS was applied on 5 consecutive days (1800 pulses/d) to a patient with a 30 year history of bilateral tinnitus. The coil was navigated to the area of increased cortical activation as identified by FDG PET- CT (right primary auditory cortex). The patient reported beneficial changes in tinnitus perception persisting up to 4 weeks and there was a statistically significant improvement in objective measures of attention and vigilance. Whereas a PET study performed two days after rTMS treatment did not show any relevant changes as compared to the baseline scan, a follow-up study by the same group which included four patients demonstrated a reduction of metabolic activity in the stimulated area after 5 sessions of low frequency rTMS. Furthermore, in all four patients, a positive effect on tinnitus and an improvement of psychomotor vigilance was observed (Smith et al 2007). Whether neuroimaging guided coil localization is a necessary condition for treatment success has been investigated by employing an easily applicable coil positioning method based on the international 10-20 EEG system (Langguth et al 2006a). This coil positioning method resulted in a significant reduction of tinnitus severity after 10 sessions of 1 Hz rTMS applied to the left auditory cortex, as suggested by imaging studies. Interestingly, a case study which systematically investigated optimal coil position over the temporal region resulted in a very similar position, although on the right side (Fierro et al 2006). A study addressed the question whether the clinical effects of rTMS on tinnitus are mediated by rhythmic stimulation of peripheral nerves (Herraiz et al 2007) or potential antidepressant effects (Rossi et al 2007). In this cross-over study, the control condition involved electrical stimulation of the facial nerve; depression scales were administered in addition to tinnitus assessment with VAS scales. By demonstrating a significant improvement of tinnitus during active stimulation, but not during the control condition, this study strongly indicates that effects on tinnitus are mediated by alterations of cortical function and not via somatosensory modulation. Furthermore, improvement of tinnitus was independent of mood changes. An auditory cortical modulatory effect is confirmed by results of extradural and intradural implantations of electrodes on the secondary and primary auditory cortex respectively, where a tinnitus suppression can be obtained (in TMS responders) as well (De Ridder et al 2007a, and chapter 21). The dura has a sensory innervation, but is coagulated during the procedure (without suppressing tinnitus permanently) to prevent postoperative pain sensation due to dural stimulations and the pia mater has no sensory innervation whatsoever (in intradural electrical stimulations). In a large study, Khedr et al (2008) compared the effects of different frequencies of stimulation. Whereas sham rTMS treatment had no effect, active stimulation over the left temporoparietal cortex resulted in a reduction of tinnitus, regardless of stimulation frequency, low or high (1 Hz, 10 Hz and 25 Hz). Tinnitus relief was still detectable four months after treatment. Further studies are necessary to validate the therapeutic potential of high frequency rTMS.
345
In contrast to these results, no reduction of tinnitus was observed in an open study which investigated the effect of five days of 0.5 Hz to the left temporoparietal cortex (Lee et al 2008). The relatively low number of pulses per session (600) might explain the negative result and support the notion that attenuation of tinnitus is dose dependent (Plewnia et al 2007a). Langguth et al (2008) compared a standard protocol of 1 Hz rTMS with a priming stimulation protocol where 1Hz rTMS was administered after 6Hz rTMS. Both treatment protocols resulted in a reduction in tinnitus severity, with no difference between the two treatments. They (Kleinjung et al 2008) also investigated a protocol where high frequency left prefrontal rTMS preceded low frequency left temporal rTMS and compared results with those from a standard protocol of low frequency temporal rTMS. Immediately after therapy, tinnitus severity was similarly improved, but 3 months later combined prefrontal and temporal rTMS appeared superior. All studies up to now found a similar percentage of responders (about 50%) (Table 2); yet, in some studies, effects outlasted the stimulation period by 3, 4 or 6 months, in others no after-effect was observed. The number of daily sessions may be the key to achieve sustained relief (Langguth et al 2003), as already seen in other TMS applications, like depression and auditory hallucinations. Lasting effects may be due to structural changes in the temporal cortex after one week of low frequency rTMS (May et al 2007).
3. SPECIFIC CONSIDERATIONS
A- The Role of Coil Localization Practically all studies addressed temporal or temporoparietal cortical areas and most used neuronavigation-guided coil localization based on different functional neuroimaging techniques in order to target areas of tinnitus-related brain activity (Eichhammer et al 2003, Kleinjung et al 2005, Londero et al 2006, Plewnia et al 2007a,b). However results from different functional imaging techniques which have been used for target detection (FDGPET, H2O-PET with lidocaine and functional MRI) were not identical: FDG-PET studies demonstrated increased metabolic activity in the left temporal lobe in the majority of the patients, independent from the laterality of the perceived tinnitus (Arnold et al 1996, Kleinjung et al 2005, Langguth et al 2006b); H2O-PET with lidocaine resulted in tinnitus related activity in temporoparietal regions (Reyes et al 2002, Plewnia et al 2007a) and fMRI has shown activation in the auditory cortex contralateral to the perceived tinnitus (Londero et al 2006, Smits et al 2007). These divergent results are probably due to crucial differences between the different imaging approaches: FDG-PET relies on permanent alterations of metabolism, whereas auditory stimulation in fMRI gives information about alterations of neuronal processing in the auditory system. The measurement of changes of cerebral blood flow with H2O-PET during lidocaine-based suppression of tinnitus allows a direct on-off comparison; yet, activation changes reflecting tinnitus perception have to be differentiated from unspecific lidocaine effects on cerebral blood flow (Reyes et al 2002, Plewnia et al 2007a) and only a subgroup of tinnitus patients respond to lidocaine. Therefore, at the
346
moment, no definitive conclusion regarding the optimal imaging method to define the TMS target for tinnitus can be drawn from currently available data. In studies where coil localization was not based on individual functional imaging data, the laterality of stimulation varied between stimulation on the left side in all patients and stimulation contralateral to tinnitus laterality. Some of these studies used a neuronavigation system in combination with structural imaging data, focussing on the primary auditory cortex (Kleinjung et al 2007, Langguth et al 2008). Easier applicable techniques include coil localization according to the 10-20 EEG coordinate system (Plewnia et al 2003, Folmer et al 2006, Fregni et al 2006, Langguth et al 2006a, Khedr et al 2008) and optimization techniques based on clinical effects (De Ridder et al 2005, Fierro et al 2006, Richter et al 2006). None of these coil localization techniques appears superior; on the other hand, no study compared directly different coil localization strategies and comparison across studies is difficult due to further differences in the study design. Although currently available neuronavigated and non-neuronavigated rTMS studies yield similar outcomes, experience with direct electrical stimulation of the auditory cortex by implanted electrodes indicates that the localization of stimulation is critical (De Ridder et al 2007a). This discrepancy might be due to TMS having a modulatory effect on a larger area (cm) than electrical stimulation (mm). Therefore, improving targeting appears necessary.
B- Study Design and Control Condition Evaluation of treatment efficacy in patients with chronic subjective tinnitus requires adequate methodologies to control for unspecific treatment effects. The majority of controlled studies published so far has used cross-over designs, with limited follow-up periods and carry-over effects as potential confounding factors. Therefore, studies using parallel group designs are needed (Landgrebe et al 2008). Different methods have been used as placebo conditions in TMS studies. Sham stimulation has been performed (1) by angulation of the active magnetic coil 45 or 90 away from the skull, (2) by using a so called sham-coil system, which mimics the sound of a real coil without generating any magnetic field, (3) by stimulating non-auditory brain areas and (4) by stimulating head- or neck-muscles without reaching brain areas. Stimulation of nonauditory brain areas as control condition is problematic, since tinnitus related changes in brain activity are not restricted to the auditory cortex (Lockwood et al 1998, Giraud et al 1999, Mirz et al 2000, plewnia et al 2007a, Schlee et al 2008) and therefore this technique cannot be considered as an inactive condition. Coil angulation has the inconvenience that a weak magnetic field effect on the brain cannot be entirely excluded (Lisanby et al 2001). On the other side, this technique has the advantage of producing skull sensations and thus a somatosensory stimulation which is similar to the active rTMS condition. In contrast, the use of a sham coil does not elicit skull sensations. This fact might be important, since, as discussed, stimulation of peripheral nerve structures may contribute to tinnitus improvement (Moller et al 1992, Herraiz et al 2007). Furthermore, in terms of blinding conditions, it is easier for patients to guess whether they receive active or sham TMS application. As a
347
potential solution, Rossi et al (2007) suggested a control condition which involves electrical stimulation of the facial nerve. All methods have in common the limitation that the operator is not blinded. In addition, the blinding of the patient is at best limited, particularly in cross-over designs, in which the patient can directly compare the different stimulation conditions. Another important issue in this context is the fact that TMS itself is a method of multi-modal sensory stimulation. In addition to site-specific effects, TMS can also elicit auditory, tactile and even visual sensations (phosphenes), singly or combined, which might modify the perception of tinnitus. All of these effects depend to a large extent on the site and intensity of the stimulation being delivered. Completely eliminating the multimodal nature of stimulation associated with TMS is not possible. C- Outcome Measurement Validated tinnitus questionnaires and visual analogue scales (VAS) serve as primary outcome measurement in the majority of studies. By quantifying severity, these methods allow to determine whether rTMS induced changes reach statistical significance, but it is not clear what amount of change is of clinical relevance. The additional use of a clinical global impression scale (CGI) may represent a first step in this direction (Plewnia et al 2007b). The development of objective markers for the assessment of treatment outcome is highly desirable. Since tinnitus severity, disability or annoyance do not correlate with its loudness or other psycho-acoustic measurements (Henry and Meikle 2000), these methods are only of limited use. Future progress in neuroimaging techniques might provide objective, comparable and reproducible methods for monitoring therapeutic effects in tinnitus treatment studies (Smith et al 2007).
D- Predictors of Treatment Outcome The high variability of treatment results, which is encountered in all studies, supports the notion of neurobiologically distinct subtypes of tinnitus. This in turn implies that no single approach will be successful in every patient. In this context, the identification of treatment predictors is of utmost importance. Several rTMS studies indicate that treatment response depends on tinnitus duration with better outcome for shorter duration tinnitus (De Ridder et al 2005, Kleinjung et al 2007, Khedr et al 2008, Plewnia et al 2007b). Hearing impairment has been identified as a negative predictor in one study (Kleinjung et al 2007), as auditory deafferentation is assumed to induce disinhibition in the central auditory system which in turn is believed to be critically involved in the pathophysiology of tinnitus (Eggermont 2007): a high degree of hearing impairment might attenuate rTMS effects by perpetually triggering neuroplastic changes in central auditory structures. More research is needed.
348
E- Neurobiological Mechanisms of rTMS Even if there is accumulating evidence that rTMS interferes with neuronal mechanisms involved in the pathophysiology of some forms of tinnitus, the exact mechanisms of action of the different applications are not clear. While it has been assumed that low frequency rTMS exerts its effects by inducing long-term-depression like effects (Eichhammer et al 2007, Langguth et al 2007c), yet, LTD-like effects should be more pronounced when areas of increased excitability are stimulated (Hoffman and Cavus 2002, Siebner et al 2004) and in tinnitus patients enhanced activity of the stimulated area seems not to increase TMS effects (Langguth et al 2006b, Plewnia et al 2007a, Langguth et al 2008). An alternative explanation could be that rTMS disrupts the malfunctioning network involved in tinnitus generation and thereby facilitates the intrinsic ability of the brain to restore normal function. This hypothesis is supported by the recent study of Khedr et al (2008), which indicates that rTMS effects in tinnitus treatment do not critically depend on stimulation frequency.
4. CONCLUSION
These encouraging results must be considered as preliminary due to small sample sizes, methodological heterogeneity and high variability of results among reported studies. Replication in multicenter trials with a large number of patients and long-term follow-up is needed (Landgrebe et al 2008). Knowledge about the duration of treatment effects is still very limited and available data are controversial. Moreover it is far from clear which stimulation parameters are optimal. Delineation of neurobiologically distinct subgroups by imaging methods may help develop more individualized treatment protocols. In sum, compared to other available treatment strategies, which mainly aim at facilitating habituation, TMS represents a promising new and more causally oriented treatment approach for chronic tinnitus. However, further research is needed before rTMS can be considered for routine clinical use.
REFERENCES
Arnold W, Bartenstein P, Oestreicher E, Romer W, Schwaiger M. Focal metabolic activation in the predominant left auditory cortex in patients suffering from tinnitus: a PET study with [18F]deoxyglucose. ORL J Otorhinolaryngol Relat Spec 1996;58:195-199 Cronlein T, Langguth B, Geisler P, Hajak G. Tinnitus and insomnia. Prog Brain Res 2007;166:227-233 De Ridder D, De Mulder G, Walsh V, Muggleton N, Sunaert S, Moller A. Magnetic and electrical stimulation of the auditory cortex for intractable tinnitus. Case report. J Neurosurg 2004;100:560-564 De Ridder D, Verstraeten E, Van der Kelen K, et al. Transcranial magnetic stimulation for tinnitus: influence of tinnitus duration on stimulation parameter choice and maximal tinnitus suppression. Otol Neurotol 2005;26:616-619
349
De Ridder D, Verstraeten E, Van der Kelen K, et al. Transcranial magnetic stimulation for tinnitus: influence of tinnitus duration on stimulation parameter choice and maximal tinnitus suppression. Otol Neurotol 2005;26:616-619. De Ridder D, De MG, Verstraeten E, et al. Auditory cortex stimulation for tinnitus. Acta Neurochir Suppl 2007a; 97:451-462 De Ridder D, De MG, Verstraeten E, et al. Primary and secondary auditory cortex stimulation for intractable tinnitus. ORL J Otorhinolaryngol Relat Spec 2006;68:48-54. De Ridder D, van der Loo E, Van der KK, Menovsky T, Van de Heyning P, Moller A. Do tonic and burst TMS modulate the lemniscal and extralemniscal system differentially? Int J Med Sci 2007b;4:242-246. De Ridder D, van der Loo E, Van der KK, Menovsky T, Van de Heyning P, Moller A. Theta, alpha and beta burst transcranial magnetic stimulation: brain modulation in tinnitus. Int J Med Sci 2007c;4:237-241. Dobie RA. A review of randomized clinical trials in tinnitus. Laryngoscope 1999;109:12021211 Dohrmann K, Weisz N, Schlee W, Hartmann T, Elbert T. Neurofeedback for treating tinnitus. Prog Brain Res 2007;166:473-485 Eggermont JJ. Pathophysiology of tinnitus. Prog Brain Res 2007;166:19-35 Eichhammer P, Kleinjung T, Landgrebe M, Hajak G, Langguth B. TMS for treatment of chronic tinnitus: neurobiological effects. Prog Brain Res 2007;166:369-375 Eichhammer P, Langguth B, Marienhagen J, Kleinjung T, Hajak G. Neuronavigated repetitive transcranial magnetic stimulation in patients with tinnitus: a short case series. Biol Psychiatry 2003;54:862-865 Fierro M, Bobadilla H, Cordoba R, Vanegas C, Olarte A. [Repeated transcranial magnetic stimulation in a patient with chronic bilateral tinnitus]. Rev Neurol 2006;43:758-759 Folmer RL, Carroll MS, Rahim A, Shi Y, Martin WH. Effects of Repetitive Transcranial Magnetic Stimulation (rTMS) on Chronic Tinnitus. Acta Otolarngologica Suppl 2006; 96-101 Fregni F, Marcondes R, Boggio PS, et al. Transient tinnitus suppression induced by repetitive transcranial magnetic stimulation and transcranial direct current stimulation. Eur J Neurol 2006;13:996-1001 Giraud AL, Chery-Croze S, Fischer G, et al. A selective imaging of tinnitus. Neuroreport 1999;10:1-5 Hallberg, L. R. and S. I. Erlandsson. Tinnitus characteristics in tinnitus complainers and noncomplainers. Br J Audiol 1993; 27: 19-27 Heller, A. J.Classification and epidemiology of tinnitus. Otolaryngol Clin North Am 2003; 36: 239-48. Henry JA, Meikle MB. Psychoacoustic measures of tinnitus. J Am Acad Audiol 2000;11:138155 Herraiz C, Toledano A, Diges I. Trans-electrical nerve stimulation (TENS) for somatic tinnitus. Prog Brain Res 2007;166:389-394. Hoffman RE, Cavus I. Slow transcranial magnetic stimulation, long-term depotentiation, and brain hyperexcitability disorders. Am J Psychiatry 2002;159:1093-1102
350
Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta burst stimulation of the human motor cortex. Neuron 2005;45:201-206 Iyer MB, Schleper N, Wassermann EM. Priming stimulation enhances the depressant effect of low-frequency repetitive transcranial magnetic stimulation. J Neurosci 2003;23:10867-10872 Jastreboff PJ. Phantom auditory perception (tinnitus): mechanisms of generation and perception. Neurosci Res 1990;8:221-254 -Khedr EM, Rothwell JC, Ahmed MA, El Atar A. Effect of daily repetitive transcranial magnetic stimulation for treatment of tinnitus: comparison of different stimulus frequencies. J.Neurol.Neurosurg.Psychiatry. 2008; 79: 212-215 Kleinjung T, Eichhammer P, Landgrebe M, et al. Combined temporal and prefrontal transcranial magnetic stimulation for tinnitus treatment: A pilot study. Otolaryngol Head Neck Surg 2008;138:497-501 Kleinjung T, Eichhammer P, Langguth B, et al. Long-term effects of repetitive transcranial magnetic stimulation (rTMS) in patients with chronic tinnitus. Otolaryngol Head Neck Surg 2005;132:566-569 Kleinjung T, Steffens T, Sand P, et al. Which tinnitus patients benefit from transcranial magnetic stimulation? Otolaryngol Head Neck Surg 2007;137:589-595 Landgrebe M, Binder H, Koller M, et al. Design of a placebo-controlled, randomized study of the efficacy of repetitive transcranial magnetic stimulation for the treatment of chronic tinnitus. BMC Psychiatry 2008;8:23 Langguth B, Zowe M, Landgrebe M, et al. Transcranial Magnetic Stimulation for the treatment of tinnitus: A new coil positioning method and first results. Brain Topogr. 2006a; 18: 241-247 Langguth B, Eichhammer P, Kreutzer A, et al. The impact of auditory cortex activity on characterizing and treating patients with chronic tinnitus - first results from a PET study. Acta Otolaryngol Suppl 2006b;84-88. Langguth B, Eichhammer P, Wiegand R, et al. Neuronavigated rTMS in a patient with chronic tinnitus. Effects of 4 weeks treatment. Neuroreport 2003;14:977-980. Langguth B, Kleinjung T, Fischer B, Hajak G, Eichhammer P, Sand PG. Tinnitus severity, depression, and the big five personality traits. Prog Brain Res 2007a;166:221-225 Langguth B, Kleinjung T, Marienhagen J, et al. Transcranial magnetic stimulation for the treatment of tinnitus: effects on cortical excitability. BMC Neurosci 2007c;8:45 Langguth B, Kleinjung T, Frank E, et al. High-frequency priming stimulation does not enhance the effect of low-frequency rTMS in the treatment of tinnitus. Exp Brain Res 2008; 184: 587-591 Lee SL, Abraham M, Cacace AT, Silver SM. Repetitive transcranial magnetic stimulation in veterans with debilitating tinnitus: a pilot study. Otolaryngol Head Neck Surg 2008;138:398-399 Lisanby SH, Gutman D, Luber B, Schroeder C, Sackeim HA. Sham TMS: intracerebral measurement of the induced electrical field and the induction of motor-evoked potentials. Biol Psychiatry 2001;49:460-463.
351
Lockwood AH, Salvi RJ, Coad ML, Towsley ML, Wack DS, Murphy BW. The functional neuroanatomy of tinnitus: evidence for limbic system links and neural plasticity. Neurology 1998;50:114-120 Londero A, Lefaucheur JP, Malinvaud D, et al [Magnetic stimulation of the auditory cortex for disabling tinnitus: preliminary results]. Presse Med 2006;35:200-206 May A, Hajak G, Ganssbauer S, et al. Structural brain alterations following 5 days of intervention: dynamic aspects of neuroplasticity. Cereb Cortex 2007;17:205-210 Mirz F, Gjedde A, Ishizu K, Pedersen CB. Cortical networks subserving the perception of tinnitus--a PET study. Acta Otolaryngol Suppl 2000;543:241-243 Moller AR. Pathophysiology of tinnitus. Otolaryngol Clin North Am 2003;36:249-66. Moller AR. Neuronal plasticity and disorders of the nervous system. Cambridge: Cambridge University Press, 2006. Moller AR. The role of neural plasticity in tinnitus. Prog Brain Res 2007a;166:37-45 Moller AR. Tinnitus and pain. Prog Brain Res 2007b;166:47-53 Moller AR, Moller MB, Yokota M. Some forms of tinnitus may involve the extralemniscal auditory pathway. Laryngoscope 1992;102:1165-1171 Moller MB, Moller AR, Jannetta PJ, Jho HD. Vascular decompression surgery for severe tinnitus: selection criteria and results. Laryngoscope 1993;103:421-427. Muhlnickel W, Elbert T, Taub E, Flor H. Reorganization of auditory cortex in tinnitus. Proc Natl Acad Sci U S A 1998;95:10340-10343. Plewnia C, Bartels M, Gerloff C. Transient suppression of tinnitus by transcranial magnetic stimulation. Ann Neurol 2003;53:263-266. Plewnia C, Reimold M, Najib A, et al. Dose-dependent attenuation of auditory phantom perception (tinnitus) by PET-guided repetitive transcranial magnetic stimulation. Hum Brain Mapp 2007a;28:238-246. Plewnia C, Reimold M, Najib A, Reischl G, Plontke SK, Gerloff C. Moderate therapeutic efficacy of positron emission tomography-navigated repetitive transcranial magnetic stimulation for chronic tinnitus: a randomised, controlled pilot study. J Neurol Neurosurg Psychiatry 2007b;78:152-156 Reyes SA, Salvi RJ, Burkard RF, Coad ML, Wack DS, Galantowicz PJ, Lockwood AH. Brain imaging of the effects of lidocaine on tinnitus. Hear Res 2002;171:43-50 Richter GT, Mennemeier M, Bartel T, et al. Repetitive transcranial magnetic stimulation for tinnitus: a case study. Laryngoscope 2006;116:1867-1872. Rossi S, De Capua A, Ulivelli M, et al. Effects of repetitive transcranial magnetic stimulation on chronic tinnitus. A randomised, cross over, double blind, placebo-controlled study. J Neurol Neurosurg Psychiatry 2007; 78: 857-863 Ryu H, Yamamoto S, Sugiyama K, Uemura K, Nozue M. Neurovascular decompression of the eighth cranial nerve in patients with hemifacial spasm and incidental tinnitus: an alternative way to study tinnitus. J Neurosurg 1998;88:232-236 Schlee W, Dohrmann K, Hartmann T, et al. Assessment and modification of the tinnitusrelated cortical network. Semin Hear 2008; 29: 270-287 Siebner HR, Lang N, Rizzo V, et al. Preconditioning of low-frequency repetitive transcranial magnetic stimulation with transcranial direct current stimulation: evidence for homeostatic plasticity in the human motor cortex. J Neurosci 2004;24:3379-3385
352
Sindhusake, D., P. Mitchell, et al. Prevalence and characteristics of tinnitus in older adults: the Blue Mountains Hearing Study. Int J Audiol 2003; 42: 289-94 Smith JA, Mennemeier M, Bartel T, et al. Repetitive transcranial magnetic stimulation for tinnitus: a pilot study. Laryngoscope 2007;117:529-534 Smits M, Kovacs S, De Ridder D, Peeters RR, Van HP, Sunaert S. Lateralization of functional magnetic resonance imaging (fMRI) activation in the auditory pathway of patients with lateralized tinnitus. Neuroradiology 2007;49:669-679. Wassermann EM. Risk and safety of repetitive transcranial magnetic stimulation: report and suggested guidelines from the International Workshop on the Safety of Repetitive Transcranial Magnetic Stimulation, June 5-7, 1996. Electroencephalogr Clin Neurophysiol 1998;108:1-16 Weisz N, Dohrmann K, Elbert T. The relevance of spontaneous activity for the coding of the tinnitus sensation. Prog Brain Res 2007a;166:61-70. Weisz N, Muller S, Schlee W, Dohrmann K, Hartmann T, Elbert T. The neural code of auditory phantom perception. J Neurosci 2007b;27:1479-1484
Chapter 21
AUDITORY CORTEX STIMULATION FOR INTRACTABLE TINNITUS

Dirk De Ridder, Berthold Langguth, Mark Plazier, Elsa van der Loo, Tim Vancamp, Jan Ost, Olivier Meeus2, Stefan Sunaert3, Silvia Kovacs3, Tomas Menovsky2, Paul Van de Heyning
BRAIN, TRI Clinic and Dept of Neurosurgery, 2TRI Clinic and Dept of ENT, University Hospital Antwerp, Belgium; 3 Dept of Radiology, University Hospital Louvain, Belgium; 4 TRI Clinic & Dept of Psychiatry, University of Regensburg, Germany; St. Jude Medical, Zaventem, Belgium.
INTRODUCTION
This chapter describes the technique of auditory cortex stimulation (A-ECS) for refractory chronic tinnitus. A-ECS has been attempted on the basis of three hypotheses: 1. Tinnitus is the result of synchronized gamma band (40 Hz) thalamocortical hyperactivity in the auditory cortex, whether as a carrier wave or neural code (Weisz et al 2007). White noise tinnitus may be caused by synchronous hyperactivity of burst firing in the non-tonotopic extralemniscal system, whereas pure tone tinnitus may be the result of increased synchronous tonic firing in the tonotopic (lemniscal) system (lemniscal dorsal cochlear nucleus, inferior colliculus and primary auditory cortex) (De Ridder et al 2007). Narrow band tinnitus could be the result of a co-activation of the lemniscal and extralemniscal pathways. This synchronized gamma band activity can be visualized using fMRI, magnetic source imaging (MSI) or low resolution brain electromagnetic tomography
2.
Correspondence concerning this article should be addressed to: Dr. Dirk De Ridder, MD, e-mail: dirk.de.ridder@neurosurgery.be/uza.be
354
Dirk De Ridder, Berthold Langguth, Mark Plazier et al.

(LORETA) (Pascual-Marqui et al 1994) (figure 1); the fMRI BOLD effect is highly coupled to gamma local field potentials (EEG) in the auditory cortex (Mukamel et al 2005, Nir et al 2007). This strongly suggests that fMRI can visualize the gamma band synchronized activity associated with tinnitus (figure 2; BOX 1).
BOX 1. fMRI of tinnitus

A scanning paradigm, using music as a stimulus adequately visualizes the auditory pathways in tinnitus patients (Smits et al 2004). fMRI activation is symmetrical in patients with bilateral tinnitus at all investigated areas of the auditory pathways (auditory cortex, thalamus and inferior colliculus). In patients with unilateral tinnitus, fMRI activation is significantly decreased all along the contralateral auditory pathway. fMRI activation always represents a difference in neural activity, instead of absolute neural activity. An increase of spontaneous neural activity, such as postulated in tinnitus patients, would mean that the affected brain area during the rest condition is more active than the unaffected side, and that the active condition (sound presentation) will only give rise to a limited increase in activity due to a ceiling effect (known as the saturation model) in comparison to the nonaffected side. This can explain the fact that constant pathological neuronal hyperactivity can be correlated to hypoactivation in fMRI (Smits et al 2004). A similar study for tinnitus using tinnitus pitch and character specific stimuli is currently being conducted. In this study we compare BOLD activation for tinnitus specific sound presentation to non-tinnitus sounds presented in the scanner. Only tinnitus specific sounds induce a significant BOLD change, as demonstrated by a lateralization effect, in contrast to non-tinnitus sounds, which generate a bilateral symmetrical BOLD activation. Even for tinnitus-specific frequencies the exact representation might be important. For a pure tone tinnitus auditory presentation of a pure tone generates a marked asymmetrical BOLD activation, a narrow band noise less so and a white noise almost no asymmetry (unpublished data).
3.
TMS can be applied to test whether this synchronized gamma band activity can be modulated non-invasively. In particular, if it is possible to modify a patients tinnitus by rTMS, an intracranial electrode can be implanted on or in the primary and/or secondary auditory cortex. High frequency (10 Hz) TMS might induce a transient disruption of neuronal activity (virtual lesion) of the underlying auditory cortex, thus temporarily desynchronizing the 40 Hz thalamocortical activity. PET/CT scans performed before and after TMS for tinnitus suppression demonstrate that symptom improvement may be related to a reduction of the cortical asymmetry associated with tinnitus (Smith et al 2007). Burst stimulation may modulate both the extralemniscal and lemniscal systems, whereas tonic stimulation only modulates the lemniscal system (De Ridder et al 2007). The frequencies at which burst and tonic TMS maximally suppress pure tone tinnitus are the same, suggesting that the extralemniscal system drives the lemniscal system. In white noise, supposedly generated in the extralemniscal system, this is not seen, as expected. Lower frequencies of narrow band tinnitus respond better to burst stimulation than higher frequencies (De Ridder et al 2007), as lower pitch sounds have a wider tuning curve and thus respond more like a non-tonotopic system in general.
Extradural Auditory Cortex Stimulation for Intractable Tinnitus

Electrical stimulation of the auditory cerebral cortex can be accomplished in three different ways. Electrodes can be implanted:
355
1- extradurally, covering a part of the secondary auditory cortex (plate electrodes). 2- intradurally, covering the intrasulcal grey matter of the primary auditory cortex (plate electrodes). 3- intraparenchymatously, in the white matter tracts leaving the primary auditory cortex (wire electrodes).
Focal electrical stimulation of the auditory cortex may exert its effects via the descending corticofugal projections from the auditory cortex to the medial geniculate body (thalamus), the inferior colliculus , superior olivary complex and cochlear nuclei (Moore and Linthicum 2004, Nieuwenhuys et al 2008). Electric stimulation creates tonotopic changes by decreasing best frequencies slightly higher and increasing best frequencies slightly lower than those electrically stimulated (Suga et al 2000), resulting in sharpened tuning of auditory neurons in the thalamus and inferior colliculus via a positive feedback, in combination with lateral inhibition. Thus, focal auditory cortex stimulation could result in an egocentric selection, in a similar way as the auditory corticofugal system controls reorganization of the thalamus and inferior colliculus by adjusting the frequency maps to auditory experience (Suga et al 2000).
Figure 1. Gamma (40 Hz) band activity in left auditory cortex associated with unilateral right-sided tinnitus in 12 patients, as assessed by LORETA EEG, which performs source analysis of the gamma band activity in tinnitus patients (Pasqual-Marqui et al. 1994). Our LORETA results demonstrate that it is possible to visualize cortical gamma band activity correlated to tinnitus loudness. Gamma band activity is localized in the auditory cortex: an ICA of the raw EEG can be performed and the independent component that co-localizes with the gamma band activity can be considered to contain tinnitus-related information. Using data from a patient implanted with an exradural electrode overlying the secondary auditory cortex , validation of the independent components measured at scalp level has
356
been obtained at the site of the intracranial electrode. 40 Hz is a carrier wave, carrying the tinnitus related information, which can potentially be represented by a co-localized independent component, thus the ICA of scalp EEG can serve as a tool to detect the neural correlate of tinnitus.
Figure 2. Tinnitus-frequency specific BOLD changes on fMRI in a patient with left sided pure tone tinnitus. Auditory presentation of non-tinnitus sound (OF: other frequency) generates a bilateral BOLD activity both on white noise (BPW: band pass wide), narrow band noise (BPS: band pass small) and pure tones. In contrast, for the tinnitus frequency (TF), a pure tone generates a marked asymmetrical BOLD activation, a narrow band noise less so and a white noise almost no asymmetry.
TECHNIQUE OF EXTRADURAL SECONDARY AUDITORY CORTEX STIMULATION

Reciprocal functional pathways have been demonstrated in man between Heschls gyrus (primary auditory cortex, A1) and the acoustically responsive posterior lateral superior temporal gyrus (PLST) (Brugge et al 2003). Thus, electrical stimulation of the PLST area may affect A1 and suppress tinnitus-associated hyperactivity. For such implantation, the patient first undergoes audiological testing determining the pitch of the perceived tinnitus (tinnitus matching). If subsequent TMS, neuronavigated or not, is capable of suppressing the tinnitus transiently in a placebo controlled way on two separate occasions, the patient becomes a candidate for implantation of an extradural electrode, using
357
fMRI guided neuronavigation (De Ridder et al 2004, 2005). With this technique, a curvilinear 6 cm long incision is made overlying the secondary auditory cortex, as determined by fMRIguided neuronavigation. The 6 cm x 2 cm craniotomy and electrode placement are tailored in the same navigated fashion. After bipolar coagulation of the dura, lesioning the sensory receptors in the dura, the lead is positioned extradurally using fMRI neuronavigation and sutured the dura. It is tunneled subcutaneously to the abdomen, connected to an extension lead and externalized. Three days after the implantation, programming of the electrode is begun, using either a trial-and-error technique or based on iEEG recordings from the implanted electrode. The normal power versus frequency plots demonstrate the typical individual alpha peak of the sensory cortices. A theta peak signals tinnitus. The electrodes where abnormal EEG data are recorded are elected as preferential targets for neuromodulation (figure 3).
Figure 3. Postoperative CT fused with preoperative fMRI. The cross-marks identical stereotactic localization. Bipolar recordings from implanted electrodes demonstrating power versus frequency. A clear theta power peak instead of the normal alpha peak is present when the patient perceives tinnitus, and is absent when the patient does not perceive tinnitus (during the period of residual inhibition resulting from electrical stimulation).
Once a stable tinnitus suppression is obtained, the extension lead is removed and a new extension lead is connected to the internal pulse generator (IPG) (EON, St. Jude Medical ANS Division, Tx,USA) that is implanted in a subcutaneous pocket in the abdomen. Extradural octopolar (Lamitrode 44, St. Jude Medical ANS Division, Tx, USA) or 16-polar electrodes (Lamitrode 88) are preferred. The Lamitrode leads (60cm in length) are made of 8 or 16 electrodes with 28 mm and 56 mm electrode spans respectively, configured with 2 offset rows of 4 or 8 electrodes, 4 mm x 2.5 mm each, and 1 mm lateral spacing and 3mm longitudinal spacing between the electrodes.
358
Results: To date, 32 patients have been implanted (15 females, 17 males; 26 unilateral, 6 bilateral tinnitus). All but two were suffering from intractable disabling tinnitus (Grade 4 on the Goebler-Hiller Tinnitus Questionnaire). Results are as follows: 13 complete failures, 9 complete/near complete tinnitus suppressions, 10 partial suppressions. The average tinnitus suppression for pure tone tinnitus has been 97%, but only 24% for unilateral white or narrow band noise. In patients with unilateral tinnitus characterized by both white noise and pure tone tinnitus, only the latter component disappeared with tonic stimulation. Recently, using a custom made program, burst electrical stimulation of the auditory cortex has been developed. Five patients with intractable tinnitus to implanted electrodes were investigated in this study. All patients suffered from a combination of both pure tone and narrow band/white noise tinnitus. Results: a significant better suppression occurs for narrow band noise tinnitus with burst stimulation in comparison to tonic stimulation, whereas for pure tone tinnitus no difference is found (similarly to what occurs in burst and tonic TMS protocols). A recent extradural electrical stimulation study of 8 patients implanted on the posterior superior temporal gyrus (Friedland et al 2007) yielded results that seem to correlate better with 1-Hz TMS data than with our cortical stimulation data using implanted electrodes. In contrast to our results, the authors found more improvement on tinnitus distress than on tinnitus intensity, even though 6 of 8 patients demonstrated a temporary complete suppression of their tinnitus at some point in time during stimulation (2 out of 8 had persistent reduction of pure tone tinnitus). This might be due to the use of a different stimulation device and different electrodes. Using an electrode with only 2 contacts limits the way the electrodes can be fully programmed. Keeping the tinnitus suppressed might require an electrode with more contacts, so more complex programming is possible (De Ridder et al 2004). A second potentially important difference between the two studies might be the use of preoperative TMS as an inclusion criterion for surgical implantation. A study looking at the predictive value of TMS for the success rate of electrical stimulation via implantated electrodes would be very helpful in that respect.
TECHNIQUE OF INTRADURAL STIMULATION

1- Intradural Intrasulcal Grey Matter Stimulation of the Primary Auditory Cortex Due to its simplicity and safety, extradural stimulation is the preferred method to start with. In some patients suppression can be obtained for only a short time (1 to 3 days) after which the effect wears off, even after reprogramming the stimulation several times. This might be due to plasticity of the secondary auditory cortex. Secondary sensory cortices and association cortices in general may be more plastic than primary sensory cortices (Moller 2003). In four patients, an intradural electrode was inserted in the Sylvian fissure on the primary auditory cortex. In two, the indication was to obtain stabilization of tinnitus suppression, because the stimulus parameters had to be reprogrammed every 2 to 3 days during 20
359
separate sessions. In both, intradural positioning resulted in stable levels of tinnitus suppression. The two other patients were implanted in the Sylvian fissure because they were refractory to secondary auditory cortex stimulation. Both presented with narrow band tinnitus. Neither of them obtained relief.
2- Intradural Intraparenchymatous Deep White Matter Stimulation of the Primary Auditory Cortex In collaboration with Drs Michael Seidman and Kost Elisevich a third approach was developed, using MEG based neuronavigation guided intraparenchymatous deep white matter stimulation below layer 6 of the primary auditory cortex. Only 2 patients have been implanted using this technique, with variable results (Seidman et al 2008).
SIDE EFFECTS
The stimulation parameters for tinnitus are selected in such a way that suppression of tinnitus can be achieved without noticeable side effects from the stimulation. As the patient has no sensory perception of the stimulation, he or she cannot feel whether the stimulation is on or off (which allows placebo studies). Side effects can be induced by high frequency, high intensity stimulation: feelings of intoxication, altered spatial localization of external sounds, word finding problems, dizziness, vertigo and hearing perception changes (sounds, including one owns voice, perceived as being clearer) (De Ridder et al 2007, 2008). In one patient in whom word finding disturbances were observed during high intensity stimulation, the location of stimulation corresponded to selective disturbances for persons, animals and tools in an anteroposterior direction, compatible with what has been reported in the topography of lexical retrieval. In patients with successful tinnitus suppression, feeling of aural pressure also decreased whenever present, but not necessarily with identical parameters (De Ridder et al 2007, 2008). In one patient, a posteriorly located electrode overlying the right-sided supramarginal-angular gyrus junction induced an out of body experience (De Ridder et al 2007) only at very specific burst stimulation parameters.
COMPLICATIONS
Epileptic seizures occurred in two of 32 implanted patients, most likely because of prolonged stimulation without stimulation free intervals. These seizures occurred when the patients were still having an external stimulator, which was under patients control. In one of these patients, a second epileptic seizure occurred when the stimulation sequence was switched to burst stimulation. He was considered unsuitable for further stimulation. One patient did not disclose previous treatment for epilepsy and it was only discovered
360
incidentally after one year of seizure-free stimulation. With the IPG implanted, no epileptic seizures were seen with the following parameters: >10 Hz, cyclical (5 sec ON/ 5 sec OFF) or <10 Hz, cyclical (as per >10 Hz or 30 sec ON up to 5 min OFF); the OFF period is tailored to the patients after-effect.
CONCLUSION
Auditory cortex stimulation for tinnitus is based on the hypothesis that pure tone tinnitus perception correlates with focal synchronized gamma band activity in the primary and/or secondary auditory cortices. Tinnitus related distress might be the result of co-activation of a right-sided distress network. Tinnitus-related activity can be visualized using LORETA EEG, magnetic source imaging or functional MRI. The high resolution images thus obtained can be used as a target for non-invasive or invasive neuromodulation. Results suggest that patients presenting unilateral tinnitus which can be suppressed in a placebo controlled way by rTMS are candidates for surgical electrode implantation. Whereas patients with tinnitus of the pure tone type most probably benefit from tonic stimulation, patients with narrow band tinnitus benefit from burst stimulation.
REFERENCES
Brugge JF, Volkov IO, Garell PC, Reale RA, Howard MA 3rd. Functional connections between auditory cortex on Heschl's gyrus and on the lateral superior temporal gyrus in humans. J Neurophysiol 2003;90: 3750-63 De Ridder D, De Mulder G, Walsh V, Muggleton N, Sunaert S, Mller A. Magnetic and electrical stimulation of the auditory cortex for intractable tinnitus. Case report. J Neurosurg 2004; 100: 560-4 De Ridder, D., E. Verstraeten, et al. Transcranial magnetic stimulation for tinnitus: influence of tinnitus duration on stimulation parameter choice and maximal tinnitus suppression. Otol Neurotol 2005; 26: 616-619. De Ridder D, Van de Heyning P. The Darwinian plasticity hypothesis for tinnitus and pain. Prog Brain Res 2007; 166: 55-60. De Ridder D, Van der Loo E, Van der Kelen K, et al. Do tonic and burst TMS modulate the lemniscal and extralemniscal system differentially? Int J Med Sci 2007; 4: 242-6. De Ridder D, Van der Loo E, Van der Kelen K, et al. Theta, alpha and beta burst transcranial magnetic stimulation: brain modulation in tinnitus. Int J Med Sci 2007; 4: 237-41 De Ridder D,Van Laere K, DuPont P, Menovsky T, Van de Heyning P. Visualizing out-ofbody experience in the brain. N Engl J Med 2007; 357: 1829-33 De Ridder D, Menovsky T, Van de Heyning P. Auditory cortex stimulation for tinnitus suppression. Otol Neurotol 2008; 29: 574-575 Friedland, D. R., W. Gaggl, Runge-Samuelson C, Ulmer JL, Kopell BH. Feasibility of Auditory Cortical Stimulation for the Treatment of Tinnitus. Otol Neurotol 2007; 28: 1005-1012.
361
Moller, A. Sensory Systems: Anatomy, Physiology, And Pathophysiology Amsterdam, Academic Press, 2003 Moore JK, Linthicum FH. Auditory system. In: Paxinos G, Mai J. The human nervous system. Amsterdam: Elsevier Academic Press, 2004, 1242-1279 Mukamel R, Gelbard H, Arieli A, Hasson U, Fried I, Malach R. Coupling between neuronal firing, field potentials, and FMRI in human auditory cortex. Science 2005; 309: 951-4. Nieuwenhuys R, Voogd J, Van Huijzen C. The human central nervous system. Berlin: Springer Verlag, 2008, 733-750 Nir Y, Fisch L, Mukamel R, et al. Coupling between neuronal firing rate, gamma LFP, and BOLD fMRI is related to interneuronal correlations. Curr Biol 2007; 17: 1275-85 Pascual-Marqui RD, Michel CM, Lehmann D. Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain. Int J Psychophysiol 1994; 18: 49-65. Seidman MD, Ridder DD, Elisevich K, et al. Direct electrical stimulation of Heschl's gyrus for tinnitus treatment. Laryngoscope 2008; 118: 491-500. Smith JA, Mennemeier M, Bartel T, et al.Repetitive transcranial magnetic stimulation for tinnitus: a pilot study. Laryngoscope 2007; 117: 529-34 Smits M, Kovacs S, De Ridder D, et al. Lateralization of signal change in the auditory pathway in patients with lateralized tinnitus studied with functional Magnetic Resonance Imaging (fMRI). Radiology 2004; 233, suppl.: A12-06 Suga N, Gao E, Zhang Y, Ma X, Olsen JF. The corticofugal system for hearing: recent progress. Proc Natl Acad Sci U S A 2000; 97: 11807-14. Weisz N, Mller S, Schlee W, Dohrmann K, Hartmann T, Elbert T. The neural code of auditory phantom perception. J Neurosci 2007; 27: 1479-84.
PROSTHETICS
Chapter 22
CORTICAL VISUAL NEUROPROSTHESIS

E. Fernndez1,, F. Pelayo2, S. Romero2, J.M. Ferrandez3, C. Botella4, J. Albisua5 and R. A. Normann6
Instituto de Bioingeniera, Universidad Miguel Hernndez, Elche, Spain; Dept. Tecnologa y Arquitectura de Ordenadores, Univ. Granada, Granada, Spain; 3 Dept. Tecnologa Electrnica, Universidad Politecnica Cartagena, Cartagena, Spain; 4 Dept. Neurosurgery, Hospital General de Alicante, Alicante, Spain; 5 Dept. Neurosurgery, Fundacin Gimenez Daz, Madrid, Spain; 6 Dept. of Bioengineering, University of Utah, Salt Lake City, USA.
2 1
INTRODUCTION
According to the World Health Organization (http://www.who.int/en), an estimated 180 million people worldwide are visually disabled and this number is expected to double by the year 2020. No treatment exist for many of these and total eye transplantation still remains at project level (Canavero 1992). A visual prosthesis could help millions of such patients, but the scientific and engineering problems are much more complex than originally believed and there are still many unresolved issues causing a delay in its development.
THE CASE FOR A CORTICAL VISUAL NEUROPROSTHESIS

The concept of artificially producing a visual sense in blind individuals is founded on our present understanding of the structure of the mammalian visual system, its processing elements, and the relationship between electrical stimulation of any part of the visual pathways and the resulting visual perceptions (Brindley and Lewin, 1968; Dobelle and
Correspondence concerning this article should be addressed to: Dr. Eduardo Fernndez, MD, e-mail: e.fernandez@umh.es.
366
E. Fernndez, F. Pelayo, S. Romero et al.
Mladejovsky, 1974; Schmidt et al., 1996; Veraart et al., 1998; Dobelle, 2000; Delbeke et al., 2001; Fernandez et al., 2002; Delbeke et al., 2003; Humayun et al., 2003). As blindness can result from an interruption of the normal flow of signals along the visual pathway, a visual prosthesis has to excite the neurons of the pathway at some point after the damaged site. The only requirement is that the device be in contact with still functioning neural elements. Since retinal diseases frequently reduce visual acuity and result in non-curable blindness, several groups worldwide are working on the development of different prostheses designed to interact with the remaining healthy retina (Eckmiller, 1997; Walter et al., 1999; Zrenner et al., 1999; Humayun, 2001; Rizzo et al., 2001; Zrenner, 2002; Humayun et al., 2003; Rizzo et al., 2003), optic nerve (Veraart et al., 1998; Delbeke et al., 2001; Veraart et al., 2003) and even the dorsal lateral geniculate nucleus (Pezaris and Reid, 2007). In this context, five groups are already pursuing human clinical trials of diverse retinal prostheses (Dagnelie, 2006; Wickelgren 2006). However, the output neurons of the eye, the ganglion cells, often degenerate in many forms of retinal blindness (Jones et al., 2003; Marc and Jones, 2003) and therefore a retinal or optic nerve prosthesis would not be always helpful. This extensive degeneration usually spares the neurons within the higher visual regions of the brain, indicating the enormous potential of a cortical prosthesis approach (Brindley and Lewin, 1968; Dobelle et al., 1974; Dobelle et al., 1976; Bak et al., 1990a; Normann et al., 1996; Schmidt et al., 1996; Normann et al., 1999; Fernandez et al., 2002; Fernandez et al., 2005a). Thus, if the higher visual centers can be stimulated with visual information in a format somewhat similar to the way they were stimulated before the onset of blindness, a blind individual could be able to use this stimulation to extract information about the physical world around him/her (Normann et al., 1996; Fernandez et al., 2005a ). This concept is supported by several studies, which demonstrate that localized electrical stimulation of the human visual cortex can excite topographically mapped visual percepts. The first report of the appearance of phosphenes after electrical stimulation of the visual cortex was published by Lwenstein and Borchart in 1918. Subsequently Penfield (Penfield and Rasmussen, 1950; Penfield and Jaspers, 1974) observed that electrical stimulation of the surface of the visual cortex evoked the perception of points of light (phosphenes). These observations led a number of investigators to propose that electrical stimulation of the visual cortex via arrays of electrodes might provide the profoundly blind with a limited form of functional vision. Experiments by the group of Giles Brindley at Cambridge University in England (Brindley and Lewin, 1968; Brindley et al., 1972; Brindley, 1982) and William Dobelle at the University of Utah (Dobelle and Mladejovsky, 1974; Dobelle et al.,1974, 1976) showed that stimulation with multiple electrodes simultaneously allowed blind volunteers to recognize simple patterns, including letters and Braille characters. The results from these studies, however, also indicate that a cortical prosthesis based on relatively large electrodes implanted subdurally have limited usefulness because of factors such as the high levels of current required to produce phosphenes (currents in the range of 1-3 mA), interactions between phosphenes, occasional elicitation of pain due to meningeal or scalp stimulation, and risk of inducing epileptic seizures. A promising approach, which can activate populations of neurons with greater spatial specificity and lower levels of stimulation than is possible with larger electrodes on the surface of the brain, is the use of intracortical microelectrodes. In this context, Schmidt et al
Cortical Visual Neuroprosthesis
367
(1996) described the implantation of 38 floating microelectrodes within the right visual cortex of a 42-year-old woman who had been blind over 22 years. 34 of the microelectrodes were able to elicit phosphenes for a period of 4 months, and most of the microelectrodes had stimulation thresholds below 25 A. Unfortunately these microelectrodes were not well suited for long term application. Thus, due to the breakage of lead wires early in the experiment, only limited tests could be done to evaluate pattern recognition. Nevertheless, taken as a whole, data suggest that passing electrical currents through an array of electrodes inserted into an appropriate location in the visual pathway can produce the perception of phosphenes and that these phosphenes could be appropriate for restoring some limited, but useful, sense of vision to the profoundly blind.
PHYSIOLOGICAL FOUNDATIONS
Before looking at the specifics of a visual neuroprosthesis, it will be helpful to review some of the physiological foundations for this neuroprosthetic approach.
1. There is Abundant and Positive Clinical Experience with many Neural Prosthetic Interfaces Deep brain stimulators have been implanted successfully in patients for pain management and for control of motor disorders such as Parkinsons disease, cochlear implants are being used for restoring auditory function and a wide variety of devices have been developed to control respiration, activate paralysed muscles or stimulate bladder evacuation (e.g. Normann 2007). Thus, it is possible to interface the nervous system with safe and effective devices.
2. Most Forms of Blindness are of Retinal Origin and Leave the higher Visual Centers Unaffected This observation is often unstated, but is key to a cortical approach to visual neuroprosthetics. The output neurons of the eye, the ganglion cells, often degenerate in many retinal blindnesses (Jones et al., 2003; Marc and Jones, 2003) and therefore a retinal or optic nerve prosthesis would not be always helpful. However, this extensive degeneration spares the neurons in the higher visual regions of the brain.
3. The Visual Pathways and Primary Visual Cortex are Organized in a Relatively Rational Scheme Receptive field centers of primary visual cortex neurons correspond in a moderately systematic fashion to locations from the fovea to the periphery (Tusa et al., 1978; Sereno et al., 1994; Warren et al., 2001; Warren et al., 2004; Kim et al., 2006a). Points located in the
368
right visual field are imaged on the temporal side of the left eye and the nasal side of the right eye. Axons from ganglion cells in these retinal cell regions make connections with separate layers in the LGN and these neurons send their outputs to cortical layers 4C and 4C respectively. Thus, the overall spatial position of the retinal ganglion cells within the retina is preserved by the spatial organization of neurons within the LGN and visual cortex. This rational mapping of visual space onto the neurons of the visual cortex is one of the fundamental cornerstones upon which a cortical visual neuroprosthesis is based.
4. Electrical Stimulation of Neurons in the Visual Pathway Evokes the Perception of Points of Light As Johannes Mller (1837) stated with his law of specific nerve energies, perceptions are determined by which nerve fibers are activated, not by how the nerve fibers are activated. For example, mechanical pressure to the eye produces a sensation of light and electrical activation of axons in the auditory nerve give rise to a sensation of sound. In this sense, studies during neurosurgical procedures have revealed that localized direct electrical stimulation of the exposed human visual cortex can evoke topographically mapped visual percepts (Penfield and Rasmussen, 1950; Brindley and Lewin, 1968; Brindley et al., 1972; Dobelle and Mladejovsky, 1974; Dobelle et al., 1974; Penfield and Jaspers, 1974; Dobelle et al., 1976; Dobelle 2000). These percepts are generally called phosphenes and are usually described as stars in the sky, clouds, pinwheels, and occasionally as complex chromatic or kinetic sensations. The induction of phosphenes by cortical stimulation establishes the visual nature of the stimulated cortex and provides the basis for the development of a cortical visual prosthesis (Normann et al., 1996; Normann et al., 1999; Warren et al., 2001; Fernandez et al., 2002; Troyk et al., 2003).
5. The Plasticity of the Brain will Foster Significant Functional Reorganization The mature visual system of primates and other mammals is capable of extensive reorganization as the roles of inputs and pathways are altered by visual experience, sensory loss, or cortical lesions. Although this plasticity declines with age (Berardi et al 2003), the adult visual cortex still responds to experience with plastic changes, as shown by the effects of perceptual learning (Schoups et al 2001) and retinal lesions (Dreher et al 2001). The understanding of these neuroplastic processes will provide the neuroscientific foundation for improved rehabilitation and teaching strategies for the blind. In addition, the modulation of such plasticity will be crucial in developing and projecting the success of novel, visual neuroprosthetic strategies. Thus, it is hoped that this neural plasticity will allow a quick reassociation of the phosphenes world with the physical world. For example, immediately after the electrode array is implanted, the evoked phosphenes are likely to engender a poor perception of an object (as it happens, for example, with cochlear implants). However, after
369
some time and with the help of adequate training the reorganization of the neural network will improve perception.
ENGINEERING PRINCIPLES
The most fundamental requirements of any neurological prosthesis are well understood. In order for a device to effectively emulate a neurological system, it has to do three things: 1. First, it must collect the same kind of information that the neuronal system normally does. Consequently, sensory prostheses always include a processing module that mimics the transfer function of biological sensory receptors so that they can be emulated by their artificial counterparts. Thus, a visual prosthesis has to capture the attributes of the visual scene. 2. Next, it has to process that information. 3. Third, it must communicate the processed information, in an appropriate way, with other parts of the nervous system. Therefore, all neural prostheses need a stable electronic/neural interface that enables selective activation of specific groups of neurons and allow for chronic injection of electrical charge, without deterioration of the electrodes or surrounding neural tissue. Furthermore, as the neural activity evoked by the processing modules is not exactly the same as expected, sensory prostheses must rely on the assumption that the nervous system is able, to a certain degree, to adapt to stimuli that differ from what was expected (Sanguineti et al 2001). A cortically based system will use a bioinspired retina able to perform some of the image pre-processing functions of the retina. This bio-inspired device will transform the visual world in front of a blind individual into electrical signals that can be used to excite neurons at the occipital cortex. These signals will be fed to intracortical microelectrodes that will excite visual cortex neurons in an appropriate way. Since signals reaching the cortex from the retina and LGN reach layer 4 at a depth of 1-2 mm (layer 4), intracortical penetrating electrodes with exposed tips located in layer 4 and with sizes of the same order of magnitude as the neurons that are intended to be stimulated are needed. An integrated telemetry system will transmit power and data (electric impulses) to the electrode array inserted into the visual cortex. The whole visual neuroprosthetic device is expected to recreate a limited, but functionally useful visual sense in blind individuals allowing them to navigate in familiar environments and to read enlarged text. The output of the bioinspired retina has to be transmitted to a remote stimulating device that electrically connects with the implanted microelectrode array. Ideally, this should be done telemetrically (i.e. without the need of attached wires) to reduce the risk of infections and an opto-coupling stage should be incorporated to protect the patient against electrical risks. Furthermore, for a durable system, material must be biocompatible and electricalcharge displacements must be at the margins of capacitive work, never in an irreversible-
370
faradaic working zone (Hu et al 2006, Srivastava and Troyk, 2006; Cogan et al., 2007; Musallam et al., 2007, Troyk et al., 2007b; Troyk et al., 2007a) (Figure 1). Several factors should be taken into account, especially safety, size and power consumptions. Thus, this stimulating device should be able to operate in real-time for a high number of electrodes, have low power requirements and be flexible enough to generate different waveforms and adapt to different stimuli and image processing conditions. Basically, it should be able to solve the problem of transmitting data and power to multiple microelectrodes, having simultaneously the possibility of remote control and sensing of microelectrode status. It implies a compromise between power level requirements of the stimulation circuits and signal bandwidth requirements (Pelayo et al., 2003; Pelayo et al., 2004; Morillas et al., 2005; Morillas et al., 2007; Romero et al 2008).
Figure 1. Basic components of a cortical visual prosthesis. (A) The system uses a bioinspired encoder able to perform some of the image pre-processing functions of the retina. This bioinspired device transforms the visual world in front of a blind individual into electrical signals that can be used to excite, in real time, the neurons at his/her visual cortex. The system includes a way to send power and control signals to implanted electronics. (B) Since signals reaching the cortex from the retina and LGN arrive not at the surface of the cortex (layer 1), but at a depth of 1-2 mm (layer 4), it is necessary to use intracortical penetrating electrodes with exposed tips located in layer 4c.
DEVELOPMENT OF AN ARTIFICAL RETINA

One of the major challenges of a cortical visual approach is the design and development of a bioinspired platform able to transform the visual world in front of a blind individual into a set of electrical signals that can be used to stimulate, in real time, the neurons in his/her visual cortex (Pelayo et al 2004). These signals should be as similar as possible to the output signals of the real retina. The question of how the information about the external world is compressed in the retina, and how this compressed representation is encoded in spike trains is therefore of seminal importance. Sampling across the retina is not uniform (Ahnelt and Kolb
371
2000) and therefore retinotopic gradients and magnification factors (Adams and Horton 2002) have to be introduced to match image representation with cortical topography. Moreover, several streams of information are processed in parallel from any retinal point by several dozens of inter-neuronal subtypes (Kolb et al 2001) before contrast, brightness, orientation movement and color are finally coded as modulation of ganglion cell action potential series. While chromatic information is not of utmost priority, a differential characterization will nevertheless be required when designing achromatic processing modules for basic representation of image components. Similarly, the high sensitivity pathways originating from rod photoreceptors may be silenced by mimicking daylight intensity (photopic) conditions using adequate pre-amplification of dimmer signals. It may be argued that emulation of cortical afferents from the lateral geniculate nucleus will provide a more appropriate sensory signal for stimulating the primary visual cortex. In particular, it is likely that significant functional transformation of the visual stream occurs at this stage, both as a result of feedforward circuitry, as well as the influence (at present poorly understood) of the very large cortical-fugal feedback pathway. However, the functional highorder relationships between the outputs of retinal ganglion cells are probably preserved to some extent at the level of the geniculate output (Reich et al., 1997; Reinagel et al., 1999; Reinagel and Reid 2000). A crucial experimental advantage of using a retinally based model is the ability to make multiple simultaneous recordings that allow derivation of firing statistics between ganglion cells, but, at present, such technology does not exist for functional (i.e.visually evoked) recordings of the lateral geniculate. One has thus to focus on providing an image of the retinal, as opposed to geniculate, output to the cortical implant. Increasing evidence suggests that the retina and the brain utilize distributed codes that can only be analyzed by simultaneously recording the activity of multiple neurons (Borst and Theunissen, 1999; Fernandez et al., 2000; Normann et al., 2001; Warren et al., 2001; Wilke et al., 2001; Greschner et al 2002; Schnitzer and Meister 2003). Far from a simple transducer of light into electrical neural impulses, the retina performs a locally-computed spatiotemporal contrast enhancement function and a very efficient compression of visual information. These tasks are essential to provide a high adaptation capability to very different lighting conditions, a high noise immunity, and to efficiently communicate the visual information by means of a limited number of optic nerve fibers. Thus, our entire experience of the external visual world derives from the concerted activity of a limited number of ganglion cells in the retinal output layer which have to represent all the features of objects in the visual world, namely their color, intensity, shape, movement, and the change of these features in time. This representation has to be unequivocal and fast in order to ensure object recognition for any single stimulus presentation within a few hundreds of milliseconds (Bialek et al., 1991; de Ruyter van Steveninck et al 1997; Smirnakis et al 1997). The development of a bioinspired visual encoder poses therefore an information processing challenge without parallel in neuroscience, computer science or communication technologies. Consequently, several researchers are developing complex retina encoders for visual implants (Eckmiller, 1997; Rizzo et al., 2001; Eckmiller et al., 2005; Fernandez et al., 2005a; Morillas et al., 2005, 2007). The bioinspired retinal model we are currently using is based on electrophysiological recordings from populations of retinal ganglion cells. Through a set of parametrized filters and functions that simulate the complex operations of the neural
372
retina, we obtain a portable model that can be easily translated into a hardware description for automatic synthesis using the appropriate tools. The input images are captured by a photosensor array (preferably a logarithmic response camera) and are processed by a set of separate spatial and temporal filters that enhance specific features of the captured visual information. The model can take into account the irregular distribution of photoreceptors within the human retina: a high density of pixels and smaller receptive field sizes in central areas; lower density and bigger receptive fields in peripheral areas. A gain factor can be specified for every individual channel as well as a global gain. The next stage reduces the information to the resolution of the electrode matrix, with the option of defining specific receptive field shapes and sizes. Finally, a mapping and neuromorphic coding (into output charge-balanced pulses that can be sent to each electrode) is carried out and feeds the radiofrequency link that goes to the microelectrode array. The model implemented in the present version of the software is a simplified version of an integrate-and-fire spiking neuron (Gerstner and Kistler 2002). Each neuron accumulates input values coming from its receptive field until it reaches a programmable threshold. Then it fires and discharges the accumulated value. The model also includes a leakage term to make the accumulated value diminish for low or null input values. Although the model is essentially analog, we have chosen a digital hardware implementation to have a more flexible and standardizable approach. In the future, this implementation could be easily customized for each implanted device. Thus, the implementation of the model in digital hardware provides a flexible design, achieving a high performance with response times several orders of magnitude lower than that of biological systems. The whole system is presently able to work properly up to 40 MHz. This means that 40,000 electrodes could be stimulated with an inter-spike temporal resolution equal to or lower than 1 msec. The use of reconfigurable circuitry (FPGA) permits to adjust or even change the spiking model easily (Pelayo et al 2004; Morillas et al 2007). Furthermore, this technology can be also used for psychophysical experimentation in order to determine the best image processing strategy and to get insights into the optimal number of electrodes, grey levels and grid size. The psychophysical experimentation can be also used to investigate if the image processor speed is adequate for ordinary tasks such as walking or reading big characters (Dagnelie et al 2007, Dagnelie 2008).
BALANCING SEFETY AND EFFICACY OF MULTIPLE INTARCORTICAL ELECTRODES

A neuroprosthetic system must be implanted into the nervous system and remain fully functional for periods that will eventually extend to many decades. Therefore, these devices must be highly biocompatible and be able to resist the attack of biological fluids, proteases, macrophages or any metabolic substances. Furthermore, it is necessary to take into account the possible damage of neural tissues by permanent charge injection using multi-electrode arrays and the most effective means of stimulating the cerebral cortex. On the other hand, although no major safety concerns emerged from pilot studies carried out to date, the risks of the surgical procedures involved are not negligible. All these considerations place unique constraints on the architecture, material, and surgical techniques used in the implementation
373
of neural interfaces (Normann et al 1999; Fernandez et al 2005a). Important attributes of this interfacing should be: Special design for insertion in any part of cerebral cortex, including sulci of highly folded cortices such as those of humans. Reaching neurons at the desired three-dimensional location. Bi-directional communication with neurons and ensembles of neurons. Providing appropriate and stable electrical interfacing. Minimizing tissue damage and scarring. Rendering the devices as inert as possible from the biocompatibility, biostability and biofouling standpoints.
Hewing to these basic requirements, great efforts have been made to develop penetrating multi-electrodes, with dimensions of the same order of magnitude as the cortical cells, that can be used to excite neurons more selectively and with electrical currents much smaller than those used by surface electrodes (Bak et al 1990b, Campbell et al., 1991; Jones et al., 1992; Kim and Wise, 1996; Nordhausen et al., 1996; Schmidt et al., 1996; Rousche and Normann, 1998; Normann et al., 1999; Stieglitz and Meyer, 1999; Troyk et 2003). In this context, the two main approaches are multiple insulated metal microwires (Troyk et al 2003, Bradley et al 2005, Kim et al 2006b, Musallam et al 2007) and thin-film penetrating microelectrode arrays such as the Utahs array (Normann et al., 1999; Maynard et al., 2000; House et al., 2006; Normann, 2007), the Michigan array (Snellings et al., 2006; Seymour and Kipke, 2007) and the new NeuroProbes arrays that are being developed in the context of an Integrated European Project (http://neuroprobes.org) and which integrate biosensors, electronics and microfluidics. Once a particular type of electrode is selected, the next step is to design a surgical procedure for electrode implantation. Although the surgical implantation of many microwires is relatively easy and can be used to compensate for the complex structure of the brain, there are still some disadvantages to use microwires, namely the low signal-to-noise ratio, the difficulty of integrating on-board electronics and the lack of methods to produce quality microwire electrodes using batch processing. For these reasons, many researchers are using silicon-based high-density microelectrode arrays to record and stimulate small populations of neurons. In this context, one microelectrode array that has been used extensively in acute and chronic recording experiments and is currently being tested in pilot clinical trials under an Investigational Device Exemption (IDE) from the Food & Drug Administration (FDA) is the Utah Electrode Array (UEA). This array of penetrating electrodes has been designed to compromise as little cortical volume as possible and has 100 microelectrodes, each 1.5 mm long, arranged in a square grid contained in a package 4.2 mm by 4.2 mm (Figure 2). Each needle has been made as slender as possible, yet retains sufficient strength to withstand the implantation procedure. However, even though the individual microelectrodes of the UEA are extremely sharp, early attempts to implant an array of 10 x 10 electrodes into the visual cortex in different animal models only deformed the cortical surface and resulted in incomplete implantation. A system that rapidly inserts the UEA into the cortex has been developed (Rousche and Normann 1992) allowing implantation in a manner that minimizes
374
dimpling and compression of the subjacent structures. The implantation is so rapid that the cortex experiences only slight mechanical dimpling and the insertion is generally complete. The most typical findings in acute experiments are occasional microhemorrhages emanating from the electrode tracks, probably due to the high probability of electrode tips encountering one or more blood vessels during implantation. This typically resolves spontaneously and, aside from a few mechanically-distorted and somewhat hyperchromic neurons, the neurons near most tracks appear normal (Figure 3). Furthermore, single unit recordings of neural activity can often by made within hours after the implantation.
Figure 2. Scanning electron micrograph of the Utah Electrode Array. This three-dimensional device contains 100 microelectrodes each of which is 1.5 mm long and tapers to a sharpened tip.
Figure 3. Cat brain implanted for six months with the Utah Electrode Array (UEA). (A) Photograph of a dorsal view of the array after implanting in the cortex. (B,C) Light micrographs at different magnifications. Although there are a few mechanically-distorted cells and some fibrotic tissue in the vicinity of the electrode tracks, neurons in close proximity to the electrodes appear normal. Calibration bar A= 1mm; B,C= 100 m.
375
An important problem reported with all available microelectrodes to date is long-term viability and biocompatibility. Recent studies of the response of neural tissue to stab wounds have shown that there are acute and chronic inflammatory reactions which affect both the neural tissue and the surface of the microelectrodes (Hoogerwerf and Wise, 1994; Woodford et al., 1996; Heiduschka and Thanos, 1998; Agnew et al., 1999; Liu et al., 1999; Majji et al., 1999; Turner et al., 1999; McCreery et al., 2002). These reactions often result in damage to neurons and microelectrodes and lead to the proliferation of a glial scar around the implanted probes which prevents neurons to be recorded or stimulated (Fawcet and Asher, 1999; Turner et al., 1999). The reasons for the inflammatory response lie in molecular and cellular reactions to foreign surfaces (Edell et al 1992; Heiduschka and Thanos, 1998). These responses can be controlled and one of the big challenges in this field is to create new, more biocompatible surfaces for use in neurosurgery and brain implants. The solution may involve coating the electrodes with bio-active molecules that are slowly released into the surrounding brain tissue. Another important issue for the design of a useful cortical visual neuroprosthesis, not yet addressed, is whether a change in retinotopic organization results from electrical stimulation. It is known that repeated sensory stimulation of either primary visual cortex or other primary sensory cortices can lead to changes in the representation of the sensory input. Additionally, it is known that changes in the cortical representation, receptive size and synaptic efficacy can occur as a result of repeated electrical stimulation of auditory, somatosensory, motor and visual cortex (Dinse et al., 1993; Sil'kis and Rapoport, 1995; Maldonado and Gerstein, 1996a, b; McCreery et al 1997, Heusler et al., 2000; Talwar and Gerstein, 2001; Teskey et al., 2002; Warren and Normann, 2005). Preliminary results are inconsistent and more work needs to be done (Warren and Normann, 2005). Finally, by implanting penetrating microelectrodes within the visual cortex, highly selective stimulation can, in principle, be achieved. Experiments to determine the levels of current injections that are required to evoke sensory percepts via intracortical microstimulation have shown that currents as low as 2 A are effective when using a microelectrode positioned in the deepest layers of V1 (Schmidt et al 1996). In spite of this, mixed reports have arisen regarding the effects of pulse polarity of phosphene induction (Tehovnik et al 2005), although cathodal pulses seem more effective than anodal pulses for inducing phosphenes (Schmidt et al 1996). Moreover, increments in current or pulse duration increase the number of elements activated, because of the higher current densities generated and because of the greater overall volume of tissue activated (Schmidt and McIntosh, 1990; Schmidt et al., 1996, Bradley et al., 2005; Tehovnik et al., 2005). These results concur with what would be expected for depth stimulation (Ranck 1975). Nonetheless, more data on the most effective means of stimulating the cerebral cortex and the possible damage of neural tissues by permanent charge injection using intracortical multielectrode arrays are still needed.
376
NONINVASIVE SELECTION OF SUITABLE CANDIDATES

The selection of a specific individual for a visual implant is not straightforward. There are no strict standardized criteria for accepting or rejecting a candidate, nor for the best rehabilitation procedure for every type of blindness. Generally, a choice could be made between different approaches and/or rehabilitation procedures depending on availability, efficacy or rejection of invasive methods (Veraart et al., 2004; Dowling, 2005, Dagnelie 2006), but a pre-surgical protocol and improved methods for predicting success with a visual neuroprosthesis need to be developed (Fernandez et al., 2005a; Merabet et al., 2007a; Dagnelie, 2008). In general, it is considered crucial, at least in these preliminary stages, that the subject should have no residual visual and had no significant benefit from a conventional visual aid. However, these seemingly straightforward criteria did not always work well in practice and there are different definitions of residual vision and significant benefit from a visual aid. Needs and wishes of individual subjects are also significant variables for implant candidacy. This issue is further complicated, because it is not possible to predict success with a visual implant in a specific subject. Clearly, our knowledge regarding visual system anatomy and function may allow for crude bio-inspired models and strategies of stimulation. However, what has not been discussed is how the type, onset, duration and temporal profile of an individual visual loss may have repercussions on the success of the device. To decide what time is the most suitable in this regard is a difficult double-faced ethical decision (Veraart et al 2004). Recognizing these limitations, it is generally acknowledged that previous visual experience is necessary for the patient to interpret and recognize the visual patterns that are generated. In this context, it has been presumed that, if blindness occurs after the age of ten years (i.e. with respect to the critical period), visual pathways should develop normally and thus remain excitable (Veraart et al., 2004; Dagnelie, 2006). Improved networking among research groups would clearly help in developing a common body of standardized test and standardized selection criteria (Dagnelie 2008). Among them, possible criteria to specify the characteristics of adults who are potential users of a visual neuroprosthesis should include (Merabet et al 2007a): Visual function criteria and pre- and post-implantation testing. Indications in favor of a cortical visual implant could include profound and bilateral visual loss, but more objective and quantitative criteria should be developed. These studies need to incorporate a more quantifiable means of estimating risk and benefit for a given candidate. Electrophysiological criteria. Measurement of electroretinogram (ERG) as well as visual evoked potentials (VEP) should be a basic component in candidate selection. Improved methods for predicting success with a visual implant need to be developed. These methods could include tests using data from TMS of visual pathways and imaging techniques such as functional magnetic resonance (fMRI). Medical, anatomical and surgical criteria. The usual candidate should be a healthy adult. The medical history should include the onset and evolution of the blindness, physical examination, and laboratory tests to include or exclude candidates and to
377
assist the implant team in planning a full-fledged program, including post-implant training. Evaluation of the risks and limitations associated with the surgery and every specific approach. These are usually small for persons in good general health, but increase with age and other pathological conditions. Quantitative evaluation of the subjects needs and performance using standardized psychophysical and behavioral methods. Psychological testing to exclude psychiatric and other possible mental disorders that could affect the success of the visual prosthesis. Another major prerequisite for the possible clinical application of this neuroprosthetic approach is related to cortical plasticity and reorganization in severe vision loss (Fernandez et al., 2005a; Merabet et al., 2005; Merabet et al., 2007a; Merabet et al., 2007b), in particular, to how the brain adapts to the loss of sight and to the physiological and functional fate of cortical areas normally associated with the processing of visual information once vision is lost (e.g. from ocular disease or trauma). Clear answers are now becoming crucial (Merabet et al 2005). The visual cortex of potential candidates for a cortical prosthesis has to be capable of processing visual information, but there is evidence that the occipital parts of the brain utilized by sighted subjects to process visual information are transformed in some blind subjects and utilized to process tactile and auditory stimuli (Cohen et al., 1997; Cohen et al., 1998; Pascual-Leone et al., 1998; Cohen et al., 1999) and even higher cognitive functions such as grammatical and linguistic processing (Rauschecker, 1995; Roder et al 1999; Van Boven et al., 2000; Bavelier and Neville, 2002). Specifically, it appears that the occipital visual cortex, normally associated with the processing of visual information, is recruited in a functional and compensatory manner to process the remaining sensory modalities (Hamilton et al., 2000; Burton et al., 2002, Amedi et al., 2003; Gizewski et al., 2003; Maeda et al., 2003), so-called cross-modal plasticity. Pascual-Leone et al (1999) reported the case of a blind woman who was an extremely proficient Braille reader working as an editor for a newsletter for the blind. In a most unfortunate event, this woman became unable to read Braille (Braille alexia) following a stroke. Indeed, she became unable to decode any complex tactile information, while otherwise remaining neurologically intact. Contrary to expectations, the lesion in this patient did not affect the somatosensory cortex, but damaged the occipital pole bilaterally. In this case, a cortical neuroprosthesis would not be useful. However, some blind or severely visually deprived subjects do not adapt well to the loss of sight. In these subjects, preliminary studies suggest that the occipital cortex is not or, at most, only partially reorganized (Fernandez et al 2002). This appears to be particularly the case in severely visually impaired, but not absolutely blind subjects. Late blind subjects are more likely to fall into this category (Alfaro et al., 2006; Alfaro and Fernandez, 2008). Such subjects never fully adapt to the world of touch and sound and for them a cortical visual neuroprosthesis might be a functionally desirable solution. In order to get insights into these issues, a reliable, non-invasive method to study the degree of cross-modal plasticity and the degree of remaining functional visual cortex in blind subjects has been developed using TMS. The procedure allows the systematic mapping of the sensations induced by focal and non invasive stimulation of the human occipital cortex and
378
provides a method for their quantification (Fernandez et al 2002) and, in combination with other brain imaging technologies and methods, could be very useful to improve our present understanding of the physiological reorganization and plastic changes in the brain of blind subjects as a consequence of their adaptation to the loss of sight. Case studies of surgical sight restoration following long-term visual deprivation (Fine et al 2003, Gregory 2003) also provide relevant insights. For example, following ocular surgical procedures aimed at regaining some degree of functional vision, patients blinded for many years experience profound difficulty in various visual tasks, particularly those requiring the identification and recognition of objects. Interestingly, if patients are allowed to explore the same object through touch, they can recognize it immediately as to register their newly acquired visual percepts with their existing senses. These findings demonstrate that, even when vision reaches the brain physiologically, visual perception remains nonetheless impaired. Thus, the simple restoration of a lost sensory input may not by itself be enough to reconstitute the sense it normally provides and active visual rehabilitation may be necessary to maximize the adaptation and get the most from these devices. As demonstrated by mapping studies after micro-infarcts, it is clear that behavior is one of the most powerful modulators of post-injury recovery and, therefore, behavioral intervention to enhance recovery is becoming increasingly popular (Pascual-Leone et al., 2005; Nudo, 2006; Alonso-Alonso et al 2007). These rehabilitation therapies have significantly improved the quality of life of many of patients after brain damage and suggest that this ability of the brain to reorganize itself by experience-dependent neural plasticity could be also used to develop new training strategies to accelerate learning and maximize the adaptation to prosthetic vision devices. At the same time, experiments with human and non-human subjects have shown that the correspondence between the spatial location of the stimulation site in the cortex and the position of the evoked phosphene in the visual field can present considerable deformations (Warren et al 2001, 2004). This is especially remarkable for high density arrays of electrodes in which the correspondence between the stimulation and the perceptual space is highly non-linear and non-conformal. This finding might reflect the complex interconnections among the neural cells that respond to stimulation in the area of influence of a specific microelectrode and agrees with the randomness in the location of phosphenes evoked in the primary visual cortex in man (Brindley and Lewin, 1968; Brindley et al., 1972; Brindley, 1982; Bak et al., 1990b; Schmidt et al., 1996). Therefore, some form of spatial remapping between the bioinspired encoder and the stimulation pattern delivered to the array of electrodes implanted in the visual cortex is highly recommendable (Normann et al 2001; Fernandez et al., 2002; Eckmiller et al., 2005; Fernandez et al., 2005a; Romero et al., 2008). As a result, appropriate learning, remapping and rehabilitation strategies could potentially help to modulate the plasticity of the brain and contribute to improved performance and more concordant perceptions. In sum, immediately after implantation, the evoked phosphenes are likely to induce a poor perception of an object: appropriate learning and rehabilitation strategies will contribute to provide concordant perceptions. Contrary to a previous belief that a given electrical pattern of stimulation generated on the surface of the visual cortex would generate a visual percept of the same pattern, current notions suggest that the relation between the pattern of stimulation and visual percept is nonlinear and nonconformal. Moreover, the patterns of stimulation are
379
likely to change over time, due to factors such as changes in electrode stimulation thresholds, the progression of disease and the effects of learning and plasticity (Normann et al 1996).
CONCLUSIONS
Whereas many attributes characterize a visual scene (for example, color, form, motion and field of view), current strategies in visual prosthesis development are aimed to address only the most basic component, that is, spatial detail. More complex perceptions could be generated by simply increasing the number of stimulating electrodes, thereby increasing the potential resolution of images being produced (Thompson et al., 2003; Dagnelie et al., 2006; Dagnelie, 2008). However, given that the complex relationship between patterns of electrical stimulation and target visual percepts remains largely unknown, this view appears to be an oversimplification. Thus, an increase in stimulus density may initially be perceptually meaningless, rather than helpful (Veraart et al., 2003; Fernandez et al., 2005b). At the same time, current psychophysical evidence suggest that the human visual system is able to identify and extract complex information (such as identifying a human faces) from relatively poor quality images by relying upon multiple salient visual features and cues (Sinha 2002). This suggests that the assumption that visual perception will improve by concentrating on increasing image resolution alone (as opposed to other visual attributes, such as, for example, the temporal encoding related to motion perception) may be incorrect. On the other hand, we should be aware that the stimulation patterns evoking visual perception can vary with a number of parameters such as pulse duration, duty cycle, amplitude, number of pulses in a train, etc. Furthermore, these parameters vary for every channel and might take on very different values for each implanted patient. Even for a given subject, these values could change due to electrode encapsulation or neuroplastic changes. The modulation of such plasticity will be crucial in developing and projecting the success of novel, visual neuroprosthetic strategies, which has implications for rehabilitative training and device development. Although the full restoration of vision seems to be impossible, the discrimination of shape and location of objects could allow blind subjects to navigate in a familiar environment and to read enlarged text, resulting in a substantial improvement in the quality of life of blind and visual impaired persons. A highly functional cortical visual prosthesis is still far into the future, but the success of the cochlear implant encourages the pursuit of this neurotechnological application.
REFERENCES
Adams DL, Horton JC. Shadows Cast by Retinal Blood Vessels Mapped in PrimaryVisual Cortex. Science 2002, 298:572-576. Agnew BJ, Duman JG, Watson CL, Coling DE, Forte JG.Cytological transformations associated with parietal cell stimulation: critical steps in the activation cascade. J Cell Sci 1999; 112:2639-2646.
380
Ahnelt PK, Kolb H. The mammalian photoreceptor mosaic-adaptive design. Prog Retin Eye Res 2000; 19:711-777. Alfaro A, Fernandez E. Transcranial magnetic stimulation: contributions to the study of the visual pathway. Rev Neurol 2008; 46:282-288. Alfaro A, Concepcion L, Merabet L, Fernandez E. An atypical presentation of visual hallucinatory experiences following prolonged blindness. Neurocase 2006; 12:212-215. Alonso-Alonso M, Fregni F, Pascual-Leone A. Brain stimulation in poststroke rehabilitation. Cerebrovasc Dis 2007; 24 Suppl 1:157-166. Amedi A, Raz N, Pianka P, Malach R, Zohary E. Early 'visual' cortex activation correlates with superior verbal memory performance in the blind. Nat Neurosci 2003; 6:758-766 Bak M, Girvin JP, Hambrecht FT, Kufta CV, Loeb GE, Schmidt EM. Visual sensations produced by intracortical microstimulation of the human occipital cortex. Med Biol Eng Comput 1990; 28:257-259. Bavelier D, Neville HJ. Cross-modal plasticity: where and how? Nat Rev Neurosci 2002; 3:443-452. Berardi N, Pizzorusso T, Ratto GM, Maffei L. Molecular basis of plasticity in the visual cortex. Trends Neurosci 2003; 26:369-378. Bialek W, Rieke F, de Ruyter van Steveninck RR, Warland D. Reading a neural code. Science 1991; 252:1854-1857. Borst A, Theunissen FE. Information theory and neural coding. Nat Neurosci 1999; 2:947957. Bradley DC, Troyk PR, Berg JA, et al. Visuotopic mapping through a multichannel stimulating implant in primate V1. J Neurophysiol 2005; 93:1659-1670 Brindley GS. Effects of electrical stimulation of the visual cortex. Hum Neurobiol 1982; 1:281-283. Brindley GS, Lewin WS. The sensations produced by electrical stimulation of the visual cortex. J Physiol (Lond) 1968; 196:479-493. Brindley GS, Donaldson PEK, Falconer MA, Rusthon DN. The extent of the region of occipital cortex that when stimulated gives phosphenes fixed in the visual field. J Physiol (Lond) 1972; 225:57-58. Brodmann K. Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues. Leizpig. 1909 Burton H, Snyder AZ, Conturo TE, Akbudak E, Ollinger JM, Raichle ME. Adaptive changes in early and late blind: a fMRI study of Braille reading. J Neurophysiol 2002; 87:589607. Campbell PK, Jones KE, Huber RJ, Horch KW, Normann RA. A silicon-based, threedimensional neural interface: manufacturing processes for an intracortical electrode array. IEEE Trans Biomed Eng 1991; 38:758-768. Canavero S. Total eye transplantation for the blind: a challenge for the future. Med Hypotheses 1992; 39: 201-211 Carmena JM, Lebedev MA, Henriquez CS, Nicolelis MA. Stable ensemble performance with single-neuron variability during reaching movements in primates. J Neurosci 2005; 25:10712-10716.
381
Cogan SF, Troyk PR, Ehrlich J, Gasbarro CM, Plante TD. The influence of electrolyte composition on the in vitro charge-injection limits of activated iridium oxide (AIROF) stimulation electrodes. J Neural Eng 2007; 4:79-86. Cohen LG, Weeks RA, Sadato N, Celnik P, Ishii K, Hallett M. Period of susceptibility for cross-modal plasticity in the blind. Ann Neurol 1999; 45:451-460. Cohen LG, Ziemann U, Chen R, et al. Studies of neuroplasticity with transcranial magnetic stimulation. J Clin Neurophysiol 1998; 15:305-324. Cohen LG, Celnik P, Pascual-Leone A, et al. Functional relevance of cross-modal plasticity in blind humans. Nature 1997; 389:180-183. Dagnelie G. Visual prosthetics 2006: assessment and expectations. Expert Rev Med Devices 2006; 3:315-325. Dagnelie G. Psychophysical Evaluation for Visual Prosthesis. Annu Rev Biomed Eng. 2008; 10: 339-368 Dagnelie G, Barnett D, Humayun MS, Thompson RW, Jr. Paragraph text reading using a pixelized prosthetic vision simulator: parameter dependence and task learning in freeviewing conditions. Invest Ophthalmol Vis Sci 2006; 47:1241-1250. Dagnelie G, Keane P, Narla V, Yang L, Weiland J, Humayun M. Real and virtual mobility performance in simulated prosthetic vision. J Neural Eng 2007; 4:S92-101. de Ruyter van Steveninck RR, Lewen GD, Strong SP, Koberle R, Bialek W. Reproducibility and variability in neural spike trains. Science 1997; 275:1805-1808. Delbeke J, Oozeer M, Veraart C. Position, size and luminosity of phosphenes generated by direct optic nerve stimulation. Vision Res 2003; 43:1091-1102. Delbeke J, Pins D, Michaux G, Wanet-Defalque MC, Parrini S, Veraart C. Electrical stimulation of anterior visual pathways in retinitis pigmentosa. Invest Ophthalmol Vis Sci 2001; 42:291-297. Dinse HR, Recanzone GH, Merzenich MM. Alterations in correlated activity parallel ICMSinduced representational plasticity. Neuroreport 1993; 5:173-176. Dobelle WH (2000) Artificial vision for the blind by connecting a television camera to the visual cortex. Asaio J 46:3-9. Dobelle WH, Mladejovsky MG. Phosphenes produced by electrical stimulation of human occipital cortex, and their application to the development of a prosthesis for the blind. J Physiol (Lond) 1974; 243:553-576. Dobelle WH, Mladejovsky MG, Girvin JP. Artificial vision for the blind: electrical stimulation of visual cortex offers hope for a functional prosthesis. Science 1974; 183:440-444. Dobelle WH, Mladejovsky MG, Evans JR, Roberts TS, Girvin JP. 'Braille' reading by a blind volunteer by visual cortex stimulation. Nature 1976; 259:111-112. Dowling J. Artificial human vision. Expert Rev Med Devices 2005; 2:73-85. Dreher B, Burke W, Calford MB. Cortical plasticity revealed by circumscribed retinal lesions or artificial scotomas. Prog Brain Res 2001; 134:217-246. Eckmiller R. Learning retina implants with epiretinal contacts. Ophthalmic Res 1997; 29:281289. Eckmiller R, Neumann D, Baruth O. Tunable retina encoders for retina implants: why and how. J Neural Eng 2005; 2:S91-S104.
382
Edell DJ, Van Toi V, McNeil VM, Clark LD. Factors influencing the biocompatibility of insertable silicon microshafts in cerebral cortex. IEEE Trans Biomed Eng 1992; 39:635643. Fawcet JW, Asher RA. The glial scar and central nervous system repair. Brain Res Bull 1999; 49:377-391 Fernandez E, Ferrandez J, Ammermuller J, Normann RA. Population coding in spike trains of simultaneously recorded retinal ganglion cells. Brain Res 2000; 887:222-229. Fernandez E, Pelayo F, Romero S, Bongard M, Marin C, Alfaro A, Merabet L. Development of a cortical visual neuroprosthesis for the blind: the relevance of neuroplasticity. J Neural Eng 2005a; 2:R1-R12. Fernandez E, Pelayo F, Romero S, Bongard M, Marin C, Alfaro A, Merabet L. Development of a cortical visual neuroprosthesis for the blind: the relevance of neuroplasticity. J Neural Eng 2005b; 2:R1-12. Fernandez E, Alfaro A, Tormos JM, et al. Mapping of the human visual cortex using imageguided transcranial magnetic stimulation. Brain Res Protoc 2002; 10:115-124. Fine I, Wade AR, Brewer AA, et al. Long-term deprivation affects visual perception and cortex. Nat Neurosci 2003; 6:915-916. Gerstner W, Kistler W. Spiking Neuron Models. Cambridge: Cambridge University Press, 2002 Gizewski ER, Gasser T, de Greiff A, Boehm A, Forsting M. Cross-modal plasticity for sensory and motor activation patterns in blind subjects. Neuroimage 2003; 19 :968-975. Gregory RL. Seeing after blindness. Nat Neurosci 2003; 6:909-910. Greschner M, Bongard M, Rujan P, Ammermuller J. Retinal ganglion cell synchronization by fixational eye movements improves feature estimation. Nat Neurosci 2002, 5:341-347. Hamilton R, Keenan JP, Catala M, Pascual-Leone A. Alexia for Braille following bilateral occipital stroke in an early blind woman. Neuroreport 2000; 11:237-240. Heiduschka P, Thanos S. Implantable bioelectronic interfaces for lost nerve functions. Prog Neurobiol 1998; 55:433-461. Heusler P, Cebulla B, Boehmer G, Dinse HR. A repetitive intracortical microstimulation pattern induces long-lasting synaptic depression in brain slices of the rat primary somatosensory cortex. Exp Brain Res 2000; 135:300-310. Hochberg LR, Serruya MD, Friehs GM, Mukand JA, Saleh M, Caplan AH, Branner A, Chen D, Penn RD, Donoghue JP. Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 2006; 442:164-171. Hoogerwerf AC, Wise KD. A 3D micro-electrode array for chronic neural recording. IEEE Trans Biomed Eng 1994; 41:1136-1146. House PA, MacDonald JD, Tresco PA, Normann RA. Acute microelectrode array implantation into human neocortex: preliminary technique and histological considerations. Neurosurg Focus 2006; 20:E4. Hu Z, Troyk PR, Detlefsen DE. Analysis of capacitive coupling within microelectrode array. Conf Proc IEEE Eng Med Biol Soc 2006; 1:3365-3368. Humayun MS. Intraocular retinal prosthesis. Trans Am Ophthalmol Soc 2001, 99:271-300. Humayun MS, Weiland J, Fujii GY, et al. Visual perception in a blind subject with a chronic microelectronic retinal prosthesis. Vision Res 2003; 43:2573-2581.
383
Jones BW, Watt CB, Frederick JM, et al. Retinal remodeling triggered by photoreceptor degenerations. J Comp Neurol 2003; 464:1-16. Jones KE, Campbell PK, Normann RA. A glass/silicon composite intracortical electrode array. Ann Biomed Eng 1992; 20:423-437. Kim C, Wise KD. A 64-site multishank CMOS low-profile neural stimulating probe. IEEE J Solid-State Circuits 1996; 31:1230-1238. Kim SJ, Manyam SC, Warren DJ, Normann RA.Electrophysiological Mapping of Cat Primary Auditory Cortex with Multielectrode Arrays. Ann Biomed Eng. 2006a; 34: 300309 Kim T, Troyk PR, Bak M. Active floating micro electrode arrays (AFMA). Conf Proc IEEE Eng Med Biol Soc 2006b; 1:2807-2810. Kolb H, Nelson R, Ahnelt P, Cuenca N. Cellular organization of the vertebrate retina. Prog Brain Res 2001; 131:3-26. Liu X, McCreery DB, Carter RR, Bullara LA, Yuen TG, Agnew WF. Stability of the interface between neural tissue and chronically implanted intracortical microelectrodes. IEEE Trans Rehabil Eng 1999; 7:315-326. Lwenstein K, Borchart M (1918) Symptomatologie und elektrische Reizung bei einer Schubverletzung des Histerhauptlappens. Dtsch Z Nervenheilk 58:264. Maeda K, Yasuda H, Haneda M, Kashiwagi A. Braille alexia during visual hallucination in a blind man with selective calcarine atrophy. Psychiatry Clin Neurosci 2003; 57:227-229. Majji AB, Humayun MS, Weiland JD, Suzuki S, D'Anna SA, de Juan E, Jr. Long-term histological and electrophysiological results of an inactive epiretinal electrode array implantation in dogs. Invest Ophthalmol Vis Sci 1999; 40:2073-2081. Maldonado PE, Gerstein GL. Reorganization in the auditory cortex of the rat induced by intracortical microstimulation: a multiple single-unit study. Exp Brain Res 1996a; 112:420-430. Maldonado PE, Gerstein GL. Neuronal assembly dynamics in the rat auditory cortex during reorganization induced by intracortical microstimulation. Exp Brain Res 1996b; 112:431441. Marc RE, Jones BW. Retinal remodeling in inherited photoreceptor degenerations. Mol Neurobiol 2003; 28:139-147. Maynard EM, Fernandez E, Normann RA. A technique to prevent dural adhesions to chronically implanted microelectrode arrays. J Neurosci Methods 2000; 97:93-101. McCreery D, Agnew WF, Bullara L. The effects of prolonged intracortical microstimulation on the excitability of pyramidal tract neurons in the cat. Ann Biomed Eng 2002; 30:107119. McCreery DB, Yuen TG, Agnew WF, Bullara LA. A characterization of the effects on neuronal excitability due to prolonged microstimulation with chronically implanted microelectrodes. IEEE Trans Biomed Eng 1997; 44:931-939. Merabet LB, Rizzo JF, 3rd, Pascual-Leone A, Fernandez E. 'Who is the ideal candidate?': decisions and issues relating to visual neuroprosthesis development, patient testing and neuroplasticity. J Neural Eng 2007a; 4:S130-135.
384
Merabet LB, Rizzo JF, Amedi A, Somers DC, Pascual-Leone A. What blindness can tell us about seeing again: merging neuroplasticity and neuroprostheses. Nat Rev Neurosci 2005; 6:71-77. Merabet LB, Swisher JD, McMains SA, Halko MA, Amedi A, Pascual-Leone A, Somers DC (2007b) Combined activation and deactivation of visual cortex during tactile sensory processing. J Neurophysiol 97:1633-1641. Morillas C, Romero S, Martnez A, Pelayo F, Fernandez E. A computational tool to test neuromorphic encoding schemes for visual neuroprosthesis. In: 8th International WorkConference on Artificial Neural Networks (IWANN 2005), pp 510-517. Barcelona. 2005 Morillas CA, Romero SF, Martinez A, Pelayo FJ, Ros E, Fernandez E. A design framework to model retinas. Biosystems 2007; 87:156-163. Mller J. Handbuch der Physiologie des Menschen. Coblenz: Verlag von J. Hlscher.1837 Musallam S, Bak MJ, Troyk PR, Andersen RA. A floating metal microelectrode array for chronic implantation. J Neurosci Methods 2007; 160:122-127. Nordhausen CT, Maynard EM, Normann RA (1996) Single unit recording capabilities of a 100 microelectrode array. Brain Res 726:129-140. Normann RA. Technology insight: future neuroprosthetic therapies for disorders of the nervous system. Nat Clin Pract Neurol 2007; 3:444-452. Normann RA, Maynard E, Guillory KS, Warren DJ. Cortical implants for the blind. IEEE Spectrum, 1996; 54-59. Normann RA, Maynard EM, Rousche PJ, Warren DJ. A neural interface for a cortical vision prosthesis. Vision Res 1999; 39:2577-2587. Normann RA, Warren DJ, Ammermuller J, Fernandez E, Guillory S. High-resolution spatiotemporal mapping of visual pathways using multi-electrode arrays. Vision Res 2001; 41:1261-1275. Nudo RJ. Plasticity. NeuroRx 2006; 3:420-427. Paninski L, Fellows MR, Hatsopoulos NG, Donoghue JP. Spatiotemporal tuning of motor cortical neurons for hand position and velocity. J Neurophysiol 2004; 91:515-532. Pascual-Leone A, Amedi A, Fregni F, Merabet LB. The plastic human brain cortex. Annu Rev Neurosci 2005; 28:377-401. Pascual-Leone A, Hamilton R, Tormos JM, Keenan JP, Catala MD. Neuroplasticity in the adjustment to blindness. In: In: J.Grafman, Christen Y (eds). Neuronal Plasticity: Building a Bridge from the laboratory to the Clinic. Berlin: Springer-Verlag1999 Pascual-Leone A, Tormos JM, Keenan J, Tarazona F, Canete C, Catala MD. Study and modulation of human cortical excitability with transcranial magnetic stimulation. J Clin Neurophysiol 1998; 15:333-343. Pelayo F, Martinez A, Romero S, Morillas CA, Ros E, Fernandez E. Cortical visual neuroprosthesis for the blind: retina-like softare/hardware preprocessor. In: First International IEEE EMBS Conference on Neural Engineering, pp 20-22. Capri, Italy, 2003 Pelayo FJ, Romero S, Morillas CA, Martinez A, Ros E, Fernandez E. Traslating image sequences into spike patterns for cortical neuro-stimulation. Neurocomputing 2004; 5860:885-892. Penfield W, Rasmussen T. The cerebral cortex of man. New York: Macmillan, 1950
385
Penfield W, Jaspers H. Epilepsy and the functional anatomy of the human brain. London, England: Churchill, 1974 Pezaris JS, Reid RC. Demonstration of artificial visual percepts generated through thalamic microstimulation. Proc Natl Acad Sci U S A 2007; 104:7670-7675. Rank JBJ (1975) Which elements are excited in electrical stimulation of mammalian central nervous system: a review. Brain Res 98:417-440. Rauschecker JP. Compensatory plasticity and sensory substitution in the cerebral cortex. Trends Neurosci 1995; 18:36-43. Reich DS, Victor JD, Knight BW, Ozaki T, Kaplan E. Response variability and timing precision of neuronal spike trains in vivo. J Neurophysiol 1997; 77:2836-2841. Reinagel P, Reid RC. Temporal coding of visual information in the thalamus. J Neurosci 2000; 20:5392-5400. Reinagel P, Godwin D, Sherman SM, Koch C. Encoding of visual information by LGN bursts. J Neurophysiol 1999; 81:2558-2569. Rizzo JF, 3rd, Wyatt J, Loewenstein J, Kelly S, Shire D. Perceptual efficacy of electrical stimulation of human retina with a microelectrode array during short-term surgical trials. Invest Ophthalmol Vis Sci 2003; 44:5362-5369. Rizzo JF, 3rd, Wyatt J, Humayun M, et al. Retinal prosthesis: an encouraging first decade with major challenges ahead. Ophthalmology 2001; 108:13-14. Roder B, Teder-Salejarvi W, Sterr A, Rosler F, Hillyard SA, Neville HJ (1999) Improved auditory spatial tuning in blind humans. Nature 400:162-166. Romero SF, Morillas CA, Pelayo FJ, Fernandez E. Computer-controlled neurostimulation for a visual implant. Biodevices 2008; 1:84-91. Rousche PJ, Normann RA. A method for pneumatically inserting an array of penetrating electrodes into cortical tissue. Ann Biomed Eng 1992; 20:413-422. Rousche PJ, Normann RA. Chronic recording capability of the Utah Intracortical Electrode Array in cat sensory cortex. J Neurosci Methods 1998; 82:1-15. Sanguineti V, Giugliano M, Grattarola M, Morasso P. Neuro-Engineering: from neural interfaces to biological computers. In: Communications Through Virtual Technology: Identity Community and Technology in the Internet Age (Riva R, Davide F, eds), pp 233246. Amsterdam: IOS Press, 2001 Schmidt EM, McIntosh JS. Microstimulation mapping of precentral cortex during trained movements. J Neurophysiol 1990; 64:1668-1682. Schmidt EM, Bak MJ, Hambrecht FT, Kufta CV, O'Rourke DK, Vallabhanath P (1996) Feasibility of a visual prosthesis for the blind based on intracortical microstimulation of the visual cortex. Brain 119:507-522. Schnitzer MJ, Meister M. Multineuronal firing patterns in the signal from eye to brain. Neuron 2003; 37:499-511. Schoups A, Vogels R, Qian N, Orban G. Practising orientation identification improves orientation coding in V1 neurons. Nature 2001; 412:549-553. Sereno MI, McDonald CT, Allman JM. Analysis of retinotopic maps in extrastriate cortex. Cereb Cortex 1994; 4:601-620. Seymour JP, Kipke DR. Neural probe design for reduced tissue encapsulation in CNS. Biomaterials 2007; 28:3594-3607.
386
Sil'kis IG, Rapoport S. Plastic reorganizations of the receptive fields of neurons of the auditory cortex and the medial geniculate body induced by microstimulation of the auditory cortex. Neurosci Behav Physiol 1995; 25:322-339. Sinha P. Recognizing complex patterns. Nat Neurosci 2002; 5 (Suppl):1093-1097. Smirnakis SM, Berry MJ, Warland DK, Bialek W, Meister M. Adaptation of retinal processing to image contrast and spatial scale. Nature 1997; 386:69-73. Snellings A, Anderson DJ, Aldridge JW. Improved signal and reduced noise in neural recordings from close-spaced electrode arrays using independent component analysis as a preprocessor. J Neurosci Methods 2006; 150:254-264. Srivastava NR, Troyk PR. Some solutions to technical hurdles for developing a practical intracortical visual prosthesis device. Conf Proc IEEE Eng Med Biol Soc 2006; 1:29362939. Stieglitz T, Schuettler M, Koch KP. Neural prostheses in clinical applications--trends from precision mechanics towards biomedical microsystems in neurological rehabilitation. Biomed Tech (Berl) 2004; 49:72-77. Talwar SK, Gerstein GL. Reorganization in awake rat auditory cortex by local microstimulation and its effect on frequency-discrimination behavior. J Neurophysiol 2001, 86:1555-1572. Tehovnik EJ, Slocum WM, Carvey CE, Schiller PH. Phosphene induction and the generation of saccadic eye movements by striate cortex. J Neurophysiol 2005; 93:1-19. Teskey GC, Monfils MH, VandenBerg PM, Kleim JA. Motor map expansion following repeated cortical and limbic seizures is related to synaptic potentiation. Cereb Cortex 2002; 12:98-105. Thompson RW, Jr., Barnett GD, Humayun MS, Dagnelie G. Facial recognition using simulated prosthetic pixelized vision. Invest Ophthalmol Vis Sci 2003; 44:5035-5042. Troyk P, Bak M, Berg J, et al. A model for intracortical visual prosthesis research. Artif Organs 2003, 27:1005-1015. Troyk PR, Hu Z, Cogan SF. Assessing polarization of AIROF microelectrodes. Conf Proc IEEE Eng Med Biol Soc 2007a:1726-1729. Troyk PR, DeMichele GA, Kerns DA, Weir RF. IMES: an implantable myoelectric sensor. Conf Proc IEEE Eng Med Biol Soc 2007b:1730-1733. Turner JN, Shain W, Szarowski DH, et al. Cerebral astrocyte response to micromachined silicon implants. Exp Neurol 1999; 156:33-49. Tusa RJ, Palmer LA, Rosenquist AC. The retinotopic organization of area 17 (striate cortex) in the cat. J Comp Neurol 1978; 177:213-235. Van Boven RW, Hamilton RH, Kauffman T, Keenan JP, Pascual-Leone A. Tactile spatial resolution in blind braille readers(1). Am J Ophthalmol 2000; 130:542. Veraart C, Wanet-Defalque MC, Gerard B, Vanlierde A, Delbeke J. Pattern recognition with the optic nerve visual prosthesis. Artif Organs 2003; 27:996-1004. Veraart C, Duret F, Brelen M, Oozeer M, Delbeke J. Vision Rehabilitation in the Case of Blindness. Expert review of Medical Devices 2004; 1:139-153. Veraart C, Raftopoulos C, Mortimer JT, et al. Visual sensations produced by optic nerve stimulation using an implanted self-sizing spiral cuff electrode. Brain Res 1998; 813:181-186.
387
Walter P, Szurman P, Vobig M, et al. Successful long-term implantation of electrically inactive epiretinal microelectrode arrays in rabbits. Retina 1999, 19:546-552. Warren DJ, Normann RA. Functional reorganization of primary visual cortex induced by electrical stimulation in the cat. Vision Res 2005; 45:551-565. Warren DJ, Fernandez E, Normann RA. High-resolution two-dimensional spatial mapping of cat striate cortex using a 100-microelectrode array. Neuroscience 2001; 105:19-31. Warren DJ, Koulakov A, Normann RA. Spatiotemporal encoding of a bar's direction of motion by neural ensembles in cat primary visual cortex. Ann Biomed Eng 2004; 32:1265-1275. Wickelgren I. Biomedical engineering. A vision for the blind. Science 2006; 312:1124-1126. Wilke SD, Thiel A, Eurich CW, et al. Population coding of motion patterns in the early visual system. J Comp Physiol [A] 2001; 187:549-558. Woodford BJ, Carter RR, McCreery D, Bullara LA, Agnew WF. Histopathologic and physiologic effects of chronic implantation of microelectrodes in sacral spinal cord of the cat. J Neuropathol Exp Neurol 1996; 55:982-991. Zrenner E. Will retinal implants restore vision? Science 2002; 295:1022-1025. Zrenner E, Stett A, Weiss S, et al. Can subretinal microphotodiodes successfully replace degenerated photoreceptors? Vision Res 1999; 39:2555-2567.
CONCLUSION
Chapter 23
CORTICAL STIMULATION VERSUS DEEP BRAIN STIMULATION IN NEUROLOGICAL AND PSYCHIATRIC DISORDERS. CURRENT STATE AND FUTURE PROSPECTS
Damianos E. Sakas and Ioannis G. Panourias
Department of Neurosurgery, University of Athens Medical School, Evangelismos Hospital, and P.S. Kokkalis Hellenic Center for Neurosurgical Research, Athens, Greece.
INTRODUCTION
Therapeutic brain stimulation can be applied: a) in networks which are located in cortical and subcortical layers, b) in deep nuclei groups (relay nodes) or c) in combination. CS and DBS belong to the domain of operative neuromodulation and particularly to the field of neural networks surgery (Sakas et al 2007a), defined as the field that studies and applies advancements in neural networks research, digitized stereotactic brain imaging and implantable electrical or electronic devices in order to alter electrically the signal transmission in the nervous system, modulate neural networks and produce therapeutic effects. Therapeutic brain stimulation methods are currently distinguished into two major categories: a) noninvasive and b) invasive. The noninvasive methods (rTMS, tDCS) stimulate electrically the brain cortex after applying electrical currents on the human scalp. The invasive methods require the surgical implantation of specially designed electrodes either in contact with the cortex of the brain, i.e. CS, or in deeply located selected targets, i.e. DBS. Instead of invasive, the term implantable or by implantable devices is much more appropriate, as it does not carry the negative association of a traumatic intervention that is implied by the term invasive CS. In this
Correspondence concerning this article should be addressed to: Dr. Damianos E. Sakas, MD, e-mail: sakasde@med.uoa.gr.
392

chapter, we describe and compare the current state and future prospects of DBS and CS in the fields of neurology and psychiatry.
CURRENT APPLICATIONS
1- Pain DBS: The rationale for the use of DBS in chronic pain dates back to the 1950s and, to date, more than 1,300 patients have been offered such a treatment. However, many factors limited its use: insufficient understanding of the neural networks that convey the perception of pain, uncertainties about optimal brain targets, inconsistent efficacy, and the success of spinal cord stimulation and opioid infusion pumps. The US Food and Drug Administration, based on the results of two multicenter trials, withdrew permission to use DBS as a treatment for pain (Burchiel 2001). The method, however, still remains in practice in Canada, Europe and Asia, but the number of reported cases ver the last decade is limited (Bartsch et al, 2008; Owen et al 2006a and b; Rasche et al 2006). Traditionally, the two most targeted areas for treatment of pain are: a) the sensory thalamus (Vc, ventrocaudalis) and b) the periaqueductal and periventricular gray (PAG and PVG). Most retrospective studies support the view that thalamic DBS has greater efficacy in neuropathic pain with 50-60% success rate (Levy et al 1987; Bendok and Levy, 1998); mesencephalic DBS has proved more efficacious in nociceptive pain (including osteoarthritis and cancer (Rezai and Lozano, 2002) and the mixed-pain failed back surgery syndrome (Young and Rinaldi, 1997; Kumar et al 1997) with 70% pain relief. However, DBS is poorly effective for central pain (Canavero and Bonicalzi 2007a,b), and contradictory results exist for phantom limb pain and postherpetic neuralgia (Owen et al 2006b; Rasche et al 2006). Recently, DBS of the posterior hypothalamic region has emerged as an effective treatment for otherwise medically-refractory cluster headache (CH) (Bartsch et al 2008; Owen et al 2007; Schoenen J et al 2005; Starr et al 2007), with half of the reported patients experiencing fair-to-excellent long-term relief in terms of frequency of pain attacks and severity of pain. The risk of intracerebral hemorrhage when bilaterally targeting such deeply-located structures remains a pervasive concern. CS: MI ECS was introduced for the management of pain following the recognition that DBS of the sensory thalamic nuclei had offered disappointing long-term results. Initially, MI ECS was proposed as a treatment for central thalamic pain (chapter 9) and facial neurogenic pain (trigeminal neuropathy) (chapter 10). MI ECS has proved most effective for trigeminal neuropathic pain with about 60% of the sufferers reporting a fair to excellent long-term pain relief (chapter 10). MI ECS is also effective in about half of all reported patients with CPSP (chapter 9). In phantom limb and brachial plexus avulsion pain, MI ECS offers an average success rate of slightly above 40% (chapter 10). Sensory cortex stimulation may also afford analgesia (chapter 9) and should be explored further. The application of either motor or sensory CS by implantable systems is moving along with that of noninvasive cortical stimulation such as TMS or tDCS (chapter 8). Despite modest effects, analgesia is usually transient, and even with protracted sessions, it lasts no more than 1 month, at best. However,
Cortical Stimulation Versus Deep Brain Stimulation
393
they may predict a positive outcome to ECS (Canavero and Bonicalzi, 2007a and b; chapter 8). The apparent superiority of tDCS over rTMS may pave the way to clinical application. CS vs DBS: Considerations, Limitations and Future Prospects The small number of patients treated in randomized placebo-controlled clinical trials, the heterogeneous nature of pain disorders, the uncertainty regarding the appropriate target for each specific pain syndrome, the lack of universal stimulation protocols and systematic follow-up data, and DBS-associated complications (diplopia, seizures, nausea, paresthesia, and, above all, intracerebral hemorrhage) are all factors that are likely to hinder an expansion of DBS in pain therapy for the foreseeable future and restrict it only to a few highlyspecialized centers (Cruccu et al 2007). Apparently, implantable CS devices offer the best alternative for the surgical treatment of intractable pain conditions. Apart from the aforementioned factors, other important criteria in selecting CS over DBS inlude: a) the profound laterality of painful symptoms, b) the extensive representation of the painful area in the lateral cortical surface, and c) advanced patients age. Bilateral pain is not a contraindication for ECS: bilateral effects from unilateral stimulation have been reported. Yet, bilateral surgery should be offered with a word of caution. Another problem arises when the pain afflicts a body region, such as the leg, which is represented in the medial interhemispheric cortical surface in several patients. In this case, only the much more invasive subdural CS would be offered. This problem may become more complex if leg pain is bilateral; in such cases, the electrodes should be inserted in both aspects of the interhemispheric fissure which may not be an easy procedure in many patients. Undoubtedly, it is much easier to stimulate the lateral aspect of the cortex unilaterally rather than the interhemispheric surface bilaterally. Fortunately, there is evidence that unilateral CS can have bilatateral beneficial effects (Canavero et al 2007b). Patients with trigeminal or arm neuropathic pains are especially suitable for ECS as compared to patients with bilateral leg pain. ECS appears to be a viable therapeutic option for drug-resistant central and trigeminal neuropathic pain. This is important, because neuropathic pain is a major concern for pain therapists due to its unresponsiveness to available drug and conservative treatments. Notably, the cut-off rate for success in pain relief is arbitrarily set at >50%; in the case of NP, however, pain improvement of 40% or even 30% has been sufficient to consider ECS effective. At this time, TMS or tDCS can be proposed as a preoperative test for selecting candidates for implantable CS procedures and possibly to alleviate painful conditions for a short period before a scheduled ECS procedure (chapter 8). A meta-analysis concluded that MI ECS appears to be more effective in alleviating chronic pain syndromes as compared to either MI rTMS or tDCS (Lima and Fregni, 2008). The availability of a noninvasive screening procedure for CS represents an important advantage of implantable (invasive) CS over DBS. Conversely, no drug-based or other test method has been developed that could predict the efficacy of DBS and, therefore, help us in selecting the right patients. DBS targets are small in size and deeply located in the brain and, at present, they can be investigated only by stereotactic approaches. CS is likely to expand its role in pain management. Patient selection will be optimized by developing drug dissection protocols in order to identify neurochemical signatures portending a successful outcome and selecting stimulation targets on the basis of anomalies
394
as seen in neuroimaging studies (Canavero and Bonicalzi 2007, chapter 9). For instance, SI cortex has been first targeted on the basis of theory and metabolic findings (Canavero and Bonicalzi 1995). The combination of anatomical, functional imaging, clinical, and intraoperative neurophysiological data (by keeping the patient awake during the procedure) will further improve target localization, together with confirmation by means of noninvasive stimulation. Securing the electrode in the predetermined position may reduce possible subsequent migration, with loss of effect. Software-guided stimulation protocols must be developed to facilitate postoperative titration and make the follow-up sessions less timeconsuming and more comfortable for both practitioners and patients; moreover, it will probably eliminate the need of intensive reprogramming or surgical repositioning of the electrode in case of tolerance-like phenomena or loss of ECS efficacy over time. On the basis of the available published studies and evidence from noninvasive brain stimulation, CS by implantable devices will likely remain at the forefront of pain therapies in the foreseeable future, particularly as the mechanisms of action and clinical efficacy will be addressed and evaluated in large-scale randomized controlled trials.
2- Parkinsons Disesae DBS: DBS has metamorphosed the treatment of PD by alleviating its troublesome symptoms and improving substantially the quality of life of sufferers. The reversibility, adjustability and programmability of DBS have favored it over the lesional procedures of the past. A change of rigidity is usually seen within 20-30 seconds with a maximum within 1 minute. Tremor can improve over weeks or months. Bradykinesia may respond after several hours or days and returns to baseline within up to 24 hours. The induction of dyskinesias with DBS in the short term predicts a favorable long-term outcome. Postoperatively, dopaminergic drugs may be reduced (about 50%) due to the additive effect; however, too rapid or too drastic a decrease in dopaminergic drugs carries the risk of unmasking apathy or depression or mania or psychosis or akinetic crisis. After optimization of parameters, levodopa may be gradually replaced with long-acting dopamine agonists. Postoperative dyskinesias or temporary worsening of PD due to drug reduction are possible. The thalamic ventral intermediate (Vim) nucleus was the first DBS target in the 1980s, with almost 90% of PD patients experiencing years-long control of their tremor (Benabid et al 1996). The minimal effect of VIM DBS on rigidity and bradykinesia spurred researchers to explore alternative nuclei as potential targets. Animal studies showed that the subthalamic nucleus (STN) is an integral relay node in the cortical-basal ganglia-thalamocortical pathways. The efficacy of STN DBS has now been documented by numerous groups and validated in prospective, randomized trials. Long-term follow-up and meta-analysis studies have demonstrated sustained improvement in the Unified PD Rating Scale (UPDRS) motor subscores of tremor, rigidity, and akinesia in the off-medication state, but minimal improvement in the onmedication state (Kleiner-Fisman et al 2006; Yu and Neimat, 2008). In prospective openlabel trials, STN DBS offered a significant benefit on the quality of life, particularly in patients younger than 65 years (Derost and al 2007; Deuschl et al 2006; Krack et al 2003). Bilateral STN DBS is considered the standard surgical therapy for medically-refractory PD;
395
unilateral STN DBS is efficacious in highly asymmetric parkinsonism (Kern and Kumar, 2007). However, concerns regarding continuous ipsilateral deterioration of PD symptoms over time restrict unilateral STN DBS only to highly selected patients. Based on reports of efficacy of pallidotomy for PD, globus pallidus internus (GPi) emerged as a target for DBS. High-frequency stimulation of the GPi alleviates rigidity, tremor and bradykinesia in the offmedication state, substantially decreases disabling levodopa-induced dyskinesias in the onand off-medication states and improves the quality of life (Deuschl et al 2006; Wichmann and DeLong, 2006; Yu and Neimat, 2008). In contrast to patients who received STN DBS, those with GPi DBS are not able to substantially reduce their medications (Rodriguez-Oroz et al 2005). Peduculopontine nucleus (PPN), a relay node in the network connecting GPi and STN with the basal ganglia and the peripheral nervous system, has recently been proposed as a new target for DBS in PD. Clinical studies support its efficacy, particularly, for postural instability and akinesia in parkinsonism (Nandi et al 2002a; Plaha and Gill, 2005). Similarly, DBS of the caudal zona incerta (ZI) has shown beneficial effects on parkinsonian tremor, rigidity and akinesia (Kitagawa et al 2005; Plaha et al 2005). CS: Functional neuroimaging studies have shown that cortical areas such as the primary motor cortex and the supplementary motor area are hypo- or hyperactive in both early and late stages of PD (chapter 12). Canavero and collaborators first reported on the clinical benefits obtained from unilateral extradural MI ECS in three PD patients (Canavero and Paolotti, 2000; Canavero et al 2002). UPDRS motor subscores (Part III) were decreased by 48% and the need for levodopa by 80%. Afterwards, the benefits of unilateral MI ECS in improving all parkinsonian symptoms (tremor, rigidity, bradykinesia, posture and gait disturbances, freezing) and reducing the required daily medications were confirmed in a large multicenter study (chapter 13 and 13b). The neurophysiologically documented prolonged modulatory activity of TMS on pathologically hyperactive or hypoactive areas of the human brain cortex has supported the assumption that TMS may also improve the motor components of PD (chapter 12). CS vs DBS: Considerations, Limitations and Future Prospects DBS has been acknowledged worldwide as the surgical treatment of choice for improving the incapacitating symptoms and the quality of life in PD patients: in 2002, it received FDAs approval as a treatment option for PD. Bilateral STN or GPi DBS is effective in patients with severe medication-refractory tremor-dominant PD. DBS of the Vim is no longer recommended for PD (Kumar et al, 1999; Kern and Kumar, 2007; Krack et al 2003). In spite of the absence of published blinded, randomized, controlled studies comparing the efficacy of DBS of the STN or GPi, the STN has gained general acceptance as the preferred target (Halpern et al, 2007; Kern and Kumar, 2007). The reasons for the supposed superiority of the STN over GPi DBS in PD include: easier targeting on MRI, longer stimulation effects of STN stimulation, larger reductions in medication, longer IPG battery life due to the lower intensity required for stimulating the smaller (as compared to GPi). Although longer followup data from randomized, controlled studies are needed to correlate the improvement induced by DBS with that due to medication, there is little doubt that DBS represents a breakthrough in the history of the treatment of PD. However, almost half of all PD sufferers are excluded from DBS mainly due to advanced age (>70 years), significant dementia or psychiatric
396
comorbidity, poor response to levodopa, low scores (<30-40) in the UPDRS part III or other general contraindication to DBS. Non-motor and non-dopaminergic symptoms of PD respond poorly to DBS, but newer targets may boost DBS potential. DBS has been associated with neuropsychiatric untoward effects (e.g. dementia, apathy, anxiety, suicidal attempts) and cognitive undesired sequelae (e.g. a decline in lexical fluency and executive functions: Temel et al, 2006). There also exists a considerable risk of surgery- and hardware-related complications estimated to be 1-3% and up to 25%, respectively (Breit et al 2004; Oh et al 2002). A hint that DBS may slow the progression of Parkinsons disease remains unproven and, if DBS is not applied early in the course of PD, such evidence will not become available (Simpson, 2006a). Due to the limited number of PD sufferers that have been treated by MI ECS so far, it is difficult to compare it to DBS. However, it seems that a considerable number of patients remain desperate and unrelieved. MI ECS appears to have some advantages over DBS in the treatment of PD: 1) it can alleviate all cardinal symptoms, but particularly the axial ones, 2) unilateral MI ECS induces beneficial effects bilaterally, 3) MI ECS is more suitable in elderly PD cases due to minimal surgery-related side effects and lower complication rates; moreover, the bilateral beneficial effects induced by unilateral MI ECS may delay the progressive deterioration of PD in the non-stimulated side, 4) MI ECS is more cost-effective because it does not require stereotactic equipment, and 5) MI ECS promises to potentially benefit almost half of the PD patients who have been excluded from DBS. Given that the number of PD sufferers aged >70 years will steadily increase, a considerable part of this patient population will require an alternative to DBS therapy; currently, CS appears the most likely option. A few caveats are in order: a) patients with levodopa-resistant parkinsonian symptoms are unlikely to respond to MI ECS, b) parkinsonism related to multisystem atrophy or progressive supranuclear palsy hardly benefits from MI ECS, and c) TMS-induced improvement of PD symptoms carries positive predictive value for MI ECS efficacy (Canavero 2007a). Nevertheless, a negative rTMS test does not predict failure of MI ECS (Cioni, et al 2007). Future efforts should focus on exploring the potential efficacy of stimulation of premotor area and the neuroprotective potential of MI ECS (Canavero and Bonicalzi, 2007a).
3- Tremor DBS: Tremor affects up to 2% of the general population worldwide, and can be essential (idiopathic) or associated with various conditions such as cerebellar dysfunction, multiple sclerosis, lesions of the brainstem (Holmes tremor) or head injury. Unilateral DBS has proved to be particularly effective in controlling contralateral idiopathic and secondary forms of tremor (Deuschl and Bain, 2002). Bilateral DBS has proved beneficial in head, trunk and voice tremor (Taha et al 1999). Traditionally, based on the existing experience of lesional surgery, the Vim thalamic nucleus has been targeted. Long-term follow-up data of unilateral Vim-DBS have documented the great improvement in essential tremor (ET) suppression, limb functionality, quality of life, and reduced need for medication (Kumar et al 2003; Sydow et al, 2003). In 8 patients with ET, unilateral DBS of the posteromedial subthalamic white
397
matter, i.e. zona incerta (ZI) suppressed contralateral proximal tremor by 81% for a mean follow-up of 22 months (Murata et al 2003). STN DBS has suppressed ET in patients with coexisting movement disorders, such as PD (Stover et al 2005). It is generally accepted that distal limb tremor responds well to Vim DBS, whereas proximal limb tremor is better controlled by ZI DBS (Nandi et al, 2002b). In a systematic review of 78 patients with tremor secondary to multiple sclerosis, unilateral Vim DBS offered postoperative benefits in motor function and daily activities in 88% and 76% of sufferers, respectively (Wishart et al 2003). Vim DBS has also successfully controlled tremor associated with unusual pathological conditions such as Holmes tremor (Romanelli et al 2003) or inherited cerebellar ataxia (Schramm et al 2005). CS: There are only few reports of successful CS application in non-parkinsonian tremor. MI ECS has shown the best efficacy in post-stroke distal rest and/or action tremor (Canavero and Bonicalzi, 2007a; Katayama et al 2002). In most cases, it reduced tremor significantly when it was offered to alleviate coexisting pain or movement disorders. Otherwise, the results of MI ECS in the management of ET have been disappointing (chapter 13). Similarly, noninvasive cortical stimulation has not provided clinical benefits in non-parkinsonian tremor (Hallett, 2007). CS vs DBS: Considerations, Limitations and Future Prospects Tremor has been part of the repertoire of stereotactic and functional surgery over the last fifty years. VIM DBS is the gold standard in the treatment of medically-refractory nonparkinsonian tremor with an efficacy up to 80% and long-lasting benefit for more than 7 years. MI ECS has not proved particularly effective in non-parkinsonian tremor, so far. This may be due to the pathophysiologic substrate of ET. Although recent reports support a cortical involvement (Raethjen et al 2007), the role of the cortex in the generation of ET has not been elucidated. On the other hand, there is sufficient demonstration of the key role of the somatosensory thalamus, and the Vim nucleus in particular, in the integration of neural loops within both the cortico-basal ganglia-spinal networks and, also, the cerebello-thalamocortical system. These networks are essential in the pathogenesis of tremor and, therefore, have offered convincing support that Vim is the optimal target for alleviating tremulous conditions, regardless of the underlying cause. Yet, given MI ECS efficacy in relieving parkinsonian tremor (although less so than akinesia and rigidity), it is possible that its limited efficacy in non-parkinsonian tremor may be due to wrong targeting (other areas rather than MI should be stimulated), suboptimal stimulation protocols or electrode-related shortcomings. The development of new electrode arrays with a larger number of poles supported by multiple programming IPGs may enhance its efficacy in the future.
4- Dystonia DBS: DBS is currently the most effective treatment for medically refractive dystonia. The main target used in primary dystonia (DYT1 gene positive) is the Globus Pallidus internus (GPi) and its efficacy has been shown in generalized dystonia, segmental dystonia, and complex cervical dystonia with movement scores typically improving by 75% (Kiss et al.
398
2007; Mueller et al, 2008; Krauss et al 1999, 2003). Certain types of truncal dystonia (camptocephalia and camptocormia) are emerging as good candidates for pallidal DBS (Sakas et al 2008b; Sakas et al 2007c). DBS maintains marked long-term symptomatic and functional improvement in the majority of patients with dystonia. (Loher et al. 2008). Upon activating DBS, phasic dystonic movements tend to improve early after surgery, but the response of tonic movements to chronic stimulation may be delayed (Krauss et al, 2002). Other targets, such as the thalamus and STN have shown encouraging results, but the available data is still scarce and sometimes contradictory (Sun et al. 2007). In sum, GPi DBS is an established treatment for idiopathic dystonia, either generalized or segmental, and its efficacy has been proven by extensive experience and by randomized, sham stimulationcontrolled trials. Future research will likely continue to address the most appropriate programming settings for various subpopulations of dystonia, the mechanism by which DBS affects this ailment, and the possibility of alternative brain targets that might have fewer associated side effects or greater efficacy than the GPi. CS: CS has been reported to be effective in certain forms of fixed dystonia (chapter 13). The reports are only a few, but preliminary data implies that the mechanism of fixed dystonia is more corticalized than in primary torsion dystonia. Some task-specific forms of dystonia, such as writers cramp or musicians cramp, may be related to maladaptive cortical plasticity as a result of overuse or improper use of the hand (Byl et al. 1996; Classen et al, 2003). Again, those focal forms of dystonia may be suitable candidates for CS. The optimal stimulation site remains to be elucidated. Based on results from rTMS on various forms of the disease, stimulation over the premotor cortex seems to yield better results than stimulation over the motor cortex (Murase et al, 2005). This may reflect both local and remote effects on distant, but connected sites (e.g. MI, the posterior parietal cortex and the basal ganglia). The assumption that stimulation over the premotor cortex may affect both the deep relay-nodes in the basal ganglia and the cortical sensory areas is in agreement with the known pathophysiology of dystonia, which involves the basal ganglia, MI and the somatosensory system in a diffuse sensory-motor processing dysfunction. Despite reports of invasive and non-invasive MI stimulation for fixed dystonia (chapter 13; Murase et al, 2005), it would be interesting to compare those effects with premotor cortex stimulation effects as well. Hence, even though MI ECS appears to have beneficial effects, it may be postulated that CS over the premotor cortex might be more effective. CS vs DBS: Considerations, Limitations and Future Prospects Bidirectional interconnectivity between motor cortical areas and deep-seated structures, such as the basal ganglia, could explain why dystonia and other movement disorders respond to neuromodulation at multiple sites. CS may induce distant neuromodulatory effects either at the cortical and/or subcortical levels through orthodromic or antidromic effects, similarly to what is postulated as a possible mechanism of action of DBS (Tisch et al, 2007). The proven efficacy of GPi DBS and, to a lesser extent, STN DBS in treating a wide spectrum of dystonic phenotypes suggests that DBS will remain the mainstay of therapeutic neuromodulation of dystonia in the near future. CS should however be better assessed by e.g. comparing unilateral versus bilateral stimulation of MI and non-primary motor areas in fixed rather than in torsion dystonia, and even combined with DBS. Initial experimental and
399
clinical data suggest that the potential efficacy of CS should be explored in cases of fixed or unilateral focal segmental dystonia involving one arm and neck or part of the trunk or leg, in cases of unilateral secondary dystonia of a multisegmental character, especially with a normal MRI scan, where DBS is likely to be less effective. rTMS may provide a useful predictor of clinical outcome and help triaging those patients who are likely to respond to CS.
5- Epilepsy DBS: 30% of epileptics are drug-resistant, due to either failure to respond to available antiepileptic drugs or medication-related adverse effects (Silanp and Schmidt, 2006; Kwan and Brodie, 2000). Resective brain surgery of the epileptic focus can be effective in judiciously selected patients; unfortunately, up to 40% of sufferers are not eligible for such surgical treatment. Moreover, post-operatively, anticonvulsant agents cannot be discontinued, seizure remissions may occur and undesired neuropsychiatric sequelae often lessen the surgical success rate. Several brain areas have been targeted by DBS (Velasco et al 2000a). The thalamus has been long considered a pivotal brain structure in the epileptogenic process and two thalamic nuclei have been targeted for suppressing intractable seizures: the centromedian nucleus of the thalamus (CMT) and the anterior nucleus of the thalamus (ANT). In addition, four other areas have been investigated: STN, cerebellum, caudate nucleus and hippocampus. The CMT has been considered a suitable DBS target on the assumption that it has an integrating role in the propagation of generalized seizures through its participation in an ascending brainstem-diencephalon-subcortical loop. CMT DBS has been successful in controlling mainly generalized tonic-clonic seizures and atypical absences (Velasco et al 1993a; Velasco et al 1993b; Velasco et al 1995; Velasco et al 2001) and, to a lesser degree, partial complex seizures (Fisher et al 1992). The role of CMT DBS, although promising, remains moot and should be further evaluated by large scale randomizedcontrolled studies (Andrade et al 2006; Fisher et al 1992). The AN, with its relatively small size, safe distance from vascular structures and documented anterograde projections to cortical, subcortical and limbic regions (Halpern et al 2008) has been targeted for DBS. Bilateral AN DBS reduced the frequency of partial complex seizures by approximately 50% in a small series and is being submitted to a controlled trial (Andrade et al 2006; Hodaie et al 2002; Kerrigan et al 2004; Halpern et al 2008). The STN has been also evaluated as a potential DBS target in drug-resistant epilepsy. In animal models, inhibition of the GABAergic neurons of the substantia nigra pars reticulata (SNr) reduces partial and generalized epileptic seizures (Iadarola and Gale 1982). Based on its excitatory effect on the nigral system, STN has been targeted in small, open-label prospective studies; notably, STN DBS reduced seizure frequency by 60-80% (Benabid et al 2001; Chabardes et al 2002; Handforth et al 2006), but large trials are needed to confirm these initial results. In the early 1970s, the cerebellum, due to its thalamic projections, was electrically stimulated for suppressing seizures (Cooper et al 1973) and Velasco et al (2005) reported a mean reduction of approximately 30% in motor seizures in five patients. At present, less than 20 patients have been treated by cerebellar stimulation in controlled studies and less than five experienced a significant reduction in their generalized tonic/clonic seizures. Chronic
400
electrical stimulation of the head of the caudate nucleus (HCN) is now being evaluated (Chkhenkeli and Chkhenkeli, 1997; Chkhenkeli et al 2004). The reported suppression of seizures has been attributed to the hyperpolarization of cortical neurons, induced by activation of the HCN by low-frequency stimulation (4-8 Hz). DBS of the aforementioned targets aims to modulate the abnormal cortical function by activating or inhibiting relay nodes which, although distant from the epileptogenic area, are critical epilepsy gating mechanisms. A different approach involves the direct stimulation of the epileptic zone per se. There is growing evidence that, in the years to come, stimulation of the amygdalohippocampal region will receive special attention. This technique may prove particularly helpful when the epileptogenic focus either involves eloquent brain areas or is not suitable for a resective procedure. The hippocampus, an integral component of Papezs circuit, plays a pivotal role in both the generation and propagation of temporal lobe seizures (Swanson, 1995; Oikawa et al 1994). Hippocampal stimulation provided excellent control of temporal lobe seizures in seven of 10 patients (Velasco et al 2000b). Vonck et al (2005) reported satisfactory suppression of seizures by bilateral amygdalohippocampal DBS; at 14month follow-up, one of seven patients was free of complex partial seizures, five enjoyed a 20-50% reduction in seizure frequency and only one did not improve. Undeniably, results are still unconvincing for this target. CS: Due to their intrinsic characteristics, CS and TMS are more likely to suppress seizures originating primarily in the cortex. Thus, if the epilepticogenic zone lies over the motor or speech area and a surgical excision is risky, CS could be an alternative method. In contrast, CS is unlikely to prove effective in cases with a deeply-located epileptogenic focus. The application of TMS to testing the mode of action and responsiveness to antiepileptic drugs (Macdonell et al 2002; Ziemann et al 1996) constituted the basis for the clinical evaluation of rTMS in epilepsy. Repetitive TMS provided promising, but unvalidated, results in a small number of patients suffering from complex partial seizures and temporal lobe epilepsy, focal cortical dysplasia or cortical myoclonus (chapter 18). Clearly, more work is needed. One issue with all commercially available DBS and CS devices is that the electrical stimulation is delivered in a pre-programmed manner with defined stimulation parameters (intensity, pulse width, frequency), mode of action (continuous versus cycling) and duration of stimulation, regardless of the subjects condition (so called open-loop stimulation). An important leap forward is the development of closed-loop stimulation systems. These devices have incorporated software with seizure detection algorithms and deliver stimulation to the epileptogenic focus on demand, i.e. once premonitory electroencephalographic (EEG) signs of imminent ictal activity are detected, well before clinical manifestation. CS combined with depth monitoring of epileptic activity is the first closed-loop system to have entered a controlled trial (chapter 19). Another controlled trial is also underway in US to test ANT stimulation in partial-onset drug-resistant epilepsy; the patient can activate the neurostimulator (InterceptTM, Medtronic, Minneapolis, US) when he/she feels the onset of seizure. CS vs DBS: Considerations, Limitations and Future Prospects Complex partial seizures, atypical absences and generalized tonic/clonic seizures seem to be responsive to DBS. Theoretically, DBS of the CMT or STN, which are relay nodes with
401
extensive cortical afferent and efferent projections, is likely to alleviate generalized seizures. In contrast, DBS of the ANT or hippocampus, which are functionally integrated into the limbic system, is expected to suppress partial seizures. Nevertheless, several critical issues should be addressed before either CS or DBS become effective therapeutic options for intractable epilepsy, such as a lack of homogeneity in the epileptic population. In clinical trials of CS or DBS, it is mandatory to include patients who have well defined anatomical, clinical, radiological and genetic characteristics, similar EEG activity in terms of intensity and frequency of spikes and also similar time elapsed since their last seizure. Standardized selection protocols based on proper phenotypical classification through video-EEG and neuroimaging studies should be developed, in order to help detect the appropriate candidates for CS or DBS, either as monotherapy or as an adjunct to other surgical treatments. At present, it remains unclear whether it is preferable to target the epileptogenic zone or the epileptogenic focus per se or whether subcortical epilepsy is best treated with DBS rather than CS. An important issue, therefore, is the definition of the optimal deep node or cortical zone to be targeted in each epileptic syndrome. H-coil TMS for depth (6 cm) stimulation may benefit some patients. Also, both high- (50 Hz) and low- (1 Hz) frequency CS of the epileptogenic zone can suppress epileptic activity and the impact of different stimulation parameters must be investigated (which applies to DBS too). In most series, patients continue their drugs during DBS therapy. It is important to evaluate the effect of DBS separately, but also compare it with that of the antiepileptic drugs. The synergistic or antagonistic actions of CS with antiepileptic drugs should also be studied. For instance, low-frequency TMS combined with valproate acid induces epileptic activity. Again, only well-conducted studies will answer these and other questions.
6- Psychiatric Disorders DBS: Early neuropsychiatric DBS practice has been largely guided by prior relevant lesioning experience (Sakas et al 2007b). Based on results of anterior capsulotomy and fMRI studies (Rauch et al 2006), DBS of the anterior limb of the internal capsule and the ventral striatum/nucleus accumbens (Nuttin et al 2003; Sturm et al 2003) has been brought to bear on treatment-refractory obsessive compulsive disorder (OCD) patients (30% of all cases: Picinelli et al 1995). In one study, four of eight patients treated by DBS enjoyed a greater than 35% reduction in the intensity of their symptoms and a substantial decrease in medication (Greenberg et al 2006). Apart from the IPG batterys short life expectancy and hypomania, DBS therapy has been well tolerated. DBS of the medial intralaminar thalamic (Visser-Vandewalle et al, 2003) and centromedian parafascicular nuclei (Houeto et al 2005), anterior limb of internal capsule (Flaherty et al 2005) and GPi (Ackermans et al 2006; Diederich et al 2005) has also shown high rates of efficacy (>70%) in suppressing drugrefractory motor and vocal tics in Tourettes syndrome, which are driven by dysfunctional limbic and cortico-basal gaglia-thalamo-cortical circuits. Patient selection guidelines (Mink et al, 2006) have been published. Severe behavioral changes may be induced in PD patients undergoing DBS (Temel et al 2006) due, most probably, to inadvertent stimulation of parts of the limbic system such as ZI or SNr; this data supports the assumption that DBS of relay
402
nodes of the limbic system may prove beneficial in treatment-resistant depression (TRD) (Fava 2003). At present, DBS of the inferior thalamic peduncle (Jimenez et al 2005), ventral internal capsule/nucleus accumbens (Schlaepfer et al 2008), and particularly the white matter in the subgenual cingulate gyrus (Cg25 area) (Mayberg et al 2005) have shown efficacy in depressed patients, with the latter target being now submitted to a controlled trial. Brodmans area 24a has also been proposed as a putative target for DBS in TRD (Sakas and Panourias, 2006). The superiority of Cg25 over the other targets may be due to its extensive connections to brain regions such as the brainstem, hypothalamus, insula, orbitofrontal, and cingulate cortex, which are all clearly implicated in the pathophysiology of depression (Marangell et al 2007). Despite all these challenging data, DBS remains highly experimental for intractable neuropsychiatric disorders. These efficacy data need validation and replication in large randomized controlled trials in order to spare it the fate of ablative psychosurgery in the past. CS: Noninvasive CS has been extensively tested and CS by implanted electrodes is now becoming reality. Over the last decade, the number of rTMS trials in psychiatric patients has steadily increased due to the high number of intractable or drug-intolerant psychiatric patients and the need to find alternatives to the rather brute-force approach of electroconvulsive therapy. The early promising results obtained from TMS of the dorsolateral prefrontal cortex (DLPFC) in intractable OCD were not replicated in other studies (chapter 17 and Sachdev et al 2007). Interestingly, in ten patients with comorbid OCD and TS, low-frequency rTMS of the supplementary motor area alleviated the symptoms and normalized overactive motor cortical regions (Mantovani et al 2006; see also Munchau et al 2002). Similar beneficial effect was shown in two other patients with TS and comorbid ADHD and MDD with an average improvement of 52%; this is highly comparable to that of approved behavioral or pharmacological treatments for TS (Mantovani et al 2007a). While low-frequency rTMS of the temporoparietal cortex has reduced hallucinations in schizophrenic patients (Aleman et al 2007), high frequency rTMS of the left prefrontal (or bilateral prefrontal) cortices has provided equivocal results (Prikryl et al 2007; chapter 17). DLPFC has been the most implicated area in the dysregulation of mood and the pathophysiology of depression. Antidepressant efficacy of high frequency rTMS of the left DLPFC has been repeatedly documented in randomized trials and meta-analyses (Fitzgerald, 2008; OReardon et al 2007). Low frequency TMS of the right DLPFC has also been effective in TRD when offered either as a single TMS treatment (Isenberg et al 2005) or combined with high-frequency rTMS of the left DLPFC (Fitzgerald et al 2006). Efficacy of rTMS has also been reported in other comorbid or depressive conditions such as after stroke (Jorge et al 2004), cerebrovascular disease-associated depression (Fabre et al 2004), PD (chapter 12) or panic disorder (Mantovani et al 2007b). tDCS of the left prefrontal region provided clinical benefits compared to sham stimulation in depressed patients (Boggio et al 2007) and studies are underway to investigate the possible efficacy of tDCS in enhancing cognitive functions such as working memory, decision making and mood (Been et al 2007). At this time, nonimplantable CS should be viewed as a treatment for the acute state rather than a realistic long-term therapy for TRD. However, TMS/tDCS studies opened the way to implantable (invasive) CS. Certainly, this latter cannot be offered as easily as TMS or tDCS. Yet, implantable CS may prove more efficacious, because the electrodes are positioned very closely to the precise cortical area which has been indicated by functional imaging studies as
403
the most appropriate to be stimulated. Given the high level of precision of implantable CS, we need to establish reliable criteria in order to select, among responders to TMS or tDCS, those who can expect the most benefit by implantable CS. CS vs DBS: Considerations, Limitations and Future Prospects In the new era of brain stimulation, the application of CS or DBS in psychiatric disorders has advanced less compared to movement disorders or pain. Undeniably, the unfortunate past of ablative psychosurgery has influenced negatively physicians attitude towards current brain stimulation efforts. DBS or CS experience in psychiatry is quite limited; therefore, definitive judgments on their clinical efficacy is impossible. As research moves forward, a few important issues must be raised. The clinical profile of neuropsychiatric disorders is diverse and symptoms of different psychiatric conditions, i.e. TS, OCD and MDD, may coexist making accurate diagnosis debatable. The establishment of consensus guidelines in selecting candidates suitable for DBS therapy is highly recommended and has already become feasible in OCD and TS (OCD-DBS Collaborative Group, 2002; Mink et al 2006). Additionally, keeping common evaluation rating scales during the patient selection process, plus preoperative and postoperative assessments ensures reliable comparative data from different centers and, hence, inference of strong conclusions. Reported TMS series include small cohorts of medically-refractory psychiatric patients at late stages of the disease. Most of the patients suffered from co-morbid conditions such as OCD and TS. In order to improve patient selection, the clinical benefits should be validated and replicated in patients with a clear psychiatric diagnosis who are enrolled in large-scale double-blind, shamcontrolled studies. Hopefully, correlative studies between non-invasive and invasive stimulation will make response prediction feasible for implantable CS, similar to what is seen in chronic pain (chapter 8). This will spur the development of the field in a way that is not apparently- possible for DBS. Compared to TMS or tDCS, implantable CS offers the opportunity to stimulate on a continuous or cyclical basis, without the patient having to return every month for therapy, and is better suited for sham-controlled studies. The development of the H-coil for TMS will make stimulation of e.g. the orbitofrontal cortex feasible, while advanced imaging may reveal stimulation patterns of CS or DBS that can be correlated with disease-related activity and clinical improvement. High-resolution imaging needs to be combined with electrophysiological mapping in order to accurately detect cortical targets or white matter bundles which may be critical in mood performance. Advancements in cortical mapping are expected to clarify the precise cortical area that is stimulated and possibly indicate other potential targets for CS therapy. Although DBS cannot stimulate a large area of brain tissue, yet, deep structures such as the STN or thalamic nuclei can be precisely delineated by current neuro-monitoring methods and perhaps are more suitable for further exploration by DBS (Visser-Vandewalle et al 2003). New relay-nodes for DBS have been discovered rather fortuitously, e.g. DBS of the ventral capsule/ventral striatum for OCD provided prominent antidepressant effect (Greenberg et al 2006). It should become a matter of priority, therefore, for investigation groups to provide detailed clinical and imaging information, as well as observations regarding target coordinates, lead location and clinical effectiveness; this will allow meaningful comparisons of targets and facilitate future developments. Short battery life is commonly reported in DBS studies for psychiatric disorders, implying either suboptimal targeting or that high-intensity stimulation is needed in
404
order to produce a clinical effect. While adjustments of stimulation parameters are largely guided by experience gained in movement disorders, efforts should be directed to identifying stimulation settings individualized for each psychiatric illness in order to maximize the clinical benefits and improve acceptance of DBS as a therapeutic tool from both practitioners and patients. As research moves forward in the field, reported results should be comparable through common sets of stimulation parameters. With respect to TMS, further studies will clarify whether alternative stimulation paradigms such as theta burst stimulation (chapter 4) or priming stimulation (Iyer et al 2003) will impact psychiatric disorders. Up to now, mostly seriously debilitated and drug-resistant psychiatric patients have been studied with TMS and DBS. In most series, efficacy of DBS is evaluated without discontinuing medications. This makes the blindness of studies doubtful and placebo effects possible. In the future, CS and DBS efficacy should be evaluated in the early stages of disease and in non-drug-resistant patients too and compared to patient cohorts treated only by medications. Since TMS may augment or hasten the clinical response of the patients to their medication (Fitzgerald 2008), it would be interesting to explore possible effects of stimulation therapies on hastening the expected progress of psychiatric symptoms over time. Although DBS has been clearly associated with psychiatric symptoms such as anxiety, fear (Okun et al 2007) and increased rates of suicide attempts (Burkhard et al 2004) in movement disorders patients, caution should be exercised when extrapolating the above observations into psychiatry-oriented DBS practice: DBS is a safe and viable therapeutic option and should be further explored in psychiatry. In sum, it should be clear for both practitioners and patients that invasive and noninvasive brain stimulation methods are still highly experimental and only much more work will establish them as powerful therapies for neuropsychiatric disorders. In this endeavor, it is critical to safeguard human rights and ethical rules as the future of the field is delineated. Therapeutic developments should evolve within a frame of evidence-based scientific knowledge, strict patient selection criteria and transparency in design and evaluation of clinical studies.
7- Stroke Rehabilitation Stroke is a debilitating disorder: roughly half of all stroke victims have some degree of residual motor disability substantially affecting their daily activities, self-independence and overall quality of life (Hendricks et al 2002). Functional recovery is mainly expected within a year after stroke; it may occur either spontaneously, possibly from partially or temporarily damaged cortical regions surrounding the stroke area, or by neurological reorganization of adjacent brain areas which take over the function of the irreversibly damaged areas. Physical therapy combined with peripheral nerve electrical stimulation is the only approved treatment for motor disability; unfortunately, many sufferers experience limited benefits at the end of the rehabilitation program. DBS has been used to ameliorate stroke-induced movement disorders such as tremor; to the best of our knowledge, however, it has not yet been used as a therapeutic tool or an addon treatment in stroke rehabilitation. DBS is exerted on well-defined deep relay-nodes; in
405
the post-stroke state, however, no important deep targets implicated in the regeneration process and suitable for therapeutic stimulation have been recognized yet. The disease per se extensively damages cortical and white matter pathways rather than specific brain nuclei; our limited understanding of the mechanisms of neuroplasticity may account for the investigators reluctance to offer DBS in stroke patients. CS has been introduced in stroke rehabilitation on the basis of its efficacy in modulating neuroplasticity and enhancing motor performance (Hummel and Cohen, 2005). Extradural MI ECS has been reported to improve motor recovery after stroke (chapter 15); however, the invasive character of the method, the uncertainty regarding potential MI ECS interferences with the natural recovery processes and the overall high cost remain considerable limitations for its wider use as an adjunctive neurorehabilitation treatment. Moreover, in a phase III study enrolling 164 patients with chronic stroke, MI ECS plus rehabilitation therapy did not provide improved motor status as compared to rehabilitation therapy alone (Levy et al 2008). Reasons can be adduced to explain such failure (chapter 15B). CS by implantable devices grew out of TMS(<1Hz/inhibitory and >10Hz/excitatory) and tDCS (anodal/excitatory and cathodal/inhibitory) studies (chapter 14). The concept of interhemispheric rivalry in the motor areas between the stroke-injured and the non-injured hemisphere has been a key hypothesis in reported series (Chapter 14 and Talleli and Rothwell 2006; but see Lotze et al 2006). Motor impairments are supposed to be elicited by reduced output of the injured hemisphere and/or increased inhibitory activity of the non-injured hemisphere over the injured hemisphere. Both the stroke-affected and non-affected hemispheres have been stimulated variably in an effort to improve recovery from post-stroke motor impairments. Both TMS and tDCS have been offered in single or multiple sessions in order either to enhance the function of the injured hemisphere or to inhibit the non-injured hemisphere; the functional improvement ranged between 10-20% (Chapter 14 and Talelli and Rothwell 2006). Notably, TMS-induced improvements were also noted in stroke victims suffering from aphasia or visuospatial neglect (chapter 14). CS vs DBS: Considerations, Limitations and Future Prospects DBS has been offered only in post-stroke tremor or hemiballism (Wichmann and Delong 2006) and thus the discussion can only be limited to CS. To date, TMS treatment has been offered only to small series of patients with subcortical lesions and rather mild motor impairment. To improve recovery by CS, it is important to define selection criteria to identify stroke patients most likely to benefit; several clinical parameters need to be defined including the somatic distribution and severity of the motor impairment and the type of brain lesion (cortical, subcortical or deep). It is mandatory to define, by means of randomized controlled studies, which type of intervention is more suitable for each type of stroke and each type of patient, taking into account the size and location of the lesion and the interval from stroke. The hand area has been the most targeted cortical region. The availability of tests evaluating hand performance, the accessibility of the hand cortical area, and the high incidence of residual hand motor deficits after stroke explain the above preference over other cortical regions. rTMS, however, has also provided moderate to good recovery of gait when the leg motor cortex was targeted (chapter 14). Clarification of whether stimulation of the ipsilesional, contralesional or both cerebral hemispheres provides the best recovery is
406
mandatory. After stroke, functional magnetic resonance imaging (fMRI) can effectively locate the hand motor region or other brain regions with residual function and discriminate between the affected and non-affected brain areas which could offer better functional recovery when stimulated. An interesting feature is the feasibility of combining neuroimaging studies with TMS-induced motor-evoked potentials, in order to determine the degree of functional recovery and guide the stimulation dose (Cramer 2008). TMS is offered within the first year after stroke, but the best interval to stimulate the motor cortex remains unknown. rTMS has been tested in patients after the first year, after the conventional rehabilitation had been completed and the motor impairments were considered permanent; improvements in motor performance of paretic extremities were reported even five years after stroke (Mally and Dinya 2008). TMS induced neuroregeneration in a Parkinsonian animal model (ArriasCarrion et al 2004); however, such a role in stroke remains speculative. It is important, therefore, to explore whether CS or DBS can influence processes such as synapse formation and outgrowth and, hence, could be used in neural regeneration and development. Future studies including larger series of patients with various degrees of severity of symptoms in both early and late stages of the disease are expected to clarify key interactions between CS efficacy and type of lesion, laterality of symptoms, time since stroke, stimulation algorithms, duration of stimulation, and stimulation combined with other forms of neurorehabilition. One recent development might improve the prospects of this therapy, according to Canavero (personal communication). A recently developed neurochip (Jackson et al 2006) creates an artificial connection between two sites in MI by using action potentials recorded on one electrode to trigger electrical stimuli delivered to another, in a chronically implanted array, over the long-term. Once configured, it operates autonomously. It can induce changes mediated by hebbian mechanisms by delivering stimuli within 50 ms (20 ms best) of recorded spikes. The linkage is explained by potentiation of horizontal pathways within MI and in cases of partial injury this could strengthen surviving projections between sites connected by the prosthesis. At present, CS for neurorehabilitation should be considered as a highly experimental, but potentially effective method that holds great promise for functional improvement of otherwise treatment-refractory stroke sufferers. CS has the potential to address the multifaceted clinical profile of stroke that includes motor, sensory, cognitive and psychiatric impairments, and favor the rehabilitation outcome of various components of stroke.
LIMITS
Compared to other neurosurgical procedures, both CS and DBS, are minimally invasive and associated with low complication rates. The stimulation can be programmed to meet patients needs or discontinued in case of undesirable effects or ineffectiveness. Yet, there is neither agreement on which targets should be stimulated in order to obtain maximum clinical benefit for each neurological or psychiatric disorder, nor sufficiently standardized stimulation protocols. These procedures have further limitations.
Cortical Stimulation Versus Deep Brain Stimulation 1- Resources and Personnel
407
Both DBS and CS require expensive high-technology equipment including frame-based stereotaxis or frameless neuronavigation, respectively, and multidisciplinary teams consisting of neurosurgeons, neurophysiologists, movement disorder neurologists, pain therapists, neuropsychologists, experienced technicians and specially-trained nurses. One key difference is that DBS for movement disorders requires the close collaboration of a neurologistneurophysiologist during surgery, unless it is performed without microrecordings monitoring and intraoperative clinical testings. Conversely, in CS, the intraoperative collaboration of a neurologist is not usually necessary, unless meticulous awake mapping is considered essential for the success of the procedure. Both TMS and tDCS are dependent on hightechnology equipment, advanced neuroimaging studies for identifying optimal targets, and experienced teams: although they may evolve into powerful predicting tools of efficacy of implantable CS, their therapeutic potential has not been elucidated yet.
2- Preoperative Evaluation The application of DBS is restricted by certain technical limitations: a) current imaging technology is still subject to spatial and geometrical distortions and b) the reproducibility and uniformity among different imaging systems and devices varies considerably. This makes high accuracy problematic and the comparison of data from different groups very difficult. In CS, an important limitation is the difficulty in the precise identification of the cortical areas in the single individual, given the huge inter-subject variability. Various techniques have been used to ensure accurate cortical targeting including fMRI, TMS and PET fused with neuronavigation data and combined with intraoperative neurophysiological recordings (sensory and motor evoked potentials, bipolar extradural stimulation). FMRI, though, has limitations: 1) cerebral areas are variably activated depending on task performance and individual thresholding, 2) large draining veins, motion artifacts, and echo planar images compromise accuracy, 3) activation protocols are not yet standardized, 4) cortical activity is hard to detect within the fissures, and 5) often, the examination must be repeated due to lack of patients cooperation. Intraoperative neurophysiological mapping too requires local anesthesia and a larger craniotomy and may not be well tolerated by chronic pain sufferers or elderly patients.
3- Electrode-tissue Interface Deeply implanted DBS electrodes are in contact with the brain, i.e. there is a direct interface of the electrode with the neural tissue, which is not the case for extradural CS; however, alterations in the impedance due to the thickness of the underlying dura or cerebrospinal fluid (CSF) can affect the stimulation results (Manola and Holsheimer, 2007). Subdural CS, like DBS, is interfaced with neural tissue, but given its low overall efficacy and association with such complications as cortical lesions, bleeding, CSF leakage, seizures,
408
secondary gliosis and scar tissue formation, it is rarely used. However, the interdural placement of the electrode (Sakas et al 2008a) in a patch-like fashion might prove an effective alternative option to standard extradural and subdural cortical stimulation techniques.
4- Electrode Design and Stimulation Delivery The stimulation of more extensive cortical zones by implanting multiple electrode arrays may improve clinical results. An important issue is whether the electrodes with a single array of stimulating poles should be replaced by electrodes with a double- or triple-array of stimulating poles. The three-array electrode (Lamitrode, ANS, Plano, Texas, US) may improve efficacy (figure 1), as it is supposed to deliver a better targeted wedge-shaped stimulation which, when necessary, can be offered in an alternating sweeping mode. It is also important to establish whether it is preferable to perform bipolar versus monopolar stimulation and what are the most effective electrical stimulation algorithms. The stimulation efficacy could also be affected by other parameters, such as the position of the electrode (perpendicular or parallel to the central sulcus) and the lack of precision in positioning or securing the electrode over the cortex.
5- Patients Clinical Condition DBS and CS call for mentally and physically able patients to undergo the demanding preoperative clinical assessments and to attend the postoperative follow-up sessions during the period of titration of the stimulation parameters. However, a considerable percentage of sufferers are excluded from DBS due to advanced age (usually more than 70 years), brain atrophy, cognitive impairment, psychiatric symptoms and medical co-morbidities. In DBS, there is a small (1-3%), but considerable risk of hardware- or surgery-related complications (intracerebral hemorrhage, seizures, infection), plus stimulation-related neuropsychiatric effects, such as apathy, hallucinations, cognitive dysfunction, decline in executive functions, depression, and even suicide attempts. CS is better tolerated because it is unilateral, does not involve penetration of brain tissue with all its associated risks (hematoma, contusion, parenchymal infection) and the risk of intracerebral hemorrhage is virtually non-existent; the risk of seizures is related to the stimulation parameters and usually controlled with changes in programming. In DBS, however, the risk of seizures may be related to the intraparenchymal presence of the electrode per se and, therefore, not easily corrected by better programming only.
409
Figure 1. The three-array electrode (Lamitrode, ANS, Plano, Texas, US) offers the option of delivering wedge-shaped stimulation in an alternating sweeping mode. Such a stimulation type is expected to achieve a better targeted stimulation focus and maximize clinical benefits.
6- Device Features Given the widespread appreciation of its unfulfilled potential, improvements and refinements in CS electrodes should be anticipated, whereas the more established and widespread use of the electrodes creates a lesser industrial incentive for development and refinement of the current quadripolar systems for DBS. Nevertheless, a higher number of poles would allow more refined stimulation. All current implantable devices for brain stimulation are susceptible to and interfere with magnetic environmental fields and they have not been cleared for safety in high-field (3-Tesla) MRI scanners. This is a problem that must be addressed by the device industry. As concerns the noninvasive technology, major shortcomings are the cost, the lack of portability of most machines that makes its routine use not easy, the patients discomfort for certain stimulation algorithms and the difficulty to conduct blinded studies. tDCS has promise over tTMS, because the required equipment is less expensive, the technique is easier in application and maintenance, the stimulators can be battery-driven and are portable, the modulatory effect appears to last longer, and shamstimulation is easily implemented as patients are unable to detect stimulation.
In conclusion: a- Only well-structured and controlled clinical trials of rTMS and tDCS will verify their therapeutic potential and justify the enthusiasm generated by their introduction in clinical practice. t present, noninvasive brain stimulation methods have a strong potential for predicting a positive response to implantable CS and guide patient selection, timing of intervention and type of stimulation. b- CS by implantable devices appears to be a promising and effective treatment modality for several neurological conditions. However, the clinical effectiveness of CS may wane over time and the long-term benefits may become less pronounced. There is a large inter-individual variability in terms of response to stimulation (in some cases leading to excitation, in others to inhibition, but a similar final result, e.g. analgesia). Moreover, some individuals on fMRI show pure SI activation during a motor task and in such patients SI must be engaged, not MI. This inconsistency of the effect and unpredictability of the outcome makes surgeons and neurologists hesitant to propose the procedure to patients. It is important to define -and possibly
410
Damianos E. Sakas and Ioannis G. Panourias standardize- optimal cortical mapping, ideal targeting and protocols of programming in order for CS to become a first line treatment for neuropsychiatric disorders. Although DBS plays a starring role in the management of PD and other neurological disorders, implantable CS presents certain significant advantages, such as a lack of requirement for a stereotactic frame, close to nil risk of intracerebral hemorrhage, less invasiveness, better acceptability to both patients and neurologists, and the availability of a noninvasive preoperative predictive test.
FUTURE PROSPECTS
An exciting frontier is the development of the capability to artificially relay environmental sensory information by CS implants, i.e. relay of hearing or visual information, directly into the human brain. From a strict anatomical and surgical safety perspective, the cerebral cortex is likely to be a much more attractive implantation site compared to the brainstem, basal ganglia or other deep brain sites. However, one of the most difficult aspects of this work is the encoding of environmental sound or vision into parameters of electrical stimulation applicable to the cortex in order to convey the information. Research has shown that patients with profound auditory loss can discriminate between electrical stimuli based on the differences in the parameters of the stimulation current with the level of the electric current, being correlated to sound loudness and the frequency of the electric current to sound pitch. Following this discovery and taking advantage of the humans discriminatory ability, it became possible to design and build effective speech processors; these devices receive input from an external microphone and then electrically encode this acoustic information in a manner specifically designed to exploit the patients ability to perceive differences in certain electrical stimulus parameters. On this basis, a cortical auditory neural prosthesis becomes possible (Howard et al 2000). However, one of the most ambitious goals is the transmission of visual perception by cortical stimulation electrodes into the visual cortex of blind volunteers, i.e. a cortical visual prosthesis (chapter 22). It is encouraging that the cerebral cortex seems to have the ability to adapt and to interpret in efficient manner electrical information that is applied in the right sequences and within a proper range of stimulation parameters by a large cortical-intracortical array. A relevant and important finding is that deaf patients do not appear to sustain deafferentation changes that would preclude the reactivation of normal auditory processing by the prosthetic stimulation device (Howard et al 2000) and a similar phenomenon could be expected to occur in the visual cortex and pathways of blind patients (chapter 22). A significant discovery that resulted from this research is that penetrating electrodes into the cerebral cortex are greatly superior in delivering more precisely the electrical stimulation compared to those electrodes that are placed in contact with the cortical surface. This opens the possibility of another field of therapeutic cortical stimulation namely deep cortical or subcortical stimulation (via an array of penetrating recording and stimulating electrodes), to be distinguished from both surface cortical stimulation (extradural or subdural) and DBS.
411
Computational neuromodulation is a field of computational biology dedicated to the study of the biophysical and mathematical characteristics of the electrochemical modulation in the nervous system. It is widely acknowledged that all types of networks and neurons (motor, sensory, and interneurons) are subject to modulation. Modulation may be induced in a circuit either extrinsically or intrinsically by circuit neurons or neural fibers (Fellous and Linster, 1998) or by artifical (manufactured) projecting systems or interfaces, including CS and DBS, with the aim to mimic natural neuromodulation. Given the complexity, convergence and divergence of neuromodulation in nature, the computational approach may provide a deeper understanding and the foundation for refined clinical applications. Areas amenable to computational analysis include synaptic drive, synaptic efficacy, and sensory encoding (Hille, 2001). With such approaches, altering the intrinsic properties of neurons, change their time-course, voltage-dependence and synaptic conductance and the strength of synaptic connections may become feasible. A single membrane current could be influenced and, depending on the conductance of the neuron membrane, a neuron brought across the boundaries of different behavior (Hille, 2001; Marder and Thirumalai, 2002). Furthermore, on the basis of current progress in computational biology, it is important to explore whether CS or DBS could be used in order: a) to offer a short synaptic input that can jump start a circuit, b) to influence the encoding of sensory information in spike trains, or c) to reconfigure an anatomically defined network into a different functional circuit by altering the intrinsic properties or the synaptic strength of the neurons within the network. Extrinsic modulation in particular (either natural or by implanted CS or DBS devices), could potentially be used to tune and configure whole networks and organize ensembles of circuits (Katz, 1995), in order to bias extensive networks or neurological systems into different functional outputs, in much the same way as changing parameters in a network model should bias or modify the output of the network (Marder and Tiramulai, 2002). Computer models of the bioelectrical and statistical aspects of neural recording and stimulation-induced recruitment can be used to model, and decipher neural coding. For instance, hippocampal functions can be replicated with a microchip implementation of the predictive mathematical models. Advances in this field hold the promise that individuals with damaged areas of the brain could be helped by CS based on such computational methods and that the use of neural interfaces could enhance normal or impaired neural function (Berger et al 2005). CS and DBS may be further refined by developing electrodes that monitor neuronal activity from multiple regions, generate network level representations of the brain and simultaneously stimulate, i.e. closed-loop integrated monitoring and stimulating implantable systems. In these systems, the stimulation is activated on demand by intrinsic signals such as specific brain activity. Hence, these devices are capable of responsive neurostimulation. A closed loop CS or DBS could be much more effective when it is activated by, i.e. it is responsive to, control signals derived from brain or body signals. Such technologies would be useful not only for epilepsy (chapter 19), but also for e.g. Parkinsons disease; research has shown that intended and self-timed movements are preceded by increased activity in the parietal cortex and sensorimotor putamen (Pancrazio et al 2006). Patients will have an advanced sensor either in contact or embedded in muscles and stimulation will be triggered by EMG signals elicited by the patients movement. The possibility to offer effective intermittent stimulation would provide extended battery life.
412
Responsive direct brain stimulation carries certain advantages over other methods as treatment is provided as needed, when needed, and for the time needed. The undesired effects of chronic stimulation are reduced, habituation is minimized, and, in epilepsy, the time of stimulation is the least necessary to suppress epileptiform activity; hence, the chances of potential failure due to reorganization of the epileptogenic circuitry are minimized. Temporal and spatial specificity, however, appear to play a pivotal role in the success of responsive stimulation. It is a high priority to develop sophisticated software that will allow early detection of epileptiform activity and enhance accuracy in lead or array implantation over the epileptogenic focus or zone. It is also important to develop seizure-prediction algorithms for each sufferer and stimulation devices able to drive several multicontact leads and arrays, and offer numerous combinations of cortical, subcortical or deep brain stimulation. Finally, all this technology should be incorporated into implantable, cosmetically acceptable devices in the most clinically effective fashion. Numerous exciting developments in neuroengineering and neurotechnology are likely to influence the future applications of both CS and DBS. Future telemetric IPGs are expected to be smaller in size, more cosmetic, contour-shaped to apply to the cranium convexity or miniaturized to be implanted through a burr hole or small craniotomy. Batteries will be based on advanced chemistry offering more power in smaller size, will be rechargeable with extended life, and capable to be connected to the electrode(s) without extension wires. Such advancements may prove particular helpful in cases of cervical dystonia where hardwarerelated complications are commonly encountered. A great advance will be the incorporation into DBS electrodes of microactuators which will enable precise implantation after the initial insertion. This will reduce substantially the operating time, since fine adjustments for maximizing the clinical effect or minimizing untoward sequelae will be easier to carry out, at a later stage, rather than in the operating theatre. The construction of smart stimulators with the capability for dynamic internal adjustments will refine CS and DBS practice (Pancrazio et al 2006). Sophisticated software with advanced graphics for precise electrode insertion and for programming the stimulation in both CS and DBS procedures is expected to optimize pole selection after permanent implantation, individualize stimulation parameters and maximize clinical efficacy. Notably, such advanced software programs that combine MRI, CT, three-dimensional (3-D) brain atlases and neurophysiological microelectrode recording data with the volume of brain tissue activated by DBS have already been tested in preliminary studies (Miocinovic et al 2007; Butson et al 2007). Finally, the ability to program stimulation devices through remote access, telephone lines or via the Internet may render patients independent from the reference center and lessen the number of required follow-up visits. Sensitive, highly specific sensors which can be applied on the skin and detect physiological biosignals (e.g. changes in the heart rate, skin temperature, respiration rate, eyelid closure, EMG) are in development, and so are implantable chronic recording microelectrode arrays incorporating on board amplification, spike detection and wireless transmission of data and power (Pancrazio et al 2006). Finally, brain computer interfaces (BCI), defined as an electronic brain implant that translates the intention to either communicate or move into communication or movement through computer cursors, robotic devices or into actual movement of paralyzed limbs, represent the next opportunity to integrate CS and DBS into a larger clinical perspective. A
413
BCI can detect changes in the users brain activity and convert them into commands for a computer application. This is achieved through the application of signal processing techniques to the signals that the patient is still able to control. The key element in a BCI is a decoding algorithm that converts the main electrophysiological signal into an output that is suitable to control an external device. The interfaces rely on the natural adaptive ability of the human brain. The users have to learn to adapt their biological response, i.e. change the amplitude or frequency of the signal monitored. BCI technology will open new possibilities for severely disabled humans and offer systems that will deliver communication or action based on data derived from incorporated cortical monitoring and stimulation systems (Angelakis et al 2007; Warwick et al 2006). In this regard, the development of miniaturized, multi-functional, chronic neural implants (BioMicroElectro-Mechanical Systems or Bio MEMS) aims to obtain and control signals extracted from neuronal activity in order to enhance residual capabilities or perform actions such as moving a prosthetic arm with near natural performance. Microelectrode arrays are special types of micro-hardware constructed by using microfabrication and microelectronics by thin film-based planar and 3D-arrays of substrate microelectrodes. Hybrid neural interfaces or cultured neural probes can be coupled in vitro to populations of cultured neurons (Sanguineti et al 2003), where each electrode is covered and surrounded by a locally confined network of cultured neurons, obtained by chemical patterning of the substrate. These microsystems are designed to make connections and communicate with regenerating neurons and can be intended as hybrid neural information prosthetic transducers for stimulation and/or recording of neural activity in the brain or the spinal cord (Rutten et al 2007). Any neurological disorder that has altered electrical conductivity could potentially be helped by nanomaterials (Webster et al 2004), which interact much more closely with cells than currently available materials. Carbon nanofibers can act as minimally traumatic CNS electrodes and carbon nanofibers arrays are expected to increase the accuracy of cerebral electrical stimulation. With such technology, our ability to offer complex, but precise patterns of stimulation may be enhanced; it will become possible to perform not only electrical microrecording, but also electrochemical recording and stimulation (Li and Andrews, 2007). Neuromimetic or neuromorphic engineering, in which very large scale integration (VLSI) systems containing electronic analog circuits mimic the architecture and design of biological nervous systems in order to replicate their function, e.g. vision or hearing (Silver et al 2007; Smith and Hamilton, 1999) may one day be integrated into CS or DBS.
CONCLUSION
Currently, the main indication of CS is central and trigeminal neuropathic pains. In the coming years, PD, dystonia, epilepsy, and psychiatric disorders are likely to become established indications for CS as well. Undoubtedly, CS will also play a starring role in neurorehabilitation, facing the great challenge of inducing movement in paretic extremities and releasing medically-refractory spasticity even years after stroke. The future of CS will be influenced greatly by increased awareness and understanding of existing indications and applications, improved case selection, introduction of new indications, more mature
414
assessment of outcome and convincing evidence regarding efficacy (Simpson 2006b). A deeper understanding of mechanisms of action will allow us to answer critical questions regarding type of stimulation (extradural, subdural or interdural), mode of stimulation (anodal or cathodal), specific sets of effective stimulation algorithms for each neurological or psychiatric condition, and prediction of degree and duration of expected clinical benefit. Certainly, the ability of unilateral CS to influence brain activity bilaterally, be it pain, PD or stroke, is a major advantage over DBS. Great advancements are likely to take place in the field of responsive CS, which will particularly impact young patients suffering from progressive disabling disease, such as epilepsy or dystonia. CS also holds the potential to enhance cognitive performances (Topper et al 1998, Olivieri et al 2001, Boroojerdi et al 2001, Montes et al 2002) and should be explored for dementia (Canavero S, personal communication). CS is a high-technology dependent field that will be largely influenced by forthcoming neuroengineering developments and a shift away from the current dependence on pharmacological treatment. A big challenge will be to integrate CS technology with relevant advancements in human neurobiology, neuroplasticity, and neural repair, as well as to explore the potential neuroprotective effects of CS. Current implantable CS devices may need to be redesigned in order to incorporate complex and intelligent miniaturized systems and sophisticated software which will enable them to integrate information exchange between the device and the patients brain, while remaining cost-effective. Progress in microsystems technologies, microelectronics, nanotechnologies, and computer modelling software are likely to create new opportunities for developing advanced stimulation systems and for expanding the therapeutic potential of CS.
REFERENCES
Ackermans L Temel Y, Cath D, et al. Deep brain stimulation in Tourettes syndrome: two targets? Mov Disord 2006;21:709-13. Aleman A, Sommer IE, Kahn RS. Efficacy of slow repetitive transcranial magnetic stimulation in the treatment of resistant auditory hallucinations in schizophrenia: a metaanalysis. J Clin Psychiatr 2007;68:416-21. Andrade DM, Zumsteg D, Hamani C, et al. Long-term follow-up of patients with thalamic deep brain stimulation for epilepsy. Neurology 2006;66:1571-3. Angelakis E, Hatzis A, Panourias IG, Sakas DE. Brain computer interfaces: a reciprocal selfregulated neuromodulation. Acta Neurochir Suppl 2007;97(2):555-9. Arrias-Carrion O, Verdugo-Diaz L, Feria-Velasco A. Neurogenesis in the subventricular zone following transcranial magnetic stimulation and nigrostriatal lesions. J Neurosci Res 2004;78:16-28. Bartsch T, Pinsker MO, Rasche D, et al. Hypothalamic deep brain stimulation for cluster headache: experience from a new multicase series. Cephalalgia 2008;28:285-95. Been G, Ngo TT, Miller SM, Fitzgerald PB. The use of t DCS and CVS as methods of noninvasive brain stimulation. Brain Res Rev 2007;56:346-61.
415
Benabid AL, Koudsie A, Benazzouz A, et al. Deep brain stimulation of the corpus luysi (subthalamic nucleus) and other targets in Parkinsons disease: extension to new indications such as dystonia and epilepsy. J Neurol 2001;248(Suppl 3):III37-47. Benabid A-L, Pollak P, Gao D, et al. Chronic electrical stimulation of the ventralis intermedius nucleus of the thalamus as a treatment of movement disorders J Neurosurg 1996;84:203-14. Bendok B, Levy RM. Brain stimulation for persistent pain management. In: Gildenberg PL, Tasker RR, editors. Textbook of stereotactic and functional neurosurgery. New York: McGraw Hill; 1998. p.1539. Berger TW, Ahuja A, Courellis SH, Deadwyler SA, Erinjippurath G, Gerhardt GA et al. Restoring lost cognitive function. IEEE Engl Med Biol Mag 2005;24:30-44. Boggio PS, Rigonatti SP, Ribeiro RB, et al. A randomized, double-blind clinical trial on the efficacy of cortical direct current stimulation for the treatment of major depression. Int J Neuropsychopharmacol 2007;11:173-83. Boroojerdi B, Phipps M, Kopylev L, Wharton CM, Cohen LG, Grafman J. enhancing analogic reasoning with rTMS over the left prefrontal cortex. Neurology 2001; 56: 526528 Breit S, Schultz JB, Benabid A-L. Deep brain stimulation. Cell Tissue Res 2004;318:275-8. Burchiel K. Deep brain stimulation for chronic pain: the results of two multi-center trials and a structured review. Pain Med 2001;2:177. Burkhard PR, Vingerhoets FJ, Berney A, et al. Suicide after successful deep brain stimulation for movement disorders. Neurology 2004;63:2170-2. Butson CR, Noecker AM, Maks CB, McIntyre CC. StimExplorer: deep brain stimulation parameter selection software system. Acta Neurochir Suppl 2007;97(2):569-74. Byl N, Wilson F, Merzenich M, Melnick M, Scott P, Oakes A, McKenzie A. Sensory dysfunction associated with repetitive strain injuries of tendinitis and focal hand dystonia: a comparative study. J Orthop Sports Phys Ther 1996;23:234-44. Canavero S, Bonicalzi V. Cortical stimulation for central pain. J Neurosurg 1995;83:1117. Canavero S and Paolotti R. Extradural motor cortex stimulation for advanced Parkinsons disease: case report. Mov Disord 2000;15:169-71. Canavero S, Paolotti R, Bonicalzi V, et al. Extradural motor cortex stimulation for advanced Parkinsons disease: report of two cases. J Neurosurg 2002;97:1208-11. Canavero S, Bonicalzi V. Extradural cortical stimulation for movement disorders. Acta Neurochir Suppl 2007a;97(2):223-32. Canavero S, Bonicalzi V. Extradural cortical stimulation for central pain. Acta Neurochir Suppl 2007b;97(2):27-36 Canavero S, Bonicalzi V. Central pain sindrome. New York: Cambridge University Press, 2007 Canavero S, Bonicalzi V. Central pain sindrome: elucidation of genesis and treatment. Exp Rev Neurother 2007; 7: 1485-1497 Chabadres S, Kahane P, Minotti L, Koudsie A, Hirsch E, Benabid A-L. Deep brain stimulation in epilepsy with particular reference to the subthalamic nucleus. Epileptic Disord 2002;4 (Suppl 3): S83-S93.
416
Chkhenkeli SA, Chkhenkeli IS. Effects of therapeutic stimulation of nucleus caudatus on epileptic electrical activity of brain in patients with intractable epilepsy. Stereotact Funct Neurosurg 1997;69:221-4. Chkhenkeli SA, Sramka M, Lortkipanidze GS, et al. Electrophysiological effects and clinical results of direct brain stimulation for intractable epilepsy. Clin Neurol Neurosurg 2004;106:318-29. Cioni B, Meglio M, Perotti V, De Bonis P, Montano N. Neurophysiological aspects of chronic motor cortex stimulation. Clin Neurophysiol 2007;37:441-7. Classen J. Focal hand dystonia-a disorder of neuroplasticity? Brain 2003;126:2571-2. Cooper IS, Amin I, Gilman S. The effect of chronic cerebellar stimulation upon epilepsy in man. Trans Am Neurol Assoc 1973;98:192-6. Cramer SC. Repairing the human brain after stroke. II. Restorative therapies. Ann Neurol 2008;63:549-60. Cruccu G, Aziz TZ, Garcia-Larrea L, et al. EFNS guidelines on neurostimulation therapy for neuropathic pain. Eur J Neurol 2007;14:952-70. Derost PP, Ouchchane L, Morand D, et al. Is DBS-STN appropriate to treat severe Parkinsons disease in an elderly population? Neurology 2007;68:267-271. Deuschl G, Bain P. Deep brain stimulation for tremor [correction of trauma]: patient selection and evaluation. Mov Disord 2002;17(Suppl 3):S102-11. Deuschl G, Schade-Brittinger C, Krack P, et al. A randomized trail of deep-brain stimulation for Parkinsons disease. N Engl J Med 2006;355:896-908. Diederich NJ, Kalteis K, Stamenkovic M, et al. Efficient internal pallidal stimulation in Gilles de la Tourette syndrome: a case report. Mov Disord 2005;20:1496-9. Fabre I, Galinowski A, Oppenheim C, et al. Antidepressant efficacy and cognitive effects of repetitive transcranial magnetic stimulation in vascular depression: an open trial. Int J Ger Psychiatr 2004;19:833-42. Fava M. Diagnosis and definition of treatment-resistant depression. Biol Psychiatry 2003;53:649-59. Fellous J-M, Linster C. Computational models of neuromodulation. Neural Comput 1998;10:771-85. Fisher RS, Uematsu S, Krauss GL, et al. Placebo-controlled pilot study of centromedian thalamic stimulation in treatment of intractable seizures. Epilepsy Res Suppl 1992;5:21729. Fitzgerald PB, Benitez J, Brown T, et al. A double-blind sham controlled trial of repetitive transcranial magnetic stimulation in the treatment of refractory auditory hallucinations. J Clin Psychopharmacol 2005;25:358-62. Fitzgerald PB, Benitez J, de Castella A, Daskalakis ZJ, Brown TL, Kulkarni J. A randomized, controlled trial of sequential bilateral repetitive transcranial magnetic stimulation for treatment-resistant depression. Am J Psychiatr 2006;163:88-94. Fitzgerald P. Brain stimulation techniques for the treatment of depression and other psychiatric disorders. Australasian Psychiatr 2008;16:183-90. Flaherty AW, Willimas ZM, Amimovin R, et al. Deep brain stimulation of the anterior limb of internal capsule for the treatment of Tourettes syndrome: technical case report. Neurosurgery 2005;57(Suppl 4):E403.
417
Greenberg BD, Malone DA, Friehs GM, et al. Three-year outcomes in deep brain stimulation for highly resistant obsessive-compulsive disorder. Neuropsychopharmacol 2006;31:2384-2393. Hallett M. Transcranial Magnetic Stimulation: a primer. Neuron 2007;55:187-98. Halpern C, Urtig H, Jaggi J, Grossman M, Won M, Baltuch G. Deep brain stimulation in neurologic disorders. Parkinsonims Relat Disord 2007;13:1-16. Halpern C, Samadani U, Litt BB, Jaggi JL, Baltuch GH. Deep brain stimulation for epilepsy. Neurotherapeutics 2008;5:59-67. Handforth A, DeSalles AA, Krahl SE. Deep brain stimulation of the subthalamic nucleus as adjunct treatment for refractory epilepsy. Epilepsia 2006;47:1239-41. Hendricks HT, van Limbeek J, Geruts AC, Zwarts MJ. Motor recovery after stroke: a systematic review of the literature. Arch Phys Med Rehab 2002;11:1629-1637. Hille B. Ion channels of excitable membranes. 3rd edition. Mass. Sinauer Associates Inc Sunderland, 2001. Hodaie M, Wennberg RA, Dostrovsky JO, Lozano AM. Chronic anterior thalamus stimulation for intractable epilepsy. Epilepsia 2002;43:603-8. Houeto JL, Karachi C, Mallet L, Pillon B, Yelnik J, Mesnage V et al. Tourettes syndrome and deep brain stimulation. J Neurol Neurosurg Psychiatr 2005;76:992-5. Howard MA, Volkov IO, Noh D, Garell C, Abbas PJ, Rubinstein JT et al. Auditory cortex neural prosthetic devices. In: Maciunas eds. Neural prostheses. Illinois: American Association of Neurological Surgeons Publications, Park Ridge. 2000; p. 273-286. Hummel FC, Cohen LG. Drivers of brain plasticity. Curr Opin Neurol. 2005a;18:667-74. Iadarola MJ, Gale K. Substantia nigra in GABA-mediated anticonvulsant actions. Adv Neurol 1982;44:343-64. Isenberg K, Down D, Pierce K, et al. Low frequency rTMS of the right frontal cortex is as effective as high frequency rTMS of the left frontal cortex for antidepressant-free, treatment-resistant depressed patients. Ann Clin Psychiatr 2005;17:153-9. Iyer MB, Schleper N, Wassermann EM. Priming stimulation enhances the derpessant effect of low-frequency repetitive transcranial magnetic stimulation. J Neurosci 2003;23:10867-72. Jackson A, Mavoori J, Fetz EE. Long-term motor cortex plasticity induced by an electronic neural implant. Nature 444 56-60 2006. Jimenez F, Velasco F, Salin-Pascual R, Hernandez JA, Velasco M, Criales JL, Nicolini H. A patient with a resistant major depression disorder treated with deep brain stimulation in the inferior thalamic peduncle. Neurosurgery 2005;57:585-93. Jorge RE, Robinson RG, Tateno K, et al. Repetitive transcranial magnetic stimulation as treatment of post-stroke depression: a preliminary study. Biol Psychiatr 2004;55:398405. Katayama Y, Oshima H, Fukaya C, Kawamata T, Yamamoto T. Congrol of post-stroke movement disorders using chronic motor cortex stimulation. Acta Neurochir Suppl 2002;79:89-92. Katz PS. Ion channels of excitable membres. Curr Opin Neurobiol 1995;5:799-808. Kern DS, Kumar R. Deep brain stimulation. The Neurologist 2007;13:237-252.
418
Kerrigan JF, Litt B, Fisher RS, et al. Electrical stimulation of the anterior nucleus of the thalamus for the treatment of intractable epilepsy. Epilepsia 2004;45:346-354. Kiss ZH, Doig-Beyaert K, Eliasziw M, et al. The Canadian multicentre study of deep brain stimulation for cervical dystonia. Brain 2007;130:2879-86. Kitagawa M, Murata J, Uesugi H, et al. Two-year follow-up of chronic stimulation of the posterior subthalamic white matter for tremor-dominant Parkinsons disease. Neurosurgery 2005;2005;56:281-9. Kleiner-Fisman G. Herzog J, Fisman DN, et al. Subthalamic nucleus deep brain stimulation: summary and meta-analysis of outcomes. Mov Disord 2006;21(Suppl /14):S290-S304. Krack P, Batir A, Van Blercom N, et al. Five-year follow-up of bilateral stimulation of the subthalamic nucleus in advanced Parkinsons disease. N Engl J Med 2003;349:1925-34. Krauss JK, Pohle T, Weber S, Ozdoba C, Burgunder JM. Bilateral stimulation of globus pallidus internus for treatment of cervical dystonia. Lancet 1999;354:837-8. Krauss JK. Deep brain stimulation for dystonia in adults. Overview and developments. Stereotact Funct Neurosurg. 2002;78:168-82. Krauss JK, Loher TJ, Weigel R, Capelle HH, Weber S, Burgunder JM. Chronic stimulation of the globus pallidus internus for treatment of non-dYT1 generalized dystonia and choreoathetosis: 2-year follow up. J Neurosurg. 2003;98:785-92. Kumar K, Toth C, Nath RK. Deep brain stimulation for intractable pain: a 15-year experience. Neurosurgery 1997;40:736-46. Kumar R, Lozano AM, Sime E, et al. Comparative effects of unilateral and bilateral subthalamic nucleus deep brain stimulation. Neurology 1999;53:561-6. Kumar R, Lozano AM, Sime E, Lang AE. Long-term follow-up of thalamic deep brain stimulation for essential and parkinsonian tremor. Neurology 2003;61:1601-04. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342:3149. Levy RM, Lamb S, Adams JE. Treatment of chronic pain by deep brain stimulation: longterm follow-up and review of the literature. Neurosurgery 1987;21:885-93. Levy R, Benson R, Winstein C, for the Everest Study Investigators. Cortical stimulation for upper-extremity hemiparesis from ischemic stroke: Everest Study primary endpoint results. International Stroke Conference, New Orleans, LA, February 20-22, 2008. Li J, Andrews RJ. Trimodal nanoelectrode array for precise deep brain stimulation: prospects of a new technology based on carbon nanofiber arrays. Acta Neurochir Suppl 2007;97(2):537-46. Lima MC, Fregni F. Motor cortex stimulation for chronic pain. Systematic review and metaanalysis of the literature. Neurology 2008;70:2329-37. Loher TJ, Capelle HH, Kaelin-Lang A, et al. Deep brain stimulation for dystonia: outcome at long-term follow-up. J Neurol 2008; 255: 881-884 Lotze M, Markert J, Sauseng P, et al. The role of multiple contralesional motor areas for complex hand movements after internal capsular lesion. J Neurosci 2006;26:6096-6102. Macdonell RA, Curatolo JM, Berkovic SF. Transcranial magnetic stimulation and epilepsy. J Clin Neurophysiol 2002;19:294-306. Mally J, Dinya E. Recovery of motor disability and spasticity in post-stroke after repetitive transcranial magnetic stimulation (rTMS). Brain Res Bull 2008;75:388-95.
419
Manola L, Holsheimer J: Motor cortex stimulation: role of computer modelling. In: Sakas DE and Simpson B, editors. Operative Neuromodulation: Neural Networks Surgery (Vol. II). New York: Springer-Verlag; 2007, 497-503. Mantovani A, Lisanby SH, Pieraccini F, Ulivelli M, Castrogiovanni P, Rossi S. Repetitive transcranial magnetic stimulation in the treatment of obsessive-compulsive disorder and Tourettes syndrome. Int J Neuropsychopharmacol 2006;9:95-100. Mantovani A, Leckman JF, Grantz H, King RA, Sporn AL, Lisanby SH. Repetitive transcranial magnetic stimulation of the supplementary motor area in the treatment of Tourettes syndrome: report of two cases. Clin Neurophysiol 2007a;118:2314-5. Mantovani A, Lisanby SH, Pieraccini F, Ulivelli M, Castrogiovanni P, Rossi S. Repetitive transcranial magnetic stimulation in the treatment of panic disorder with comorbid major depression. J Affect Disord 2007b;102:277-80. Marangell LB, Martinez M, Jrdi RA, Zboyan H. Neurostimulation therapies in depression: a review of new modalities. Acta Psychiatr Scand 2007;116:174-81. Marder E, Thiramulai V. Cellular, synaptic and network effects of neuromodulation. Neural Netw 2002;15:479-93. Mayberg HS, Lozano AM, Voon V, et al. Deep brain stimulation for treatment-resistant depression. Neuron 2005;45:651-60. Mink JW, Walkup J, Frey KA, et al. Patient selection and assessment recommendations for deep brain stimulation in Tourette syndrome. Mov Disord 2006;21:1831-8. Miocinovic S, Noecker AM, Maks CB, Butson CR, McIntyre CC. Cicerone: stereotactic neurophysiological recording and deep brain stimulation electrode placement software system. Acta Neurochir Suppl 2007;97(2):569-74. Montes C Mertens P Convers P Peyron R Sindou M Laurent B Mauguiere F Garcia-Larrea L Cognitive effects of precentral cortical stimulation for pain control: an ERP study, Neurophysiol. Clin./Clin. Nurophysiol 2002 ; 32 : 313-325 Mueller J, Skogseid IM, Benecke R, , et al. Pallidal deep brain stimulation improves quality of life in segmental and generalized dystonia: results from a prospective, randomized sham-controlled trial. Mov Disord. 2008;23:131-4. Munchau A, Bloem BR, Thilo KV, Trimble MR, Rothwell JC, Robertson MM. Repetitive transcranial magnetic stimulation for Tourette syndrome. Neurology 2002;59:1789-91. Murata J, Kitagawa M, Uesugi H, et al. Electrical stimulation of the posterior subthalamic area for the treatment of intractable proximal tremor. J Neurosurg 2003;99:708-15. Murase N, Rothwell JC, Kaji R, et al. Subthreshold low-frequency repetitive transcranial magnetic stimulation over the premotor cortex modulates writer's cramp. Brain 2005;128:104-15. Nandi D, Liu X, Winter JL, Aziz TZ, Stein JF. Deep brain stimulation of the peduculopontine region in the normal non-human primate. J Clin Neurosci 2002a;9:170-4. Nandi D, Chir M, Liu X, et al. Electrophysiological confirmation of the zona incerta as a target for surgical treatment of disabling involuntary arm movements in multiple sclerosis: use of local field potentials. J Clin Neurosci 2002b;9:64-8. Nuttin BJ, Gabriels LA, Cosyns PR. Long-term electrical capsular stimulation in patients with obsessive-compulsive disorder. Neurosurg 2003;52:1263-72.
420
OCD-DBS Collaborative Group. Deep brain stimulation for psychiatric disorders. Neurosurgery 2002;51:519. Oh MY, Abosch A, Kim SH, Lang AE, Lozano AM. Long-term hardware-related complications of deep brain stimulation. Neurosurgery 2002;50:1268-74. Oikawa H, Sasaki M, Tamakawa Y, Kamei A. The circuit of Papez in mesial temporal sclerosis: MRI. Neuroradiology 2001;43:205-10. Okun MS, Mann G, Foote KD, et al. Deep brain stimulation in the internal capsule and nucleus accumbens region: responses observed during active and sham programming. J Neurol Neurosurg Psychiatry 2007;78:310-4. Olivieri M, Turriziani P, Carlesimo Ga, et al. Parieto-frontal interactions in visual-object and visual-spatial working memory: evidence from trancranial magnetic stimulation. Cereb Cortex 2001; 11: 606-618 OReardon JP, Solvason B, Janicak PG, et al. Efficacy and safety of transcranial magnetic stimulation in the acute treatment of major depression: A multisite randomized controlled trial. Biol Psychiatry 2007;62:1208-16. Owen SL, Green AL, Nandi D, Bittar RG, Wang S, Aziz TZ. Deep brain stimulation for neuropathic pain. Neuromodulation 2006a;9:100-6. Owen SL, Green AL, Stein JF, et al. Deep brain stimulation for the alleviation of post-stroke neuropathic pain. Pain 2006b;120:202-6. Owen SL, Green AL, Davies P, et al. Connectivity of an effective hypothalamic surgical target for cluster headache. J Clin Neurosci 2007;14:955-60. Pancrazio JJ, Chen D, Fertig SJ, et al. Towards neurotechnology innovation: report from the 2005 Neural Interfaces Workshop. An NIH-sponsored event. Neuromodulation 2006;9:17. Picinelli M, Pini S, Bellantuono C, Wilkinson G. Efficacy of drug treatment in obsessive compulsive disorder. A meta-analysis review. Br J Psychiatry 1995;166:424-43. Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine nucleus for Parkinsons disease. Neuroreport 2005;16:1883-87. Prikryl R, Kasparek T, Skotakova S, Ustohal L Kucerova H, Ceskova E. Treatment of negative symptoms of schizophrenia using repetitive transcranial magnetic stimulation in a double-blind, randomized controlled study. Schizophr Res 2007;95:151-7. Raethjen J, Govindan RB, Kopper F, Muthuraman M, Deuschl G. Cortical involvement in the generation of essential tremor. J Neurophysiol 2007;97:3219-28. Rasche D, Rinaldi PC, Young RF, Tronnier VM. Deep brain stimulation for the treatment of various chronic pain syndromes. Neurosurg Focus 2006;21:E8 Rauch SL Dougherty DD, Malone D, et al. A functional neuroimaging investigation of deep brain stimulation in patients with obsessive-compulsive disorder. J Neurosurg 2006;104:558-65. Rezai AR, Lozano AM. Deep brain stimulation for chronic pain. In: Burchiel KJ, ed. Surgical management of pain. New York: Thieme; 2002. p. 565-76. Rodriguez-Oroz MC, Obeso JA, et al. Bilateral deep brain stimulation in Parkinson's disease: a multicentre study with 4 years follow-up. Brain 2005;128:2240-9.
421
Romanelli P, Bronte-Stewart H, Courtney T, Heit G. Possible necessity for deep brain stimulation of both the ventralis intermedius and subthalamic nuclei to resolve Holmes tremor. Case report. J Neurosurg 2003;99:566-71. Rutten WLC, Ruardij TG, Marani E, Roelofsen BH. Neural networks on chemically patterned electrode arrays: towards a cultured probe. Acta Neurochir Suppl 2007;97(2):547-54. Sachdev PS, Loo CK, Mitchell PB, McFarquhar TF, Malhi GS. Repetitive transcranial magnetic stimulation for the treatment obsessive-compulsive disorder. A double-blind controlled investigation. Psychol Med 2007;37:1645-9. Sakas DE, Panourias IG. Rostral cingulate gyrus: A putative target for deep brain stimulation in treatment-refractory depression. Med Hypothes 2006;66:491-4. Sakas DE, Panourias IG, Simpson BA. An introduction to neural networks surgery, a field of neuromodulation which is based on advances in neural networks science and digitised brain imaging. Acta Neurochir Suppl 2007a;97(2):3-13. Sakas DE, Panourias IG, Singounas E, Simpson BA. Neurosurgery for psychiatric disorders: from the excision of brain tissue to the chronic electrical stimulation of neural networks. In: Sakas DE, Simpson BA, editors. Operative Neuromodulation: Neural networks surgery. New York: Springer-Verlag; 2007b, 365-74. Sakas DE, Panourias IG, Boviatsis ES et al. Treatment of primary camptocormia (dystonic bent spine or bent neck) by electrical stimulation of the globus pallidus internus: report of 3 cases. Abstract book, The 8th International Neuromodulation Society and the 11th Annual Meeting of the North American Neuromodulation Society, Acapulco, 2007, December 7-12. Sakas DE, Flaskas P, Panourias IG, Georgakoulias N. Cortical electrical stimulation by interdural (patch-like) implantation of the electrodes: a new technique for increased efficacy and safety. Neurosurgery 2008a (under review) Sakas DE, Panourias IG, Boviatsis E, Themistocleous M, Stavrinou L, Gatzonis S. Treatment of idiopathic head-drop syndrome (camptocephalia) by bilateral chronic electrical stimulation of the globus pallidum internus. J Neurosurg 2008b [in press]. Sanguineti V, Giugliano M, Grattarola M, Morasso P. Neuroengineering: from neural interfaces to biological computers. In: Riva G, Davide F, editors. Communications throuhg virtual technology: identity community and technology in the internet age. Amsterdam, IOS Press; 2003. p. 233-46. Schlaepfer TE, Cohen MX, Frick C,et al. Deep brain stimulation to reward circuitry alleviates anhedonia in refractory major depression. Neuropsychopharmacology 2008 ; 33:368-77. Schoenen J, Di Clemente L, Vandenheede M, et al. Hypothalamic stimulation in chronic cluster headache: a pilot study of efficacy and mode of action. Brain 2005;128:940-7. Schramm P, Scheihing M, Rasche D, Tronnier VM. Behr syndrome variant with tremor treated by VIM stimulation. Acta Neurochir (Wien) 2005;147:679-83. Silanp M, Schmidt D. Natural history of treated childhood-onset epilepsy: prospective, long-term population-based study. Brain 2006;129:617-24. Silver R, Boahen K, Grillner S, Kopell N, Olsen KL. Neurotech for neuroscience: unifying concepts, organizing principles, and emerging tools. J Neurosci 2007;27:11807-19.
422
Simpson BA. The role of neruostimulation: the neurosurgical perspective. J Pain Symptom Manage 2006a;31:S3-S5. Simpson BA. Challenges for the 21st century: the future of electrical neuromodulation. J Pain Symptom Manage 2006b;31 (Suppl 4):S3-S5. Smith LS, Hamilton A. Neuromorphic systems. Introduction. Int J Neural Syst 1999;9;371-3 Starr PA, Barbaro NM, Raskin NH, Ostrem JL. Chronic stimulation of the posterior hypothalamic region for cluster headache: technique and 1-year results in four patients. J Neurosurg 2007;106:999-1005. Stover NP, Okun MS, Evatt ML, Raju DV, Bakay RA, Vitek JL. Stimulation of the subthalamic nucleus in a patient with Parkinsons disease and essential tremor. Arch Neurol 2005;62:141-3. Sturm V, Lenartz D, Koulousakis A, et al. The nucleus accumbens: a target for deep brain stimulation in obsessive-compulsive and anxiety disorders. J Chem Neuroanat 2003;26:293-9. Sun B, Chen S, Zhan S, Le W, Krahl SE. Subthalamic nucleus stimulation for primary dystonia and tardive dystonia. Acta Neurochir Suppl 2007;97:207-14. Swanson TH. The pathophysiology of human mesial temporal lobe epilepsy. J Clin Neurophysiol 1995;12:2-22. Sydow O, Thobois S, Alesch F, Speelman JD. Multicenter European study of thalamic stimulation in essential tremor: a six year follow-up. J Neurol Neurosurg Psychiatr 2003;74:1387-91. Taha JM, Jansen MA, Favre J. Thalamic deep brain stimulation for the treatment of head, voice, and bilateral limb tremor. J Neurosurg 1999;91:68-72. Talelli P, Rothwell J. Does brain stimulation after stroke have a future? Curr Opin Neurol 2006;19:543-50. Temel Y, Kessels A, Tan S, Topdag A, Boon P, Visser-Vandewalle V. Behavioural changes after bilateral subthalamic nucleus stimulation in advanced Parkinsons disease: A systematic review. Parkinsonism Relat Disord 2006;12:265-72. Tisch S, Rothwell JC, Limousin P, Hariz MI, Corcos DM. The physiological effects of pallidal deep brain stimulation in dystonia. IEEE Trans Neural Syst Rehabil Eng 2007;15:166-72. Topper R, Mottaghy FM, Brugmann M, Noth J, Huber W. facilitation of picture naming by focal transcranial magnetic stimulation of wernickes area. Exp Brain Res 1998; 121: 371-378 Velasco F, Velasco AL, Velasco G, Jimenez F. Effect of chronic electrical stimulation of the centromedian thalamic nuclei on various intractable seizures patterns: I. Clinical seizures and paroxysmal EEG activity. Epilepsia 1993a;34:1065-74. Velasco F, Marquez I, Velasco G. Effect of chronic electrical stimulation of the centromedian thalamic nuclei on various intractable seizure patterns: II. Psychological performance and background EEG activity. Acta Neurochir Suppl 1993b;58:201-4. Velasco M, Velasco AL, Jimenez F et al. Role of the centromedian thalamic nucleus in the genesis, propagation and arrest of epileptic activity. An electrophysiological study in man. Epilepsia 1995;36:63-71.
423
Velasco F, Velasco M, Jimenez F, et al. Predictors in the treatment of difficult-to-control seizures b electrical stimulation of the centromedian thalamic nucleus. Neurosurgery 2000a;47:295-304. Velasco AL, Velasco M, Velasco F, et al. Sucabute and chronic electrical stimulation of the hippocampus on intractable temporal lobe seizures: preliminary report. Arch Med Res 2000b;31:316-28. Velasco M, Velasco AL, Menez D, Rocha L. Centromedian-thalamic and hippocampal electrical stimulation for the control of intractable epileptic seizures. Stereotactic Funct Neurosurg 2001;77:223-7. Velasco F, Carrillo-Ruiz JD, Brito F, et al. Double-blind, randomized controlled pilot study of bilateral cerebellar stimulation for treatment of intractable motor seizures. Epilepsia 2005;46:1071-81. Visser-Vandewalle V, Temel Y, Boon P, et al. Chronic bilateral thalamic stimulation: a new therapeutic approach in intractable Tourettes syndrome. Report of three cases. J Neurosurg 2003;99:1094-1100. Vonck K, Boon P, Claeys P, Dedeurwaerdere S, Achten R, Van Roost D. Long-term deep brain stimulation for refractory lobe epilepsy. Ann Neurol 2005;52:556-65. Warwick K, Gasson MN, Spiers AJ. Therapeutic potential of computer to cerebral cortex implantable devices. Acta Neurochir Suppl 2007;97(2):529-35. Weaver F. Follett K, Hur K, Ippolito D, Stern M. Deep brain stimulation in Parkinsons disease: a meta-analysis of patient outcomes. J Neurosurg 2005;103:956-67. Webster TJ, Waid MC, McKenzie JL, Price RL, Ejiofor JU. Nano-biotechnology: carbon nanofibers as improved nerual and orthopaedic implants. Nanotechnology 2004;15:4854. Wichmann T, DeLong MR. Deep brain stimulation for neurologic and neuropsychiatric disorders. Neuron 2006;52:197-204. Wishart HA, Roberts DW, Roth RM, et al. Chronic deep brain stimulation for the treatment of tremor in multiple sclerosis: review and case reports. J Neurol Neurosurg Psychiatr 2003;74:1392-7. Young R, Rinaldi PC. Brain stimulation. In: North RB, Levy RM, editors. Neurosurgical management of pain. New York: Springer; 1997. p. 283-301. Yu H, Neimat J. The treatment of movement disorders by deep brain stimulation. Neurotherapeutics 2008;5:26-36. Ziemann U, Lonnecker S, Steinhoff BJ, Paulus W. Effects of antiepileptic drugs on motor cortex excitability in humans: transcranial magnetic stimulation study. Ann Neurol 1996;40:367-78.
INDEX
A
abatement, 120, 124, 147, 148, 174, 334 abdomen, 357 abduction, 210 abnormalities, 190, 204, 212, 225, 238 ACC, 144, 170, 171, 174, 211, 338 accessibility, 133, 405 accidental, 26, 207 accommodation, 92 accounting, 289 accuracy, xiii, 33, 48, 56, 234, 240, 407, 411, 413 acid, 49, 401 acoustic, 211, 293, 294, 295, 347, 410 action potential, 47, 53, 76, 77, 82, 83, 86, 97, 100, 110, 179, 371, 406 acute, xii, 54, 62, 110, 174, 178, 194, 202, 210, 217, 232, 234, 237, 239, 242, 244, 245, 249, 254, 263, 269, 270, 271, 272, 304, 305, 310, 320, 324, 334, 373, 375, 402, 420 acute ischemic stroke, 245, 269 adaptation, 371, 378 addiction, xiii ADHD, 402 adhesion, 26 adhesions, 383 adjunctive therapy, 193 adjustment, 384 administration, 49, 59, 268 adolescents, 291 adult, 8, 52, 111, 272, 291, 368, 376 adults, 249, 269, 286, 351, 376, 418 adverse event, 204, 207, 208, 212, 322, 332 affective disorder, 304 affective states, 173 age, xiii, 144, 153, 202, 204, 212, 231, 236, 252, 267, 303, 337, 368, 376, 377, 393, 395, 408, 421 agents, 297, 399 aging, xiii agonist, 134, 204 aid, 18, 243, 265, 376 air, 24, 329 akinesia, 183, 203, 209, 210, 212, 215, 223, 394, 397 alcohol, 207, 299, 300 alcoholism, 154 alexia, 377, 383 algorithm, 320, 321, 323, 329, 412 allodynia, 120, 142, 143, 145, 151, 152, 166, 174 alloys, 29 alpha, 268, 271, 277, 284, 302, 338, 349, 357, 360 ALS, 211, 212 alternative, 167, 194, 310, 313, 348, 351, 393, 394, 396, 398, 400, 404, 408 alternatives, xi, 402 alters, 47, 196 aluminum, 329 ambiguity, 51 amelioration, 187, 210 American Heart Association, 243, 270 amplitude, 20, 21, 22, 36, 38, 39, 40, 47, 58, 60, 61, 62, 63, 65, 76, 77, 82, 85, 91, 92, 100, 101, 104, 106, 107, 108, 109, 135, 138, 154, 155, 173, 176, 177, 186, 187, 206, 234, 263, 312, 323, 324, 379, 412 amputation, 164 amputees, 23, 36 Amsterdam, 111, 361, 385, 421 AMT, 234, 239, 240 amygdala, 171, 296, 297, 338 amyotrophic lateral sclerosis, 65, 194, 211, 215 anaesthesia, 67, 110, 212
426
Index
Asia, 392 aspiration, 283 assessment, 39, 135, 165, 174, 185, 186, 187, 202, 207, 208, 215, 223, 241, 252, 270, 272, 279, 280, 296, 344, 347, 381, 413, 419 assessment tools, 272 assignment, 235 assumptions, 105 astrocyte, 386 asymmetry, 338, 354, 356 asynchronous, 310 ataxia, 397 Athens, 391 athetosis, 211, 224 atlas, 15 atrophy, 24, 26, 66, 153, 165, 202, 203, 204, 212, 213, 264, 291, 383, 396, 408 attacks, 121, 269, 296, 305, 309, 392 auditory cortex, ix, xii, 277, 293, 300, 338, 340, 341, 342, 343, 344, 345, 346, 348, 349, 350, 351, 353, 354, 355, 356, 357, 358, 359, 360, 361, 383, 386 auditory hallucinations, 293, 300, 301, 302, 303, 304, 305, 345, 414, 416 auditory nerve, 368 auditory stimuli, 377 aura, 4, 311 Australasia, 231 availability, 3, 12, 26, 173, 376, 393, 405, 409 avoidance, vii, 33, 296 avoidance behavior, 296 awareness, 276, 277, 278, 279, 285, 286, 413 axon, 49, 58, 59, 61, 64, 70, 77, 78, 83, 86, 97, 98, 103, 107, 108 axonal, 70, 98, 107, 286, 298 axons, 46, 54, 57, 58, 59, 60, 63, 64, 65, 70, 71, 73, 77, 78, 83, 87, 93, 97, 105, 108, 109, 111, 175, 177, 219, 279, 368
analgesia, 22, 41, 42, 109, 118, 128, 132, 133, 134, 135, 140, 145, 153, 154, 155, 170, 171, 174, 175, 267, 392, 409 analgesic, 38, 42, 118, 132, 133, 134, 135, 137, 140, 141, 144, 149, 154, 155, 179 analog, 38, 124, 372, 413 analysis of variance, 207 anatomy, xi, 71, 78, 87, 110, 157, 168, 226, 244, 284, 376, 385 anesthetics, 312 angiogram, 18 angulation, 339, 340, 342, 346 animal models, 373, 399 animal studies, 105, 108, 134 animals, 47, 105, 359 anisotropy, 88, 279 anode, 22, 24, 46, 51, 71, 73, 78, 79, 80, 81, 84, 87, 91, 92, 94, 99, 101, 103, 104, 105, 107, 109, 154, 155, 177, 178, 206, 207, 208, 324 anodes, 76, 154, 251 anomalous, 134 ANS, 23, 280, 357, 408, 409 antagonism, 322 antagonist, 49 antagonistic, 401 antibiotic, 329, 332 antibiotics, 41 anticonvulsant, 309, 399, 417 anticonvulsants, 42 antidepressant, 188, 290, 291, 303, 304, 305, 344, 403, 417 antidepressant medication, 291 antidepressants, 291, 304 anxiety, 54, 204, 207, 295, 296, 297, 301, 302, 396, 404, 422 anxiety disorder, 295, 422 apathy, 394, 396, 408 aphasia, 27, 147, 231, 237, 245, 246, 247, 251, 254, 255, 260, 261, 262, 266, 269, 271, 272, 405 apoptosis, 268 application, xii, 39, 41, 46, 47, 50, 52, 56, 87, 89, 110, 118, 133, 135, 157, 193, 235, 243, 276, 289, 329, 337, 338, 346, 367, 377, 379, 381, 392, 397, 400, 403, 407, 409, 412 arousal, 276, 279, 280, 284 arrest, 41, 211, 221, 422 arteries, 249 artery, 19, 234, 237, 261 arthritis, 126 aseptic, 41
B
back pain, 117, 138 bandwidth, 320, 370 banks, 78 basal ganglia, 143, 171, 184, 189, 191, 194, 197, 207, 218, 219, 220, 224, 235, 294, 297, 299, 394, 397, 398, 410 Basal ganglia, 172, 196, 225 basket cells, 70 batteries, 86
Index
battery, 28, 50, 69, 83, 92, 155, 166, 205, 283, 322, 325, 327, 333, 395, 401, 403, 409, 411 BDNF, 62, 65, 190, 194, 300 behavior, xii, xiii, 51, 59, 97, 108, 204, 222, 234, 242, 250, 266, 276, 280, 299, 311, 378, 386, 411 behavioral assessment, 207, 276, 278 behavioral change, xii, 238, 401 behavioral effects, 50 behavioral recovery, 232, 246, 259, 272 Belgium, 33, 337, 353 benchmarks, 258 bending, 65 beneficial effect, 189, 200, 212, 232, 301, 311, 315, 344, 393, 395, 396, 398, 402 benefits, 188, 205, 283, 395, 396, 397, 402, 403, 404, 409 benign, 41, 117 bias, 238, 411 binding, 173, 185, 208, 223 biocompatibility, 373, 375, 382 biocompatible, 369, 372, 375 biological systems, 372 biophysics, 176 biosensors, 373 biotechnology, 423 bipolar, 20, 21, 22, 24, 27, 35, 36, 38, 41, 42, 48, 80, 81, 87, 92, 105, 108, 139, 154, 177, 202, 203, 205, 211, 212, 250, 265, 304, 311, 322, 324, 357, 407, 408 bladder, 367 bleeding, 24, 38, 165, 407 blindness, 366, 367, 376, 380, 382, 384, 404 blink reflex, 280 blocks, 312, 330 blood, 72, 171, 184, 190, 197, 199, 226, 249, 258, 265, 276, 316, 330, 345, 374 blood flow, 171, 258, 345 blood pressure, 276, 330 blood supply, 249 blood vessels, 374 BOLD, 22, 354, 356, 361 bolus, 153, 341 boredom, xiii Boston, xii, 226, 318 botulinum, 191, 196, 210 boundary conditions, 72, 95 brachial plexus, 20, 36, 128, 162, 164, 168, 170, 392 bradykinesia, 184, 186, 187, 189, 194, 200, 203, 219, 394, 395 bradykinetic, 184
427
Braille, 366, 377, 380, 381, 382, 383 brain activity, 278, 285, 305, 345, 346, 411, 412, 413 brain damage, 26, 39, 169, 246, 276, 278, 378 brain injury, 159, 275, 276, 280, 285, 286 brain stem, 133, 249 brain structure, 317, 399 brainstem, 51, 122, 124, 128, 143, 146, 147, 148, 149, 150, 151, 171, 272, 396, 399, 402, 410 browser, 323 buffer, 322 bulbar, 157, 267, 271 burning, 38, 42, 50, 145, 149, 151
C
C++, 320 cables, 41, 42, 43, 91, 165 cadaver, 9, 11 caliber, 93, 103, 105 calvaria, 18 cAMP, 134 Canada, 290, 392 cancer, 134, 392 candidates, 97, 146, 268, 310, 324, 342, 360, 376, 377, 393, 398, 401, 403 capacitance, 74, 75, 76, 98, 99 capsule, 141, 142, 175, 259, 260, 401, 403, 416, 419 carbon, 329, 413, 418, 423 cardiac pacemaker, 52, 310 carrier, 353, 356 case study, 344, 351 cast, 268 catecholamine, 67 cathode, 22, 24, 46, 51, 73, 77, 80, 81, 84, 87, 91, 92, 94, 99, 100, 101, 103, 104, 105, 107, 109, 154, 177, 178, 207, 208, 324 Catholic, 57, 231 cauda equina, 122 C-C, 265 ceiling effect, 188, 257, 354 cell, 58, 59, 61, 63, 64, 70, 71, 83, 87, 92, 93, 96, 97, 98, 99, 100, 101, 103, 105, 106, 107, 108, 109, 111, 176, 296, 297, 368, 371, 379, 382 cell body, 58, 59, 61, 63, 64, 83, 96, 97, 98, 99 cement, 253 central nervous system, 85, 88, 89, 111, 179, 238, 309, 361, 382, 385 cerebellar ataxia, 397 cerebellum, 50, 56, 172, 192, 210, 399
428
Index
codes, 371 coding, 177, 352, 372, 380, 382, 385, 387, 411 cofactors, 31, 55, 88 cognition, 254 cognitive capacities, 278 cognitive dysfunction, 408 cognitive function, 194, 202, 285, 377, 402, 414 cognitive impairment, xiii, 269, 310, 408 cognitive performance, xiii, 268, 271, 413 cognitive process, 28, 277, 278 cognitive processing, 277 coherence, 219, 284, 286 cohort, 267, 317 coil, xii, 46, 47, 48, 50, 52, 58, 59, 60, 64, 65, 66, 67, 118, 119, 120, 121, 122, 124, 126, 127, 128, 135, 137, 157, 185, 186, 187, 192, 239, 240, 242, 290, 291, 292, 303, 314, 315, 317, 325, 339, 340, 341, 342, 344, 345, 346, 350, 401, 403 collaboration, 246, 359, 407 collateral, 269 coma, 278, 285 communication, 26, 39, 140, 163, 167, 204, 213, 258, 261, 267, 268, 280, 285, 323, 371, 373, 406, 412, 414 communication technologies, 371 community, 421 co-morbidities, 408 comorbidity, 295, 396 competition, 268 complement, 279 complex partial seizure, 315, 332, 400 complex regional pain syndrome, 137, 167, 210 complexity, 76, 410 complications, 28, 36, 38, 150, 155, 165, 166, 204, 207, 213, 266, 310, 393, 396, 407, 408, 412, 419 components, 20, 42, 70, 73, 191, 277, 280, 320, 322, 323, 330, 355, 370, 371, 395, 406 composition, 93, 98, 381 comprehension, 255, 258, 278 computed tomography, 15, 169, 183 computer science, 371 concentration, 219 conditioning, 49, 296 conductance, 98, 99, 176, 411 conduction, 48, 110, 220 conductive, 71, 72, 73, 84, 100 conductivity, 71, 72, 73, 75, 78, 95, 99, 103, 413 conductor, 65, 76, 92, 93, 97, 98, 108 configuration, 13, 28, 38, 73, 154, 155, 206, 324 confusion, 51
cerebral blood flow, 170, 184, 190, 197, 199, 226, 316, 345 cerebral cortex, 31, 47, 54, 55, 58, 60, 62, 63, 66, 87, 88, 89, 110, 111, 224, 226, 232, 341, 355, 372, 373, 375, 382, 384, 385, 410, 423 cerebral hemisphere, 70, 73, 222, 405 cerebral hemorrhage, 28 cerebral metabolism, 277, 296 cerebrospinal fluid, 27, 71, 93, 407 cerebrovascular, 269, 402 cerebrovascular disease, 269, 402 cerebrum, xii, 5, 7, 15 channels, 76, 86, 93, 98, 107, 322, 323, 417 chaos, 319 charge density, 29, 83, 324 charged particle, 72 childhood, 52, 421 children, 52, 291 cholecystokinin, 305 choreoathetosis, 418 chronic pain, viii, xiii, 38, 88, 111, 117, 118, 119, 132, 133, 134, 136, 137, 140, 166, 169, 243, 250, 392, 393, 403, 407, 415, 418, 420 cigarette smoking, 299, 300 classes, 70, 84 classical, 17, 224 classification, 159, 232, 401 clavicle, 24 clinical assessment, 185, 187, 207, 279, 408 clinical examination, 276 clinical symptoms, 298 clinical trial, 189, 233, 243, 289, 290, 294, 295, 296, 298, 299, 300, 317, 333, 349, 366, 373, 393, 401, 409, 415 clinical trials, 233, 243, 289, 290, 294, 295, 296, 298, 299, 349, 366, 373, 393, 401, 409 clinician, 33, 38, 325 closed-loop, xii, 316, 320, 333, 400, 411 closure, 27, 36, 41, 330, 332, 412 clouds, 368 cluster headache, 392, 414, 420, 421 clusters, 205 CMOS, 383 CNS, 15, 111, 253, 385, 413 CNS tumors, 15 CO2, 24, 173 coagulation, 27, 36, 357 cocaine, 299, 300 cochlear implant, 367, 368, 379 Cochrane, 298, 303, 304
Index
Congress, 136, 158, 159 conjecture, xi connectivity, 70, 93, 172, 176, 178, 198, 225, 243, 254, 255, 271, 272, 276, 284 conscious perception, 276, 278, 338 consciousness, 275, 276, 278, 279, 280, 283, 284, 285, 286, 312 consensus, 18, 134, 338, 403 consent, 207 consolidation, 232 constraints, 322, 323, 372 construction, 412 consumption, 104 contiguity, 221 continuity, 72 contractions, 59, 142, 253, 263 contralateral hemisphere, 254, 260 control, ix, xiii, 30, 31, 40, 41, 44, 47, 48, 50, 88, 89, 104, 111, 133, 135, 137, 142, 143, 146, 150, 157, 163, 165, 166, 172, 175, 178, 179, 191, 192, 193, 211, 214, 217, 221, 226, 231, 244, 252, 271, 279, 309, 310, 312, 313, 315, 316, 319, 320, 334, 339, 340, 342, 343, 344, 346, 359, 367, 370, 382, 394, 400, 411, 412, 419, 422 control condition, 339, 340, 342, 343, 344, 346 control group, 193, 252, 315 controlled studies, 154, 167, 192, 296, 299, 346, 395, 399, 403, 405 controlled trials, 240, 290, 315, 394, 398, 402 contusion, 408 contusions, 283 convergence, 177, 410 convex, 93, 328 Copenhagen, 270 coping strategies, 338 copper, 46, 47 corona, 271 corpus callosum, 19, 220, 225, 283, 284 correlation, 20, 22, 40, 135, 137, 172, 213, 222, 259, 337, 339 correlation analysis, 213 correlations, 168, 179, 189, 361 corrosion, 29, 85 cortical asymmetry, 354 cortical hyperexcitability, 191 cortical inhibition, 49, 65, 196, 199, 309 cortical localization, vii, 3, 4 cortical neurons, 60, 62, 70, 71, 82, 84, 85, 88, 104, 111, 220, 253, 384, 400 cortical processing, 277
429
cortical silent period, 186 corticospinal, 22, 30, 32, 48, 55, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 88, 109, 112, 153, 175, 179, 190, 196, 200, 231, 234, 240, 242, 244, 247, 249, 250, 265, 271 cosine, 48 cost-effective, 396, 414 costs, 53, 231 coupling, 38, 175, 232, 243, 277, 278, 284, 369, 382 covering, xi, 36, 38, 165, 355 cranial nerve, 161, 220, 351 craniotomy, 4, 17, 19, 23, 24, 26, 34, 35, 38, 41, 212, 265, 327, 328, 330, 357, 407, 412 cranium, 7, 412 craving, 299, 300, 301, 305 credentials, 218 critical period, 376 cross-fertilization, xi cross-sectional, 247 CRPS, 120, 164, 167 CRS, 280 crying, xii CSF, 71, 72, 73, 78, 93, 94, 99, 100, 101, 103, 108, 109, 407 CSR, 280 CT scan, 11, 38, 209, 329, 354 cues, 379 curettage, 155 cycles, 173, 310 cycling, 42, 92, 144, 165, 166, 400 cyst, 179 cytoarchitecture, 32, 70
D
daily living, 204, 252, 268, 337 Dallas, 243 data analysis, 23 data set, 150, 278 database, 128 de novo, 195 death, 66, 149, 164, 202, 231, 276 decay, 47 decision making, 299, 402 decisions, 270, 383 decoding, 412 decompression, 351 decortication, 277
430
Index
disability, xiii, 193, 231, 239, 347, 404, 418 disabled, 26, 34, 231, 276, 365, 412 discharges, xii, 49, 310, 311, 313, 314, 317, 333, 372 discomfort, 409 discretization, 95 discrimination, 175, 379, 386 discriminatory, 410 discs, 73 disease progression, 212 diseases, 45, 257, 366 disinhibition, 191, 192, 193, 245, 347 dislocation, 27, 165 disorder, 69, 85, 191, 209, 234, 258, 289, 295, 297, 300, 301, 303, 304, 401, 404, 406, 407, 413, 415, 417, 420 dispersion, 222 displacement, 20, 26, 222 disposition, 221 dissatisfaction, 140 distortions, 407 distress, 337, 338, 358, 360 distribution, 36, 48, 62, 73, 76, 81, 82, 89, 91, 93, 119, 120, 121, 122, 127, 128, 154, 220, 225, 278, 279, 312, 338, 372, 405 divergence, 410 diversity, 222 division, 296 dizziness, 309, 359 dogs, 383 doors, xiii dopamine, 134, 185, 188, 189, 199, 200, 204, 208, 219, 295, 305, 394 dopamine agonist, 204, 394 dopaminergic, 189, 190, 199, 294, 299, 394, 396 dopaminergic neurons, 189, 190 dorsal horn, 175 dorsolateral prefrontal cortex, xiii, 132, 183, 185, 195, 203, 205, 279, 280, 290, 292, 297, 298, 300, 304, 305, 343, 402 dosage, 151, 204 dosing, 51, 291 double blind study, 241, 303 double-blind trial, 15, 44, 159, 162, 164, 168 downregulating, 233 down-regulation, 173 drowsiness, 309 DRS, 187, 280 drug abuse, 165, 207 drug addict, 154 drug addiction, 154
deep brain stimulation, ix, xi, xiii, 28, 69, 85, 88, 202, 225, 276, 297, 391, 412, 414, 415, 416, 417, 418, 419, 420, 422, 423 deep-sea, 398 defibrillation, 88 deficiency, 189 deficit, 144, 151, 152, 184, 254, 258, 259, 266, 267, 268, 271, 300, 305, 316 deficits, 23, 153, 217, 226, 238, 244, 251, 269, 405 definition, 177, 401, 416 degenerate, 366, 367 delivery, 50, 62, 71, 265, 312, 321, 322, 323, 334, 408 deltoid, 51 delusion, 293 dementia, xiii, 202, 395, 414 dendrites, 70, 71, 76, 77, 88, 93, 109 dendritic spines, 198 denervation, 36 density, 29, 31, 50, 52, 55, 71, 72, 73, 78, 79, 80, 88, 89, 91, 98, 100, 101, 372, 373, 378, 379 depolarization, 77, 86, 99, 101, 234, 313 deposition, 29 depressed, 51, 52, 67, 195, 305, 402, 417 depression, xii, 6, 50, 52, 54, 55, 57, 62, 67, 165, 179, 185, 187, 188, 194, 196, 199, 204, 208, 220, 232, 269, 289, 290, 291, 292, 299, 300, 301, 302, 303, 304, 305, 337, 344, 345, 348, 350, 382, 394, 402, 408, 416, 417, 419, 420 depressive disorder, 54, 289, 295, 304 depressive symptoms, 208, 231, 289, 294, 296 deprivation, 378, 382 derivatives, 77 destruction, 164 desynchronization, 219 detachment, 24, 26 detection, 27, 35, 136, 138, 175, 319, 320, 321, 322, 323, 324, 326, 332, 333, 334, 341, 345, 400, 411, 412 diabetes, 28 diabetes mellitus, 28 diagnostic criteria, 285 differential equations, 98 differentiation, 184, 275, 277, 278 diffusion tensor imaging (DTI), 39 diplopia, 393 dipole, 20, 51, 174 direct measure, 58 direct observation, 40 dirichlet boundary conditions, 95
Index
drug therapy, 212 drug treatment, 290, 309, 310, 420 drug-induced, 184 drug-resistant, 294, 295, 304, 324, 393, 399, 400, 404 drugs, xiii, 140, 141, 143, 144, 172, 188, 198, 291, 299, 309, 312, 317, 394, 399, 400, 401, 423 DuPont, 360 dura mater, 23, 24, 26, 34, 71, 72, 78, 93, 99 duration, 20, 30, 36, 38, 49, 60, 77, 82, 83, 92, 97, 102, 117, 119, 120, 121, 124, 127, 128, 133, 135, 137, 144, 155, 165, 213, 219, 222, 232, 235, 236, 240, 251, 257, 263, 264, 291, 302, 310, 311, 312, 313, 321, 322, 323, 333, 338, 339, 341, 342, 343, 347, 348, 360, 375, 376, 379, 400, 406, 413 dynamical system, 310 dysarthria, 27 dyskinesia, 186, 187, 199, 205 dysphagia, 209, 260 dysplasia, 314, 316, 317, 318, 400 dysregulated, 296 dysregulation, 402 dystonia, 45, 183, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 210, 215, 250, 397, 398, 412, 413, 414, 415, 417, 418, 419, 422
431
E
ECS stimulating array, 33 edema, 52 Eden, 194 EEG, 13, 20, 30, 51, 62, 88, 207, 208, 225, 236, 268, 276, 277, 279, 285, 286, 311, 315, 316, 317, 319, 320, 322, 323, 324, 333, 338, 339, 342, 343, 344, 346, 354, 355, 357, 360, 400, 401, 422 EEG activity, 311, 401, 422 efferent nerve, 93 Egypt, 183 elderly, 28, 291, 303, 304, 396, 407, 416 elderly population, 416 electric charge, 46 electric current, 46, 53, 410 electric field, 46, 47, 48, 52, 62, 64, 71, 72, 73, 75, 85, 89, 177, 310 electric potential, 72, 78 electrical conductivity, 103, 413 electrical fields, 177 electrical properties, 71, 93 electrical pulses, 321 electrical resistance, 43, 46
electrochemical reaction, 29, 74 electrode placement, 14, 20, 22, 141, 317, 357, 419 electroencephalography, 51, 320 electrographic seizures, 311, 321 electrolysis, 47 electrolyte, 28, 381 electromagnetic, xii, 28, 46, 48, 66, 353, 361 electromagnetic tomography, 353, 361 electromagnetism, xii, 72 electromyography, 132, 284 electron, 88, 89, 176, 179, 374 electrons, 73 electrophysiological study, 178, 422 electrophysiology, 234, 326 email, 3 EMG, 21, 49, 67, 176, 278, 411, 412 emission, 12, 169, 178, 183, 195, 197, 198, 226, 244, 247, 296, 338, 351 emotional, 170, 173, 278, 338 emotional distress, 338 emotional processes, 170 emotional responses, 173 empyema, 40, 41 encapsulated, 93 encapsulation, 379, 385 encoding, 225, 379, 384, 385, 387, 410, 411 endocrinological, xiii energy, 47, 48, 83, 86, 87, 104, 224, 324 energy consumption, 104 energy efficiency, 83, 87 engagement, 278 England, 206, 366, 385 enrollment, 252 enthusiasm, 409 environment, 39, 98, 275, 379 ependymoma, 149, 151 epidemiology, 349 epidural hematoma, 24, 26 epilepsy, ix, xii, 51, 207, 224, 257, 309, 310, 312, 315, 316, 317, 318, 319, 320, 322, 324, 333, 334, 359, 399, 400, 401, 411, 413, 414, 415, 416, 417, 418, 421, 423 epileptic seizures, 20, 28, 144, 156, 208, 333, 334, 360, 366, 399, 422 episodic memory, 279 erosion, 329 ERPs, 27, 276, 277, 278 escitalopram, 300 essential tremor, 210, 215, 396, 420, 421, 422 estimating, 376
432
Index
fear, 296, 404 feature selection, 333 feedback, 175, 219, 310, 355, 371 feeding, 283 feelings, 295, 297, 359 females, 289, 358 fetus, 52 fiber, 13, 65, 76, 77, 78, 84, 89, 91, 96, 97, 98, 99, 101, 102, 103, 105, 107, 108, 177, 283, 284, 329 fiber bundles, 13 fibers, 72, 73, 76, 77, 78, 79, 80, 82, 84, 85, 89, 93, 94, 97, 102, 104, 105, 106, 108, 109, 134, 175, 176, 177, 178, 212, 219, 220, 221, 225, 253, 260, 261, 368, 371, 410 fibromyalgia, 117, 136, 137 fibrosis, 42, 155, 212 film, 18, 373 filters, 371 finite element method, 75 fire, 47, 372 fires, 86, 372 fish, xii fixation, 35, 36, 37, 41, 275, 325, 329 flashbacks, 296 flex, 251 flexibility, 104 floating, 367, 383, 384 flow, 28, 46, 50, 60, 71, 72, 73, 76, 77, 78, 79, 103, 132, 170, 184, 190, 197, 199, 211, 221, 226, 258, 269, 316, 345, 366 fluctuations, 204, 207, 212, 285 fluid, 27, 71, 76, 93, 407 fluoroscopy, 329, 332 fluoxetine, 187, 194, 196 fMRI, 12, 22, 23, 25, 34, 35, 36, 37, 39, 153, 169, 170, 174, 183, 195, 221, 236, 244, 246, 247, 251, 253, 254, 259, 260, 261, 262, 263, 264, 265, 269, 271, 277, 278, 340, 341, 345, 352, 353, 354, 356, 357, 361, 376, 380, 401, 406, 407, 409 focal seizure, 41, 166, 311 focusing, 199, 252, 343 folding, 3, 39, 60, 91 food, 299, 305 Food and Drug Administration, 392 foramen, 14 fovea, 367 FPGA, 372 fracture, 28, 121, 126 framing, 279 France, 91
etiology, 119 etiopathogenesis, 297 Europe, 231, 392 evacuation, 367 event-related potentials (ERPs), 27, 285 evoked potential, 17, 22, 30, 31, 32, 35, 39, 48, 53, 55, 58, 66, 109, 110, 111, 112, 132, 135, 138, 140, 154, 157, 173, 178, 183, 187, 194, 195, 199, 227, 267, 270, 278, 350, 376, 406, 407 evolution, 376 excision, 310, 311, 400, 421 excitability, xii, 46, 47, 48, 50, 51, 53, 54, 55, 56, 57, 58, 59, 62, 63, 65, 66, 67, 85, 99, 108, 109, 118, 138, 183, 184, 185, 187, 189, 190, 191, 192, 193, 194, 195, 196, 197, 199, 200, 219, 225, 226, 232, 233, 234, 235, 238, 241, 242, 244, 245, 247, 255, 267, 272, 298, 310, 313, 348, 350, 383, 384, 423 excitation, 30, 58, 61, 63, 64, 65, 82, 88, 89, 91, 97, 99, 100, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 118, 177, 199, 200, 219, 242, 309, 409 exclusion, 52, 154, 207, 266 execution, 189, 219, 222, 240, 262 executive function, 207, 396, 408 executive functions, 207, 396, 408 expertise, 139 exposure, xii, 47, 54, 327 extinction, 246, 296 extraction, 162 eye movement, 382, 386 eyelid, 412 eyes, 39, 275, 345
F
facial expression, 283 facial nerve, 342, 344, 347 facial pain, 13, 35, 36, 132, 156, 164, 167, 211 failure, 37, 43, 140, 147, 151, 152, 154, 163, 166, 206, 218, 238, 268, 290, 396, 399, 405, 411 false negative, 279 false positive, 320 family, 283, 309 family members, 283 fat, 43, 72, 93, 153 fatigue, xiii, 27, 155, 264 FDA, 218, 320, 373, 395 FDA approval, 320 FDG, 285, 343, 344, 345 FDI, 239
Index
freedom, 316 freezing, 203, 204, 206, 212, 395 Freud, xii frontal cortex, 55, 189, 197, 198, 296, 297, 417 frontal lobe, 224, 265, 284, 315 functional activation, 198 functional changes, 218 functional imaging, 15, 153, 157, 168, 179, 264, 279, 290, 295, 296, 297, 299, 341, 345, 346, 394, 402 functional magnetic resonance imaging, 32, 44, 158, 168, 169, 183, 196, 285, 352, 406 functional MRI, 11, 12, 22, 199, 244, 269, 270, 345, 360 fusion, 11, 329 FWHM, 170
433
G
GABA, xiii, 49, 56, 134, 191, 197, 233, 253, 417 GABAB, 49 GABAergic, 49, 59, 62, 87, 134, 184, 233, 399 gait, 203, 204, 210, 212, 213, 395, 405 gambling, 299 Gamma, 355 ganglia, 143, 171, 179, 184, 189, 191, 194, 197, 207, 218, 219, 220, 224, 235, 294, 297, 299, 394, 397, 398, 410 ganglion, 366, 367, 368, 371, 382 gauge, 328, 329, 330 GDNF, 190 gene, 50, 134, 397 general anesthesia, 21, 22, 24, 34, 38, 139, 257, 265, 280 generalization, 316 generalized seizures, xii, 42, 155, 310, 399, 401 generalized tonic-clonic seizure, 332, 399 generation, 29, 61, 65, 97, 99, 155, 220, 295, 296, 319, 348, 350, 386, 397, 400, 420 generators, 28 Georgia, 319, 332 Ger, 416 geriatric, 214 Germany, 161, 289, 337, 353 gestures, 154 glass, 383 glial, 375, 382 glial scar, 375, 382 glioma, 149 gliosis, 40, 408 globus, 191, 276, 395, 418, 421
glucose, 190, 195, 198, 290, 303 glucose metabolism, 190, 195, 198, 290, 303 glutamate, 49 glutamatergic, 47, 189, 219 goal-directed, 226 goals, 84, 410 gold, 397 gold standard, 397 government, iv gray matter, 70, 73, 77, 93, 94, 111 Greece, 319, 391 grey matter, 57, 58, 355, 358 grid resolution, 95 Grip strength, 257 groups, 70, 118, 128, 140, 147, 171, 193, 204, 217, 218, 222, 243, 252, 295, 315, 366, 369, 376, 391, 394, 403, 407 growth, 190, 284 growth factor, 190 growth factors, 190 guidance, 12, 14, 15, 20, 39, 157, 168 guidelines, 36, 52, 56, 87, 223, 233, 352, 401, 403, 416 guns, 218 gyri, 10, 11, 14, 23, 24, 26, 40, 70, 81, 87, 91, 93, 279, 327 gyrus, 3, 10, 11, 12, 13, 14, 15, 18, 19, 20, 23, 26, 32, 36, 39, 70, 75, 78, 80, 81, 87, 92, 93, 165, 177, 221, 224, 237, 316, 356, 359, 360, 361, 402, 420
H
H2, 170 HA, 194, 350, 423 habituation, 38, 166, 213, 224, 337, 338, 348, 411 hallucinations, 202, 293, 294, 295, 300, 301, 302, 303, 304, 305, 345, 402, 408, 414, 416 handedness, 225 hands, 154, 234, 240, 259 handwriting, 192, 193 hanging, 39 harmful effects, 50 head injury, 396 head trauma, 143, 144 headache, 27, 51, 144, 309 healing, 41 health, xi, 377 hearing, 51, 338, 342, 343, 347, 359, 361, 410, 413 hearing impairment, 342, 347
434
Index
243, 244, 245, 246, 247, 253, 270, 271, 272, 305, 310, 318, 322, 333, 349, 351, 361, 366, 368, 372, 377, 379, 380, 381, 382, 384, 385, 391, 395, 404, 410, 412, 414, 416, 422 human brain, xii, 29, 53, 55, 56, 57, 58, 62, 190, 200, 226, 253, 270, 272, 384, 385, 395, 410, 412, 416 human cerebral cortex, 31, 32, 65, 70, 227 human rights, 404 human subjects, 60, 65, 198, 378 humans, xii, 47, 53, 54, 55, 56, 58, 61, 66, 67, 110, 135, 138, 167, 169, 175, 178, 194, 219, 225, 227, 244, 299, 317, 320, 334, 338, 360, 373, 381, 385, 412, 423 hybrid, 333, 413 hydrocephalus, 279 hydrolysis, 29 hyperactivity, 184, 238, 253, 295, 297, 341, 353, 354, 356 hyperalgesia, 152 hyperarousal, 296 hypesthesia, 142 hypoperfusion, 170 hypothalamic, 392, 420, 421 hypothalamus, 172, 290, 402 hypothesis, 58, 105, 135, 175, 225, 250, 290, 295, 341, 348, 360, 405 hypoxia, 278 hypoxia-ischemia, 278
hearing loss, 51, 338 heart, 243, 270, 276, 412 heart rate, 276, 412 heartbeat, 310 heat, 52 heaths, 111 heating, 48 height, 11 hematoma, 41, 143, 149, 408 hematomas, 37, 155, 283 hemicraniectomy, 280 hemiparesis, 231, 236, 245, 250, 259, 260, 267, 269, 271, 418 hemiplegia, 260, 261, 262, 263 hemiplegic, 258, 270 hemisphere, 11, 20, 24, 40, 49, 51, 222, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 254, 255, 262, 264, 405 hemodynamic, 179 hemorrhage, 28, 139, 141, 142, 147, 148, 149, 150, 213, 330, 408 hemostasis, 36 hernia, 144 heterogeneity, 135, 191, 295, 320, 348 heterogeneous, 293, 393 heuristic, 340 high resolution, 205, 360 high-frequency, 31, 48, 118, 132, 137, 141, 189, 202, 245, 290, 291, 292, 294, 295, 296, 297, 298, 299, 304, 305, 311, 312, 321, 322, 338, 340, 342, 402 high-speed, 329 high-tech, 218, 407, 414 hippocampal, 326, 327, 328, 331, 332, 411, 422 hippocampus, xiii, 194, 296, 399, 401, 422 histological, 29, 382, 383 homogeneity, 401 hormone, 305 horse, 4 hospital, 39, 53 hospitalized, 321 host, 222 hot spots, 22, 268 House, 373, 382 household, 28 housing, 324 human, xii, xiii, 29, 30, 31, 32, 45, 46, 48, 53, 54, 55, 56, 57, 58, 60, 62, 65, 66, 67, 70, 88, 91, 93, 98, 105, 108, 110, 111, 112, 169, 175, 179, 189, 197, 198, 199, 200, 219, 224, 225, 226, 227, 233,
I
IASP, 25, 27, 159 iatrogenic, 161, 163 IBM, 35, 39 ICMS, 381 id, 173, 202, 204, 260, 315, 331, 405 identification, 14, 15, 19, 30, 67, 110, 173, 208, 347, 378, 385, 407, 418 identity, 421 idiopathic, 167, 195, 196, 198, 207, 396, 398, 421 idiosyncratic, 221, 249 Illinois, 309, 417 imagery, 27, 54, 195, 279 images, 10, 12, 13, 14, 40, 259, 279, 281, 282, 297, 329, 360, 372, 379, 407 imaging, 9, 12, 14, 15, 18, 23, 26, 31, 32, 44, 150, 153, 157, 158, 168, 169, 179, 183, 184, 196, 213, 254, 259, 260, 262, 264, 265, 277, 279, 283, 285, 290, 293, 295, 296, 297, 299, 338, 341, 344, 345,
Index
346, 348, 349, 351, 352, 353, 360, 376, 378, 391, 394, 402, 403, 406, 407, 421 imaging systems, 407 imaging techniques, 341, 345, 376 Immanuel Kant, xi immunity, 371 impairments, 405, 406 implants, xii, 41, 146, 218, 367, 368, 371, 375, 381, 384, 386, 387, 410, 413, 423 implementation, 243, 372, 411 impulsivity, 299 in vitro, 67, 381, 413 in vivo, xi, 67, 194, 385 inactive, 346, 383, 387 inattention, 245 incentive, 409 incidence, 246, 276, 405 inclusion, 52, 207, 212, 213, 268, 298, 358 income, 231 independence, 404 indication, xi, 133, 266, 358, 413 indicators, 276 indices, 98, 295 indirect effect, xiii, 86 individualization, xi, 267, 297 induction, xiii, 24, 46, 50, 165, 233, 243, 245, 247, 255, 280, 350, 368, 375, 386, 394 inductor, 47 industrial, 409 industry, xi, 250, 409 ineffectiveness, 406 inert, 373 infancy, 317 infarction, 32, 142, 147, 148, 250, 271, 272, 290 infection, 28, 38, 40, 144, 155, 163, 166, 202, 206, 212, 252, 330, 408 infections, 28, 41, 146, 251, 369 infectious, 36, 40 inferior frontal gyrus, 11, 221, 237, 247 infinite, 73, 82 inflammation, xiii, 41, 161 inflammatory, 268, 375 inflammatory response, 375 information exchange, 414 information processing, 93, 371 informed consent, 207 ingestion, xiii inherited, 383, 397 inhibition, 49, 54, 55, 56, 62, 65, 66, 118, 134, 137, 138, 175, 177, 184, 189, 191, 192, 193, 196, 197,
435
199, 200, 213, 219, 225, 226, 232, 233, 236, 238, 244, 253, 291, 309, 317, 338, 355, 357, 399, 409 inhibitory, 49, 58, 59, 65, 70, 85, 86, 88, 174, 175, 176, 177, 178, 184, 187, 191, 219, 232, 234, 235, 236, 238, 243, 253, 269, 290, 294, 297, 318, 405 inhibitory effect, 49, 235, 269, 318 inhomogeneity, 174 initiation, 64, 76, 77, 83, 107, 209, 267 injection, 341, 369, 372, 375, 381 injections, 375 injuries, 276, 415 injury, 4, 31, 55, 88, 126, 128, 136, 139, 146, 149, 159, 162, 167, 243, 249, 276, 284, 320, 378, 396, 406 innervation, 344 innovation, 420 insertion, 166, 329, 373, 374, 412 insight, 54, 58, 384 insomnia, 337, 348 inspection, 20, 43 instability, 205, 395 institutions, 22 instruction, 277 insulation, 94 integration, 39, 157, 168, 191, 193, 199, 222, 237, 277, 397, 413 integrity, 267 intensive care unit, 38, 275 interaction, 49, 81, 178, 187, 268, 275, 283, 317 interactions, 49, 198, 200, 224, 226, 233, 246, 366, 406, 420 interface, 28, 29, 74, 75, 188, 320, 367, 369, 380, 383, 384, 407 interference, 28, 81, 296 internalization, 173 Internet, 323, 385, 412, 421 interneuron, 70 interneurons, 49, 63, 64, 70, 71, 79, 106, 174, 175, 176, 177, 224, 410 interval, 19, 38, 48, 49, 58, 63, 66, 146, 173, 191, 205, 235, 239, 240, 252, 267, 276, 405 intervention, 144, 247, 270, 320, 351, 378, 391, 405, 409 intoxication, 359 intracerebral, 26, 212, 350, 392, 393, 408, 409 intracerebral bleed, 26 intracerebral hemorrhage, 212, 392, 393, 408, 409 intracortical microstimulation, 375, 380, 382, 383, 385
436
Index
kinematics, 236, 241, 246 kinetics, 97
intracranial, 4, 31, 46, 51, 67, 110, 166, 320, 321, 322, 333, 354, 356 intramuscular, 30 intraoperative, 17, 21, 22, 35, 44, 140, 150, 152, 154, 155, 158, 165, 168, 174, 217, 266, 268, 329, 330, 332, 394, 407 intravenous, 24, 41, 341 intravenous antibiotics, 41 intrinsic, 105, 109, 110, 176, 188, 211, 221, 249, 284, 348, 400, 411 intrusions, 297 invasive, xi, 27, 40, 45, 83, 167, 175, 196, 265, 291, 310, 316, 326, 327, 360, 376, 391, 393, 398, 402, 403, 404, 405, 406 inversion, 20 ion channels, 76, 98, 107 ionic, 29, 47, 93, 98 ions, 72, 73, 86, 98 ipsilateral, 19, 22, 40, 54, 141, 147, 153, 156, 170, 171, 175, 195, 207, 208, 219, 220, 222, 225, 233, 238, 242, 245, 247, 250, 253, 258, 272, 326, 395 Ireland, 15 iridium, 29, 326, 381 IRP, 7, 8, 9, 10 irrigation, 329 irritation, xii ischaemia, 54 ischemia, 66, 249, 278 ischemic, 203, 212, 213, 250, 268, 269, 271, 418 ischemic stroke, 268, 269, 271, 418 Islam, 197 isotropic, 73 Israel, 290 Italy, 3, 45, 57, 117, 139, 169, 201, 217, 231, 275, 289, 309, 319, 384
L
lack of control, 75 lambda, 7 lamina, 70, 93, 94, 97, 178 laminar, 70, 93, 178 language, xii, 4, 233, 237, 247, 251, 252, 254, 255, 258, 260, 262, 269, 271, 272, 310, 341 language impairment, 254, 260 laptop, 326 large-scale, xi, 277, 394, 403 laser, 135, 138, 173 latency, 58, 59, 60, 61, 64, 67, 105, 106, 107, 173, 197, 199, 219, 321, 323 laterality, 153, 267, 280, 286, 338, 345, 346, 393, 406 late-stage, 188 law, xii, 98, 368 leakage, 27, 98, 372, 407 learning, xiii, 232, 234, 235, 242, 243, 254, 264, 378, 381 learning process, 264 left hemisphere, 11, 204, 208, 237, 254, 260 lesion, 128 lesioning, 40, 357, 401 lesions, 20, 28, 36, 54, 141, 142, 144, 164, 174, 213, 232, 238, 245, 290, 315, 317, 368, 381, 396, 405, 407, 414 levodopa, 185, 187, 188, 194, 195, 196, 199, 202, 206, 207, 215, 394, 395, 396 lidocaine, 329 life expectancy, 401 lifetime, 289 limbic system, 350, 401 limitation, 36, 77, 108, 298, 347, 407 limitations, 84, 108, 204, 205, 286, 376, 377, 393, 395, 397, 398, 400, 403, 405, 406, 407 linear, 13, 26, 80, 103, 236, 329, 378 linear acceleration, 236 linguistic, 278, 377 linguistic processing, 377 linkage, 406 links, 350 lipid, 76, 98 liquids, 283 local anesthesia, 22, 23, 24, 34, 39, 165, 330, 407 local anesthetic, 328
J
Japan, 17 job dissatisfaction, 154 joints, 221, 267 joystick, 203 judgment, 265
K
K+, 142 Kant, xi ketamine, 142, 146, 153
Index
local area network, 329 localization, vii, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 17, 18, 19, 22, 23, 30, 32, 48, 140, 150, 171, 238, 291, 315, 321, 326, 344, 345, 346, 357, 359, 394 location, 10, 11, 12, 13, 14, 17, 18, 20, 21, 22, 24, 36, 40, 48, 69, 71, 77, 84, 96, 97, 98, 99, 101, 103, 104, 109, 133, 179, 198, 245, 264, 317, 324, 326, 359, 367, 373, 378, 379, 403, 405 locomotion, 259, 266 locus, 134, 170 London, 179, 197, 334, 385 long distance, 263 long period, 49, 190, 212 longevity, 92 long-term potentiation, 50, 88, 190, 198, 232 loss of consciousness, xii low back pain, 138 low power, 280, 370 low-intensity, 47, 62, 193 low-tech, 218 LSI, 413 LTD, 62, 232, 348 LTP, 62, 63, 194, 232, 234, 243 lubrication, 111 luminosity, 381 lungs, 24 lying, 3, 41, 258, 276
437
M
M1, 119, 120, 121, 122, 124, 126, 127, 128, 240, 241, 242, 297 machine learning, 319 machines, 409 macrophages, 372 magnesium, 86 magnet, 323 magnetic, xii, 10, 15, 18, 31, 39, 45, 46, 47, 48, 49, 50, 52, 53, 54, 55, 56, 58, 60, 61, 63, 64, 65, 66, 67, 118, 135, 136, 137, 157, 158, 168, 184, 191, 194, 195, 196, 226, 234, 243, 245, 246, 272, 285, 295, 296, 298, 300, 301, 302, 303, 304, 305, 318, 333, 346, 348, 350, 353, 360, 376, 380, 409, 418, 420 magnetic field, xii, 39, 46, 47, 48, 52, 346 magnetic field effect, 346 magnetic resonance, 10, 15, 18, 31, 32, 44, 158, 168, 169, 183, 191, 196, 285, 296, 352, 376, 406 magnetic resonance imaging, 15, 18, 31, 32, 44, 158, 168, 169, 183, 196, 285, 296, 352, 361, 406
magnetic resonance spectroscopy, 191, 296 magnetoencephalography, 12 magnets, 28 maintain balance, 266 maintenance, 117, 289, 291, 304, 409 major depression, 154, 165, 199, 289, 291, 299, 300, 301, 303, 304, 305, 415, 417, 419, 420, 421 major depressive disorder, 54, 304 maladaptive, 190, 191, 192, 232, 237, 338, 398 males, 358 malignant, 156, 167 mammals, 368 management, 44, 136, 139, 140, 156, 167, 178, 283, 286, 323, 367, 392, 393, 397, 409, 414, 420, 423 mandibular, 18 mania, 292, 300, 301, 394 manic, 292 manipulation, 59 manufacturer, 218, 330 manufacturing, 380 mapping, 22, 26, 34, 35, 39, 44, 150, 152, 155, 158, 168, 208, 225, 253, 264, 265, 311, 320, 368, 372, 377, 380, 384, 385, 387, 403, 407, 409 market, xiii marketing, 218 mask, 85, 188 masking, 337 MAST, 258 mastoid, 6, 18, 42, 330 matrices, 202 matrix, 95, 339, 372 MCA, 261 MCC, 172 MCS, 27, 35, 38, 124, 141, 142, 143, 144, 145, 146, 147, 149, 150, 151, 163, 164, 171, 177, 275, 276, 277, 278, 279, 316 measurement, 22, 184, 197, 257, 323, 324, 345, 347, 350 measures, xiii, 22, 41, 83, 119, 184, 188, 192, 193, 203, 207, 234, 235, 239, 271, 344, 349 media, 71, 72 medial prefrontal cortex, 291, 296 median, 19, 20, 21, 22, 36, 208, 222, 236, 313, 321, 332 mediators, 220 medication, 52, 150, 166, 188, 189, 204, 206, 207, 219, 223, 291, 299, 301, 302, 309, 312, 314, 394, 395, 396, 399, 401, 404 medications, 150, 156, 202, 205, 207, 289, 291, 395, 404
438
Index
mobility, 205, 259, 381 MOD, 280 modalities, 291, 377, 419 modality, 135, 409 modeling, 77, 89, 93, 95, 97, 103, 104, 105, 107, 108, 109, 111, 179 models, 76, 88, 89, 91, 97, 98, 99, 100, 102, 105, 107, 111, 169, 233, 373, 376, 399, 411, 416 moderators, 302 modulation, xi, xiii, 47, 50, 53, 56, 174, 183, 189, 193, 197, 198, 200, 213, 214, 218, 219, 226, 232, 238, 239, 268, 295, 296, 300, 309, 312, 344, 349, 360, 368, 371, 379, 384, 410 modules, 277, 369, 371 mold, 253 molecules, 375 monkeys, 58, 178, 179, 225, 250 monoamine, 195 monotherapy, 401 mood, 134, 173, 202, 204, 207, 221, 223, 260, 290, 305, 309, 344, 402, 403 mood change, 134, 204, 309, 344 mood disorder, 202 morning, 38, 332 morphine, 141, 142, 146, 148, 153, 165, 172 morphological, 87, 232 mosaic, 221, 380 motion, 23, 41, 53, 250, 252, 379, 387, 407 motor activity, 275 motor area, 22, 30, 32, 51, 61, 132, 183, 184, 185, 187, 192, 196, 197, 199, 202, 209, 219, 227, 232, 233, 236, 239, 254, 255, 258, 279, 297, 298, 395, 398, 402, 405, 418 motor behavior, 51, 222, 234, 242 motor control, 175, 222, 226 motor fiber, 175, 260, 261 motor function, 22, 69, 141, 152, 153, 156, 167, 187, 188, 190, 196, 197, 232, 233, 235, 236, 237, 240, 241, 242, 244, 245, 246, 255, 258, 261, 262, 264, 269, 272, 316, 397 motor neurons, 237, 253 motor strip, 13, 19, 34, 36, 41, 163, 166, 263 motor system, 50, 184, 235, 254 motor task, 23, 34, 203, 221, 234, 240, 250, 262, 264, 268, 409 mouth, 186 movement, viii, xiii, 24, 26, 30, 32, 34, 35, 38, 40, 44, 69, 85, 87, 110, 152, 156, 157, 167, 175, 183, 185, 186, 189, 191, 195, 196, 197, 201, 202, 203, 209, 211, 212, 213, 214, 215, 218, 219, 221, 222,
medicine, xiii, 121, 214 medulla, 144 MEG, 12, 39, 88, 153, 338, 359 membranes, 109, 417 memory, xiii, 55, 56, 190, 194, 198, 207, 208, 232, 233, 255, 279, 310, 322, 326, 328, 380, 402, 420 memory deficits, xiii memory loss, 310 memory performance, 380 MEMS, 413 men, 53, 289 meninges, 46 mental disorder, 377 mental image, 278 mental imagery, 278 mental representation, 175 mesencephalon, 170 messengers, 134, 220 meta-analysis, 132, 137, 196, 290, 302, 305, 318, 393, 394, 414, 417, 418, 420, 423 metabolic, 161, 169, 174, 193, 195, 259, 276, 277, 278, 300, 344, 345, 348, 372, 394 metabolic dysfunction, 278 metabolic rate, 276, 277 metabolism, 190, 195, 198, 211, 277, 283, 290, 291, 296, 303, 345 metabolites, 195 metals, 29 microchip, 327, 411 microelectrode, 139, 369, 370, 372, 373, 375, 378, 382, 383, 384, 385, 387, 412 microelectrodes, 73, 366, 369, 370, 373, 374, 375, 383, 386, 387, 413 microelectronics, 413, 414 microfabrication, 413 microhemorrhages, 374 microscope, 89, 176 Microsoft, 320 midbrain, 142, 144, 173, 276 migraine, 117, 121 migration, 27, 265, 394 military, xiii mimicking, 371 Minnesota, 272 minority, 28, 59 mirror, 23 misconception, 81, 85 misleading, 133 Mississippi, 258 MMSE, 202, 207, 208
Index
224, 225, 226, 234, 236, 245, 246, 250, 252, 257, 262, 266, 269, 272, 284, 371, 397, 398, 403, 404, 407, 411, 412, 413, 414, 415, 417, 423 movement disorders, viii, xiii, 24, 26, 30, 34, 35, 38, 40, 44, 69, 85, 87, 110, 152, 156, 157, 167, 175, 183, 195, 201, 211, 212, 213, 214, 218, 224, 226, 250, 269, 397, 398, 403, 404, 407, 414, 415, 417, 423 MRI, 10, 11, 12, 14, 15, 18, 19, 22, 31, 39, 41, 43, 44, 48, 52, 55, 100, 103, 148, 169, 183, 199, 203, 204, 212, 244, 259, 260, 262, 265, 269, 270, 283, 285, 286, 296, 329, 345, 352, 360, 376, 395, 399, 406, 409, 412, 419 MRS, 56, 283 MSI, 353 multidisciplinary, 194, 407 multiple sclerosis, xiii, 139, 396, 419, 423 Multi-Society Task Force, 276 muscle, xii, 17, 20, 21, 22, 24, 27, 43, 50, 53, 55, 57, 58, 59, 65, 67, 109, 132, 142, 154, 156, 179, 184, 190, 191, 195, 219, 222, 265, 266, 291 muscle contraction, 142, 154 muscle power, 266 muscle relaxant, 24 muscles, 48, 51, 59, 211, 219, 220, 221, 231, 237, 266, 346, 367, 411 musculoskeletal, 137 musculoskeletal pain, 137 music, 341, 354 mutations, 210 myelin, 98, 111 myelination, 91, 93, 279 myocardial infarction, 296, 305 myoclonus, 210, 315, 318, 400
439
N
NAA, 283 naloxone, 134, 153 naming, 237, 246, 422 nanofibers, 413, 423 nanomaterials, 413 nanotechnologies, 414 narcotic, 150 natural, 13, 94, 175, 211, 271, 272, 405, 410, 412 nausea, 393 navigation system, 12, 17, 23 neck, 12, 33, 41, 43, 192, 193, 210, 346, 399, 421 negativity, 277, 278 neglect, 238, 244, 246, 405
neocortex, 111, 178, 317, 326, 382 nerve, 19, 20, 21, 22, 30, 36, 46, 65, 87, 88, 89, 92, 93, 94, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 108, 109, 110, 111, 112, 120, 122, 124, 128, 161, 162, 165, 167, 177, 208, 222, 234, 239, 290, 320, 346, 349, 366, 368, 382, 404 nerve cells, 96, 97, 102, 103, 104, 105, 109 nerve fibers, 89, 93, 94, 97, 102, 104, 105, 108, 109, 177, 368 nerve trunk, 124, 128 nerves, 36, 161, 220 nervous system, 20, 232, 351, 361, 367, 369, 372, 384, 391, 410, 413 Netherlands, 91 network, 93, 172, 189, 191, 192, 231, 237, 253, 254, 263, 264, 267, 272, 279, 290, 295, 296, 297, 310, 316, 329, 338, 348, 351, 360, 369, 395, 411, 413, 419 networking, 253, 376 neural connection, 94 neural function, xiii, 411 neural mechanisms, 224, 277 neural network, xiii, 117, 279, 369, 391, 392, 421 neural networks, xiii, 117, 391, 392, 421 neural regeneration, 406 neural tissue, 32, 47, 48, 95, 111, 177, 324, 369, 372, 375, 383, 407 neuralgia, 121, 158, 162, 163, 164, 168, 392 neuroanatomy, 350 neurobehavioral, 280 neurobiological, 289, 295, 296, 349 neurobiology, 414 neurochemistry, 135 neurogenesis, 253, 270 neurogenic, 34, 38, 40, 43, 131, 133, 136, 139, 174, 392 neuroimaging, xi, xiii, 10, 153, 169, 174, 194, 220, 232, 259, 269, 278, 279, 290, 302, 344, 345, 347, 394, 395, 401, 406, 407, 420 neuroimaging techniques, 169, 345, 347 neurologic disorders, 416 neurological condition, 409 neurological deficit, 26, 254, 260, 266, 268, 310 neurological disease, 257 neurological disorder, 65, 285, 309, 409, 413 neurological rehabilitation, 386 neurologist, 204, 218, 407 neuroma, 162, 167 neuromodulation, xi, 47, 52, 357, 360, 391, 398, 410, 414, 416, 419, 421
440
Index
noise, 50, 337, 338, 340, 341, 353, 354, 356, 358, 371, 373, 386 non invasive, 377 non-human, 226, 378, 419 non-human primates, 226 non-invasive, ix, 45, 51, 52, 56, 232, 243, 289, 291, 354, 360, 377, 398, 403 non-linearity, 76 noradrenaline, 134 normal, xii, 31, 43, 52, 55, 72, 143, 144, 153, 175, 184, 190, 191, 192, 197, 198, 199, 225, 231, 232, 233, 238, 254, 268, 278, 296, 310, 348, 357, 366, 374, 399, 410, 411, 419 normal children, 52 normal conditions, 310 normalization, 191, 193, 254, 279, 283, 297 North America, 421 nuclei, xi, 93, 94, 109, 220, 276, 355, 391, 392, 394, 399, 401, 403, 405, 420, 422 nucleus, 88, 109, 190, 200, 205, 214, 220, 224, 225, 226, 299, 321, 353, 366, 371, 394, 396, 397, 399, 401, 414, 415, 416, 417, 418, 419, 420, 421, 422 nucleus accumbens, 401, 419, 422 nurse, 165 nurses, 38, 165, 407
neuronal death, 66 neuronal excitability, 383 neuronavigation, 3, 12, 14, 23, 25, 34, 39, 44, 132, 150, 151, 157, 158, 165, 168, 265, 340, 342, 343, 345, 346, 357, 359, 407 neurons, 29, 30, 46, 47, 48, 49, 53, 59, 60, 61, 62, 63, 64, 65, 69, 70, 71, 79, 80, 81, 84, 85, 87, 88, 89, 97, 99, 103, 104, 105, 106, 108, 109, 111, 169, 175, 179, 189, 190, 219, 220, 222, 225, 249, 276, 310, 341, 355, 366, 367, 368, 369, 370, 371, 373, 374, 375, 383, 385, 386, 399, 410, 413 neuron-specific enolase, 52 neuropathic pain, viii, 15, 30, 31, 43, 44, 110, 117, 118, 119, 122, 132, 134, 136, 137, 156, 157, 158, 159, 161, 162, 163, 164, 167, 168, 170, 173, 174, 178, 179, 392, 393, 413, 416, 420 neuropathy, 35, 128, 392 neuropeptides, 195 neurophysiology, 22, 45, 137, 169 neuroplasticity, 277, 351, 381, 382, 383, 384, 405, 414, 415 neuroprostheses, 384 neuroprotective, xiii, 62, 67, 211, 396, 414 neuropsychiatric disorders, 402, 403, 404, 409, 423 neuropsychological assessment, 202 neuroscience, 371, 421 neuroses, xii neurosurgeons, xi, 17, 18, 21, 23, 140, 217, 218, 220, 407 neurosurgery, xi, 45, 61, 194, 226, 353, 375, 414 neurotoxicity, 167 neurotransmission, 185, 294 neurotransmitter, 253, 277 neurotransmitters, 50, 134, 185 neurotrophic, 67, 190 neurotrophic factors, 67 New Orleans, 418 New York, 15, 44, 87, 88, 89, 136, 156, 157, 178, 195, 197, 245, 384, 414, 415, 418, 420, 421, 423 nicotine, 299 nigrostriatal, 414 NIH, 234, 420 NMDA, 49, 86, 134, 190 NMDA receptors, 86, 190 noble metals, 29 nociceptive, 134, 144, 161, 167, 173, 175, 392 node of Ranvier, 98, 103, 111 nodes, 84, 98, 103, 106, 107, 391, 398, 400, 402, 403, 404
O
object recognition, 371 observations, 38, 72, 107, 135, 156, 167, 185, 206, 254, 255, 289, 315, 366, 403 obsessive-compulsive, 297, 422 obsessive-compulsive disorder (OCD), 299, 301, 303, 305, 401, 402, 403, 416, 418, 419, 420 obstruction, 26 occipital cortex, 340, 342, 343, 369, 377, 380, 381 occupational, 266, 296 oil, 339, 340, 342 older adults, 351 onset latency, 60 operator, 28, 47, 347 opioid, 146, 173, 179, 392 opioids, 172, 175 opposition, xi, 22 optic nerve, 366, 367, 371, 381, 386 optimization, 30, 110, 178, 188, 203, 297, 326, 346, 394 oral, 165, 207, 283 oral antibiotic, 165 orbit, 18
Index
orbitofrontal cortex, 170, 171, 290, 297, 338, 403 orientation, 20, 24, 35, 40, 47, 59, 60, 64, 65, 67, 91, 93, 97, 104, 105, 165, 177, 371, 385 orthopaedic, 423 oscillation, 219 oscillations, 31, 219, 225, 338 oscillatory activity, 134, 224, 225, 226 osteoarthritis, 392 osteomyelitis, 117, 120 outpatients, 155 overload, 278 oxide, 381 oxygenation, 22
441
P
P300, 278 PACE, 179 pacemaker, 12 pacemakers, 52, 310 pain management, 140, 167, 367, 393, 414 pain reduction, 22, 26, 121, 122, 126, 127, 128, 133, 140, 141, 146, 151, 164, 166 pain score, 172 pain therapy, 393 pancreatitis, 122, 132 panic attack, 296, 305 panic disorder, 302, 305, 402, 419 parallel processing, 194 parameter, vii, 42, 69, 91, 298, 348, 360, 381, 415 parenchymal, 408 paresis, 235, 249, 252, 271 paresthesias, 22, 140, 141, 142, 147, 154, 165 parietal, 174 parietal cortex, xii, 38, 42, 54, 132, 133, 138, 153, 170, 178, 224, 238, 245, 246, 258, 262, 272, 279, 339, 398, 411 parietal lobe, xii, 142, 143, 238, 253, 263, 265 parietal lobes, xii, 265 Paris, 14, 15, 55, 143, 218 Parkinson, viii, xii, 20, 32, 45, 136, 174, 183, 185, 187, 190, 194, 195, 196, 197, 198, 199, 200, 201, 207, 213, 214, 215, 217, 218, 224, 225, 226, 283, 284, 367, 394, 396, 411, 414, 415, 416, 417, 420, 421, 422, 423 Parkinson disease, xii, 197, 198, 207, 213 Parkinsonian symptoms, 198 parkinsonism, 184, 202, 214, 395, 396 Parkinsonism, 30, 199, 201, 202, 203, 215, 225, 422 PART, 249, 257
partial seizure, 309, 311, 312, 400 particles, 72 passive, 98, 99, 107, 175, 222, 278 pathogenesis, 296, 397 pathology, 52, 279 pathophysiology, xiii, 45, 174, 183, 196, 224, 225, 238, 295, 338, 347, 348, 398, 402, 422 pathways, 60, 105, 167, 184, 219, 220, 232, 250, 253, 258, 337, 353, 354, 356, 365, 367, 368, 371, 376, 381, 384, 394, 405, 406, 410 pattern recognition, 367 patterning, 413 Pavlovian, 271 Pavlovian conditioning, 271 PCS, 93, 146 pediatric, 52, 236, 317 pediatric patients, 236 PEK, 380 pelvis, 13 perception, 43, 53, 137, 186, 187, 197, 270, 276, 277, 278, 338, 340, 341, 344, 345, 347, 350, 351, 352, 359, 360, 361, 366, 367, 368, 378, 379, 382, 392 perceptions, 365, 368, 378, 379 perceptual learning, 368 perforation, 43 perfusion, 196, 205, 209, 214, 224 periodic, 289 periodicity, 61, 63, 66, 67, 247 peripheral, 161, 234, 239, 266 peripheral nerve, 65, 122, 162, 344, 346, 404 peripheral nervous system, 85, 161, 190, 395 periventricular, 392 permit, 174 personal communication, 26, 39, 140, 163, 167, 213, 267, 268, 406, 414 personal computers, 320 personality, 154, 350 personality disorder, 154 personality traits, 350 perturbations, 316 PET, 12, 54, 153, 158, 169, 170, 174, 178, 179, 183, 189, 192, 193, 197, 200, 203, 209, 210, 219, 225, 226, 237, 263, 272, 276, 277, 278, 285, 291, 296, 299, 338, 340, 341, 342, 343, 344, 345, 348, 350, 351, 354, 407 PET scan, 170, 172, 192 PFC, 173, 294 PG, 54, 251, 304, 350, 420 phantom limb pain, 138, 158, 162, 168, 179, 392
442
Index
pools, 221 poor, 22, 140, 141, 143, 144, 163, 171, 175, 210, 267, 291, 312, 337, 368, 378, 379, 396 population, 52, 84, 85, 86, 140, 212, 234, 259, 293, 309, 337, 396, 401, 416, 421 portability, 409 positive feedback, 355 positron, 12, 169, 198, 226, 244, 247, 296, 351 positron emission tomography, 12, 169, 198, 226, 244, 247, 296, 351 posterior cingulated, 172 postoperative, 144, 165, 166, 173, 208, 211, 332, 344, 394, 397, 403, 408 poststroke, 121, 158, 211, 214, 237, 247, 269, 303, 380 post-stroke, 28, 32, 45, 112, 153, 156, 157, 157, 159, 178, 201, 211, 214, 233, 236, 237, 238, 239, 250, 266, 267, 270, 397, 405, 417, 418, 420 postsynaptic, 86, 208, 234 posttraumatic stress, 300, 301, 303, 304 posttraumatic stress disorder, 300, 301, 303, 304 postural instability, 205, 395 posture, 203, 205, 210, 212, 251, 395 potassium, 98 power, 28, 43, 83, 87, 110, 174, 185, 266, 278, 280, 284, 311, 312, 314, 321, 323, 324, 357, 369, 370, 412 PPS, 66 preconditioning, 243 prediction, 77, 108, 124, 271, 319, 320, 333, 403, 411, 413 predictors, 31, 158, 291, 300, 343, 347 preference, 405 prefrontal cortex, xiii, 55, 119, 121, 124, 127, 132, 173, 183, 185, 195, 199, 200, 203, 205, 209, 258, 279, 280, 289, 291, 292, 294, 295, 296, 297, 298, 300, 301, 302, 304, 305, 343, 402, 415 pregnancy, 52, 55 pregnant, 52 pregnant women, 52 premotor cortex, 3, 93, 133, 140, 172, 184, 187, 192, 193, 194, 196, 197, 198, 200, 203, 205, 217, 219, 235, 244, 245, 251, 253, 258, 259, 261, 262, 263, 268, 291, 398, 419 pressure, 151, 193, 276, 330, 359, 368 prestige, 218 presynaptic, 86, 189 prevention, 134, 316 primary care, 231 primary visual cortex, 367, 371, 375, 378, 387
pharmaceutical, 297 pharmacological, 47, 59, 402, 414 pharmacological treatment, 402, 414 pharmacotherapy, 304 phenotypes, 398 phenytoin, 166 philosophical, 218 phonological, 208 photon, 169, 183 photoreceptor, 380, 383 photoreceptors, 371, 372, 387 phrenology, 169 physical world, 366, 368 physicians, 38, 165, 286, 323, 403 physics, 45 physiological, xii, 36, 55, 57, 58, 223, 232, 243, 277, 367, 377, 378, 412, 422 physiology, 225 physiotherapists, 283 physiotherapy, 266 pia mater, 93, 344 pilot studies, 270, 372 pilot study, 137, 197, 215, 291, 300, 303, 304, 318, 338, 350, 351, 352, 361, 416, 421, 422 pitch, 354, 356, 410 placebo, 38, 118, 119, 133, 135, 146, 151, 154, 165, 166, 167, 193, 196, 212, 236, 239, 240, 245, 290, 295, 296, 298, 299, 300, 301, 303, 305, 340, 346, 350, 351, 356, 359, 360, 393, 404 planar, 407, 413 planning, 3, 9, 34, 167, 262, 264, 329, 377 plasma, 62, 65, 189, 194 plasma levels, 62, 65, 189, 194 plastic, 36, 42, 46, 191, 194, 247, 272, 338, 358, 368, 378, 384 plasticity, 62, 67, 190, 191, 192, 195, 196, 199, 220, 232, 233, 243, 245, 246, 247, 253, 254, 255, 264, 266, 268, 270, 271, 272, 350, 351, 358, 360, 368, 377, 379, 380, 381, 382, 385, 398, 417 platinum, 29, 326 platysma, 43 play, 3, 109, 134, 190, 238, 253, 254, 267, 338, 411, 413 plexus, 23, 122, 124, 128, 161, 162, 164, 172 polarity, 24, 30, 69, 81, 84, 85, 95, 103, 104, 156, 375 polarization, 28, 29, 47, 53, 54, 56, 74, 89, 97, 98, 103, 107, 111, 220, 386 polymorphisms, 300 pons, 172
Index
primate, 89, 179, 380, 419 primates, 250, 272, 368, 380 priming, 236, 243, 291, 345, 350, 404 probability, 23, 374 probands, 296 probe, 36, 329, 383, 385, 420 production, 173 prognosis, 312 prognostic factors, 213 prognostic marker, 140 prognostic value, 269 program, 165, 323, 358, 377, 412 programmability, 394 programming, 41, 140, 152, 154, 204, 310, 357, 358, 397, 398, 408, 409, 412, 419 progressive supranuclear palsy, 396 proliferation, 375 propagation, 94, 98, 107, 109, 399, 422 property, 232 prophylactic, 41, 165 propofol, 119, 120, 134, 145, 146, 148, 149, 153, 165, 175 prostheses, 89, 333, 366, 369, 386, 417 prosthesis, 365, 366, 367, 368, 369, 370, 377, 379, 381, 382, 384, 385, 386, 406, 410 prosthetics, 381 proteases, 372 protection, 54, 66, 166 proteins, 47 protocol, 29, 33, 38, 52, 62, 63, 65, 118, 165, 167, 174, 188, 209, 237, 238, 246, 268, 295, 296, 311, 312, 313, 345, 376 protocols, 47, 49, 50, 52, 65, 135, 153, 191, 234, 239, 243, 249, 262, 290, 293, 310, 345, 348, 358, 393, 397, 401, 406, 407, 409 psychiatric diagnosis, 403 psychiatric disorder, ix, 45, 165, 295, 391, 403, 406, 413, 416, 419, 421 psychiatric disorders, ix, 45, 165, 295, 391, 403, 413, 416, 419, 421 psychiatric illness, 293, 404 psychiatric patients, 402, 403 psychiatrists, xii psychologist, 165 psychosis, 165, 394 psychosocial stress, 154 psychosurgery, 298, 402, 403 psychotherapy, 296, 297 PTSD, 296 public, 310
443
public health, 310 pulse, 24, 28, 29, 30, 34, 38, 43, 46, 47, 48, 49, 50, 56, 58, 60, 62, 63, 65, 66, 67, 69, 73, 75, 82, 83, 85, 86, 87, 88, 91, 92, 97, 99, 100, 102, 103, 105, 107, 108, 109, 154, 155, 184, 185, 186, 187, 192, 200, 204, 243, 245, 247, 251, 263, 280, 312, 313, 318, 320, 321, 324, 334, 340, 357, 375, 379, 400 pulses, xii, 29, 36, 38, 47, 48, 49, 56, 60, 62, 85, 92, 108, 118, 119, 120, 121, 122, 124, 126, 127, 128, 132, 135, 175, 185, 188, 191, 192, 209, 232, 235, 238, 239, 240, 241, 243, 290, 314, 315, 316, 324, 340, 343, 344, 345, 372, 375, 379 pumps, 52, 392 PVS, 276, 277, 283 pyramidal, 49, 57, 60, 61, 64, 65, 67, 70, 71, 84, 87, 88, 93, 94, 96, 97, 98, 99, 102, 105, 106, 107, 108, 109, 110, 111, 176, 177, 178, 219, 222, 267, 383 pyramidal cells, 70, 84, 93, 94, 96, 97, 102, 105, 106, 109, 176, 177, 178, 219
Q
quality of life, 140, 202, 269, 295, 309, 337, 378, 379, 394, 395, 396, 404, 419 questionnaire, 203, 343 questionnaires, 337, 347
R
radiation, 278 radio, 28 radiofrequency, 372 radiological, 401 Raman, 196 randomness, 378 range, xi, 13, 14, 28, 46, 59, 70, 82, 100, 101, 102, 108, 122, 151, 155, 165, 175, 190, 202, 204, 219, 222, 231, 238, 240, 241, 252, 268, 286, 311, 321, 366, 410 rash, 309 rat, 88, 382, 383, 386 rating scale, 185, 192, 280, 403 ratings, 118, 132, 133, 135, 171 rats, 67, 243 reaction time, 185, 186, 234, 235, 239 reading, 218, 255, 258, 372, 380, 381 real time, 39, 321, 322, 370 reality, xiii, 71, 402
444
Index
resistivity, 46, 88 resolution, 39, 95, 170, 171, 205, 232, 278, 319, 353, 360, 361, 372, 379, 384, 386, 387, 403 resources, 268 respiration, 367, 412 respiratory, 276 response time, 372 responsiveness, 173, 195, 275, 280, 400 resting potential, 76, 86 retina, 366, 368, 369, 370, 371, 381, 383, 384, 385 retinal disease, 366 retinitis, 381 retinitis pigmentosa, 381 retirement, 218 returns, 394 Rhode Island, 3 rhythm, 62 rhythmicity, 313 rhythms, xii, 213, 310 right hemisphere, 11, 35, 208, 222, 232, 237, 238, 263, 279, 285 right visual field, 368 rigidity, 184, 186, 189, 195, 203, 204, 206, 212, 217, 394, 395, 397 rings, 327, 330 risk, 24, 26, 28, 38, 40, 41, 51, 52, 85, 144, 212, 233, 289, 291, 293, 322, 329, 366, 369, 376, 392, 394, 396, 408, 409 risks, 324, 369, 372, 377, 408 robotic, 412 ROI, 174 Rome, 57, 137, 231
reasoning, 207, 415 receptive field, 372, 386 receptors, 46, 49, 86, 134, 173, 190, 208, 357, 369 recognition, 10, 277, 285, 333, 367, 371, 378, 386, 392 recovery, viii, 45, 167, 211, 232, 233, 234, 235, 236, 237, 238, 239, 243, 244, 246, 247, 249, 250, 251, 252, 253, 254, 258, 259, 262, 263, 264, 267, 269, 270, 271, 272, 275, 276, 278, 279, 280, 283, 284, 285, 286, 378, 404, 405, 417 recovery processes, 249, 405 recreational, xiii recruiting, 237 recurrence, 210, 312 redistribution, 316 redness, 52 referees, 218 reflexes, 62, 174, 193 refractoriness, 49 refractory, ix, 31, 59, 67, 108, 122, 133, 139, 158, 175, 290, 291, 297, 301, 303, 309, 311, 312, 313, 317, 318, 337, 353, 359, 392, 394, 397, 401, 403, 406, 413, 416, 418, 420, 421, 423 regeneration, 405, 406 regional, 121, 159, 168, 170, 177, 184, 190, 195, 197, 209, 226, 278, 296, 305 regrowth, 283, 284, 286 regular, 98, 155 regulation, 134, 173, 284 rehabilitation, viii, xii, 45, 150, 218, 231, 234, 237, 242, 243, 249, 250, 251, 252, 257, 260, 261, 263, 266, 267, 268, 269, 270, 271, 284, 368, 376, 378, 380, 404, 405, 406 rehabilitation program, 404 rejection, 376 relapse, 145, 148, 149, 315 relationship, 3, 7, 72, 82, 83, 84, 142, 154, 184, 193, 235, 242, 264, 279, 295, 343, 365, 379 relationships, 15, 31, 72, 84, 191, 371 relaxation, 95 relevance, 291, 297, 333, 338, 347, 352, 381, 382 reliability, 13, 280 remission, 41 remodeling, 190, 383 renormalization, 172, 174, 279 repair, 382, 414 repetitive behavior, 297 replication, 193, 402 resistance, 43, 46, 76, 87, 98, 99, 142 resistive, 98
S
saccadic eye movement, 386 safeguard, 404 safety, 29, 51, 52, 53, 54, 55, 56, 213, 236, 251, 269, 271, 291, 302, 304, 321, 324, 332, 352, 358, 370, 372, 409, 410, 420, 421 saline, 42, 50 sample, 19, 135, 166, 188, 208, 270, 299, 302, 323, 339, 343, 348 satisfaction, 43 saturation, 354 scaling, 99 scalp, 4, 5, 15, 18, 20, 21, 23, 25, 43, 46, 47, 48, 51, 52, 56, 66, 133, 174, 237, 291, 328, 329, 330, 339, 341, 355, 366, 391 scanning electron, 374
Index
scar tissue, 28, 36, 40, 408 schema, 98 Schiff, 285 schizophrenia, 293, 294, 295, 300, 301, 302, 303, 304, 305, 414, 420 schizophrenic patients, 295, 300, 304, 402 school, 40 scientific knowledge, 404 scleroderma, 162, 167 sclerosis, 397, 419 scores, 140, 150, 172, 186, 187, 189, 190, 202, 204, 205, 206, 223, 234, 239, 252, 259, 264, 396, 397 SCS, 142, 143 search, xi, 33, 91, 140, 155, 217, 223 searching, 266 Seattle, 78, 159, 218, 320 secretion, 173 sedation, 51, 279, 329 segregation, 277 seizure, xii, 38, 41, 152, 155, 156, 166, 233, 251, 266, 309, 310, 312, 313, 314, 315, 316, 317, 318, 319, 320, 322, 323, 326, 328, 332, 333, 334, 359, 399, 400, 401, 411, 422 seizure-free, 320, 360 seizures, xii, 21, 28, 30, 40, 41, 51, 141, 142, 147, 152, 155, 166, 167, 233, 235, 266, 309, 310, 311, 312, 314, 315, 316, 317, 318, 319, 320, 321, 322, 324, 332, 333, 359, 386, 393, 399, 400, 407, 408, 416, 422 selecting, 69, 83, 393, 403 selectivity, 24, 30, 83, 84, 85, 87 self, 197, 285 self-awareness, 277, 278, 279 self-care, 252, 259 self-consciousness, 283 self-management, 283 self-report, 337 semantic, 208, 255, 277 semi-permeable membrane, 98 sensation, 13, 38, 41, 42, 43, 69, 110, 150, 151, 154, 175, 202, 344, 352, 368 sensations, 20, 38, 42, 43, 50, 51, 52, 144, 147, 151, 156, 321, 346, 347, 368, 377, 380, 386 sensing, 323, 370 sensitivity, 36, 39, 59, 142, 200, 320, 323, 371 sensorimotor cortex, 31, 32, 35, 55, 191, 192, 203, 224, 258, 262 sensorimotor system, 32 sensors, 236, 412 sensory cortices, 179, 357, 358, 375
445
sensory experience, 279 sensory modalities, 377 sensory systems, 175 separation, 26 sequelae, 396, 399, 412 series, 24, 26, 28, 36, 48, 49, 57, 76, 118, 132, 134, 136, 140, 150, 152, 153, 156, 163, 164, 165, 167, 202, 210, 213, 266, 276, 341, 349, 371, 399, 401, 403, 405, 414 serotonin, 134 serum, 52, 67, 185 severity, 20, 83, 222, 266, 271, 332, 344, 345, 347, 350, 392, 405 sex, 153, 231, 267 shape, 11, 48, 66, 70, 71, 73, 76, 77, 97, 104, 325, 371, 379 shock, 234 shocks, 19, 46, 233, 239 short period, 38, 133, 393 short term memory, 207 short-term, 66, 145, 148, 164, 196, 200, 244, 254, 291, 323, 324, 342, 385 shoulder, 210, 221 side effects, 27, 52, 69, 83, 92, 150, 266, 309, 310, 321, 359, 396, 398 sigmoid, 14, 280 sign, 19, 80, 81, 140, 153, 202, 267, 275, 283 signals, 19, 39, 70, 73, 85, 109, 211, 222, 236, 338, 357, 366, 369, 370, 411, 412 signal-to-noise ratio, 373 signs, 267, 275, 276, 283, 293, 311, 316, 400 SII, 149 silicon, 373, 380, 382, 383, 386 silk, 27, 36, 165, 330 similarity, 108, 189 simulation, 81, 89 sinus, 14, 209 sites, 43, 50, 52, 60, 64, 103, 107, 133, 171, 217, 313, 398, 406, 410 skill acquisition, 234, 245 skills, xiii skin, 12, 26, 33, 34, 37, 41, 42, 46, 50, 52, 276, 280, 327, 330, 332, 412 skin conductance, 276 sleep, xiii, 51, 150, 202, 221, 232, 279 slow-wave, 312 SMA, 170, 172, 183, 185, 186, 187, 192, 194, 203, 209, 211, 219, 224, 279, 298 smoke, xiii smoking, 299, 300, 301
446
Index
statistical analysis, 203, 207 statistics, 243, 371 status epilepticus, 312, 314, 318, 332 stellate cells, 176 stem cells, 284 sterile, 39, 165, 328, 332 stiffness, 27 stimulus, 28, 29, 48, 49, 50, 57, 58, 59, 60, 69, 74, 77, 82, 83, 85, 86, 87, 88, 91, 92, 97, 98, 99, 101, 102, 103, 104, 105, 107, 108, 109, 110, 141, 169, 173, 177, 184, 191, 213, 235, 238, 240, 241, 243, 266, 311, 312, 313, 350, 354, 358, 371, 379, 410 storage, 322, 323 strain, 415 strategies, 184, 239, 243, 319, 338, 346, 348, 368, 376, 378, 379 streams, 371 strength, 47, 50, 82, 83, 236, 240, 241, 257, 373, 411 stress, 41 stressors, 154 striatum, 173, 189, 219, 290, 401, 403 strokes, 52, 231, 245, 257 structural changes, 250, 345 subacute, 140, 153, 254 subarachnoid hemorrhage, xiii, 26 subcortical infarcts, 249, 253, 258, 261, 271 subcortical structures, 70, 78, 294, 297 subdural hematoma, 28 subgroups, 133, 315, 348 subjective, 144, 150, 166, 202, 205, 260, 337, 346 substances, 372 substantia nigra, 191, 399, 417 substitution, 109, 385 substrates, 194 subtraction, 172 subventricular zone, 414 success rate, 33, 40, 166, 358, 392, 399 suffering, 35, 42, 118, 136, 164, 170, 172, 188, 201, 202, 210, 296, 305, 348, 358, 400, 405, 413 suicidal, 154, 291, 396 suicide, 404, 408 suicide attempts, 404, 408 summer, 277, 280 superior temporal gyrus, 356, 358, 360 superiority, 266, 393, 395, 402 superposition, 22, 72, 81 suppression, 50, 59, 65, 242, 311, 312, 318, 333, 339, 340, 341, 344, 345, 348, 349, 351, 354, 357, 358, 359, 360, 396, 400 surface area, 29
social roles, 297 sodium, 98 software, 170, 172, 176, 178, 320, 326, 372, 400, 411, 412, 414, 415, 419 somatosensory, 17, 22, 31, 32, 35, 55, 88, 92, 110, 131, 132, 134, 136, 145, 157, 168, 174, 175, 178, 179, 183, 194, 199, 221, 233, 244, 254, 270, 271, 272, 277, 278, 344, 346, 375, 377, 382, 397, 398 sounds, 354, 359 South Korea, 249, 257 Spain, 365 spastic, 250, 259, 260 spasticity, 215, 250, 259, 280, 413, 418 spatial, 39, 54, 73, 76, 77, 78, 84, 85, 87, 207, 238, 244, 245, 246, 254, 278, 279, 317, 319, 320, 359, 366, 368, 372, 378, 379, 385, 386, 387, 407, 411, 420 spatial location, 84, 378 spatiotemporal, 171 specificity, 190, 278, 323, 366, 411 SPECT, 153, 169, 170, 174, 183, 199, 205, 208, 209, 211, 214, 224, 277 spectroscopy, 52, 297 spectrum, 73, 311, 398 speculation, 338 speech, 41, 144, 185, 186, 187, 195, 246, 270, 277, 278, 400, 410 speech perception, 270 speed, xiii, 186, 188, 196, 257, 329, 372 spinal cord, xii, 22, 48, 55, 57, 61, 62, 70, 89, 94, 97, 105, 110, 118, 120, 122, 124, 126, 128, 136, 137, 139, 157, 158, 159, 162, 164, 168, 175, 179, 243, 276, 283, 387, 392, 413 spinal cord injury, 136, 139, 159, 162, 243 spinal cord stimulation, 89, 105, 164, 276, 283, 392 spine, 164, 421 spines, 88, 197, 198 spiritual, xiii splenium, 19 spontaneous recovery, 238, 276, 280 sporadic, 275 square wave, 29, 38, 313 St. Louis, 88 stab wounds, 375 stability, 202, 205 stabilization, 222, 260, 358 stabilize, 155 stages, 166, 376, 395, 403, 406 standardization, 337 stars, 368
Index
surgeons, 13, 23, 24, 28, 409, 417 surgery, xi, xii, 3, 7, 15, 17, 20, 24, 25, 27, 32, 38, 39, 41, 67, 83, 110, 155, 158, 202, 204, 205, 207, 210, 211, 214, 218, 220, 224, 249, 253, 265, 280, 283, 309, 310, 312, 314, 318, 320, 324, 351, 377, 391, 392, 393, 396, 397, 398, 399, 407, 408, 418, 421 surgical, vii, xiii, 3, 9, 17, 18, 24, 33, 34, 35, 38, 39, 40, 42, 43, 104, 167, 169, 194, 203, 217, 252, 253, 259, 261, 293, 309, 310, 311, 316, 320, 328, 332, 342, 358, 360, 372, 373, 376, 378, 385, 391, 393, 394, 395, 399, 400, 401, 410, 419, 420 surprise, 132, 220 survival, 62, 284 surviving, 231, 268, 406 survivors, 231, 234, 235 susceptibility, 246, 381 suture, 5, 9, 10, 11, 13, 14, 18, 27, 37, 43, 165, 265, 330 swallowing, 279, 283 switching, 27, 36, 80, 171, 205, 207, 280 symmetry, 76 symptom, 92, 223, 354 symptoms, xii, 20, 40, 183, 188, 189, 190, 193, 194, 195, 198, 202, 203, 204, 207, 210, 212, 213, 214, 217, 218, 219, 222, 223, 231, 289, 292, 293, 294, 295, 296, 297, 298, 302, 303, 304, 305, 311, 316, 332, 393, 394, 395, 396, 401, 402, 403, 406, 408, 420 synapse, 63, 79, 176, 179, 406 synapses, 50, 62, 86, 88, 93, 106, 176, 190, 212 synaptic plasticity, 67, 220, 247 synaptic strength, 411 synaptic transmission, 232 synchronization, 219, 382 synchronous, 58, 234, 286, 313, 353 syndrome, 44, 121, 136, 137, 138, 139, 156, 157, 158, 162, 167, 178, 194, 198, 199, 210, 271, 277, 294, 303, 305, 392, 393, 401, 414, 416, 417, 418, 419, 421, 423 synergistic, 220, 401 synthesis, 372 syringomyelia, 145 systems, 3, 12, 48, 64, 134, 176, 184, 222, 224, 277, 319, 354, 392, 400, 409, 410, 411, 412, 413, 414, 421
447
T
targets, xi, 26, 35, 37, 70, 80, 81, 117, 132, 177, 262, 281, 282, 300, 357, 391, 392, 393, 394, 396, 398, 400, 402, 403, 405, 406, 407, 414 task performance, 190, 407 TBI, 276 teaching, 368 teaching strategies, 368 technicians, 407 technology, xi, 7, 45, 110, 140, 218, 323, 371, 372, 407, 409, 412, 413, 414, 418, 421 telephone, 412 television, 381 temperature, 412 temporal, xiii, 4, 7, 8, 14, 39, 52, 54, 86, 171, 271, 278, 279, 290, 293, 295, 296, 311, 315, 316, 318, 321, 322, 326, 327, 338, 339, 341, 342, 343, 344, 345, 350, 368, 371, 372, 376, 379, 384, 400, 419, 422 temporal lobe, 4, 52, 54, 171, 311, 316, 318, 345, 400, 422 temporal lobe epilepsy, 311, 318, 400, 422 tendinitis, 415 tension, 166 territory, 172, 221, 237, 261 Tesla, 46, 47, 409 test-retest reliability, 280 Texas, 280, 408, 409 textbooks, xi, 220 Thai, 302 thalamocortical pathways, 394 thalamocortical system, 397 thalamus, 70, 93, 94, 97, 134, 142, 144, 170, 174, 176, 178, 191, 205, 214, 235, 276, 279, 290, 299, 354, 355, 385, 392, 397, 398, 399, 414, 417 theft, 28 therapeutic approaches, 238, 310 therapeutic practice, 135 therapeutics, 317 therapists, 393, 407 therapy, xiii, 47, 54, 56, 135, 140, 157, 159, 168, 189, 194, 197, 204, 213, 235, 238, 240, 242, 245, 251, 270, 272, 276, 286, 289, 302, 304, 319, 320, 321, 324, 332, 334, 345, 393, 394, 396, 401, 402, 403, 404, 405, 406, 416 theta, 50, 62, 212, 234, 277, 338, 357, 404 thin film, 413 thoracic, 199 thorax, 13
448
Index
196, 197, 198, 199, 200, 225, 226, 243, 244, 245, 246, 247, 271, 286, 289, 299, 300, 301, 302, 303, 304, 305, 317, 318, 333, 349, 350, 351, 352, 360, 361, 381, 382, 384, 414, 416, 417, 418, 419, 420, 422, 423 transducer, 371 transfer, 88, 369 transformation, 371 transformations, xiii, 379 transistor, 48 transition, 73, 108, 275 transmembrane, 47, 76 transmission, 175, 232, 284, 391, 410, 412 transparency, 404 transplantation, 365, 380 trauma, 121, 126, 144, 161, 162, 210, 278, 296, 377, 416 traumatic brain injury, 159, 276, 280, 285, 286 travel, 49 treatment-resistant, 291, 300, 305, 402, 416, 417, 419 trees, 93, 97 tremor, 158, 188, 196, 203, 205, 206, 210, 211, 212, 215, 217, 224, 394, 395, 396, 397, 404, 405, 416, 417, 418, 419, 420, 421, 422, 423 trial, 23, 38, 52, 69, 124, 134, 136, 149, 151, 152, 153, 165, 166, 167, 189, 209, 212, 218, 236, 239, 240, 244, 245, 250, 266, 268, 270, 277, 299, 300, 301, 302, 304, 305, 314, 315, 317, 320, 321, 333, 357, 399, 400, 402, 415, 416, 419, 420 trial and error, 69 trigeminal, viii, 20, 30, 31, 35, 36, 117, 122, 124, 128, 157, 161, 162, 163, 164, 167, 168, 173, 392, 393, 413 trigeminal nerve, 36, 122, 124, 128, 161, 164 trigeminal neuralgia, 117, 162, 163, 164, 167 triggers, 173, 211, 325 TSH, xiii tumor, 290, 316 tumors, 15, 161, 164 two-dimensional, 387
threat, 276, 296 three-dimensional, 75, 157, 168, 373, 374, 380, 412 threshold, 17, 29, 30, 38, 48, 49, 50, 51, 52, 58, 60, 61, 62, 63, 76, 77, 82, 83, 84, 86, 92, 97, 99, 100, 102, 103, 104, 105, 107, 108, 109, 110, 118, 132, 141, 142, 144, 147, 154, 171, 175, 184, 185, 187, 192, 194, 200, 205, 207, 217, 232, 235, 239, 242, 251, 263, 266, 267, 268, 291, 293, 294, 297, 298, 300, 303, 311, 314, 315, 320, 323, 324, 340, 343, 372 thresholds, 46, 84, 85, 89, 102, 103, 104, 105, 107, 108, 109, 117, 136, 138, 144, 153, 175, 176, 192, 221, 237, 298, 367, 379 thyroid, 305 thyroid stimulating hormone, 305 tics, 401 time consuming, 39, 220 timing, 188, 236, 242, 310, 385, 409 tin, 38, 42, 50, 52, 154, 156, 321 tinnitus, ix, xii, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361 tissue, 28, 29, 31, 32, 36, 40, 46, 47, 48, 52, 73, 74, 75, 83, 85, 88, 95, 111, 177, 220, 254, 269, 276, 310, 311, 312, 316, 320, 324, 369, 373, 374, 375, 383, 385, 403, 407, 408, 412, 421 titration, 394, 408 Tokyo, 32, 112 tolerance, 309, 394 tonic, 86, 184, 232, 340, 341, 349, 353, 354, 358, 360, 398, 399, 400 topographic, 3, 14, 15, 93, 193 torticollis, 190, 196, 198 toxic, 29 toxic products, 29 toxin, 191, 196, 210 toxins, 190 tracking, 244, 269, 276 tractography, 23, 30, 39 trade, 323 trade-off, 323 training, xi, 23, 39, 232, 236, 241, 244, 246, 250, 251, 254, 259, 260, 263, 264, 266, 267, 268, 269, 270, 272, 322, 369, 377, 378, 379 trajectory, 12, 20, 43, 101, 222, 329, 331 trans, 57, 59, 107, 174, 236, 412 transcranial direct current stimulation, 45, 54, 55, 136, 194, 196, 246, 299, 301, 341, 349, 351 transcranial magnetic stimulation, 53, 54, 55, 56, 57, 65, 66, 67, 135, 136, 137, 138, 168, 194, 195,
U
ulcer, xii uncertainty, 108, 393, 405 underlying mechanisms, 309 uniform, 71, 73, 84, 101, 370 United States, 231 unmasking, 253, 394
Index
unpredictability, 409 Utah, 365, 366, 373, 374, 385 water, xii, 73 wavelet, 319, 321 weakness, 26, 142, 150, 154, 211, 267 wealth, 169 wear, 320
449
V
vacuum, 23, 150 vagal nerve, 290, 320 validation, 39, 257, 317, 355, 402 validity, 280 values, 13, 37, 91, 92, 93, 95, 171, 204, 277, 372, 379 variability, 13, 18, 56, 91, 100, 220, 221, 224, 226, 231, 249, 259, 263, 289, 343, 347, 348, 380, 381, 385, 407, 409 variables, 52, 193, 222, 271, 376 variance, 203, 207 variation, 39, 183, 197, 199 vascularization, 249 vasospasm, xiii vector, 98 vegetative state, viii, 28, 275, 276, 278, 280, 284, 285, 286 vein, 26 velocity, 107, 222, 241, 384 verbal fluency, 207, 208, 210, 212 vertigo, 359 vessels, 26, 329 vestibular system, 46 veterans, 350 vibration, 154, 191 victims, 404, 405 virtual reality, xiii visible, 11, 18, 41, 234, 239 vision, 309, 366, 367, 376, 377, 378, 379, 381, 384, 386, 387, 410, 413 visual acuity, 366 visual field, 378, 380 visual perception, 53, 365, 378, 379, 382, 410 visual system, 365, 368, 376, 379, 387 visuospatial, 238, 244, 246, 405 vitamin E, 12 vocabulary, 252 vocational, 309 voice, 195, 278, 285, 359, 396, 422
W
walking, 186, 187, 188, 205, 206, 207, 209, 257, 270, 372
450
weight gain, 309 white matter, 10, 14, 58, 63, 70, 71, 73, 77, 78, 88, 93, 100, 174, 203, 212, 279, 283, 355, 359, 397, 402, 403, 405, 417 wireless, 323, 412 wires, 37, 38, 41, 42, 367, 369, 412 withdrawal, 144, 150, 253 women, 52 workers, 337 working memory, 55, 190, 194, 198, 232, 402, 420 workstation, 36, 329 World Health Organization, 289, 365 wound dehiscence, 41 wound healing, 37, 41 wound infection, 28, 166 writing, 186, 193, 319
Index
X
x-rays, 7, 38, 165, 281, 282 X-rays, 43, 209
Y
yes/no, 144 yield, 135, 276, 346, 398 young men, 53
Z
zygomatic arch, 5, 7, 8, 10, 13, 14

Textbook of Therapeutic Cortical Stimulation

Uploaded by

Document Information

Original Description:

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

Textbook of Therapeutic Cortical Stimulation

Uploaded by

Copyright:

Available Formats

TEXTBOOK OF THERAPEUTIC CORTICAL STIMULATION

TEXTBOOK OF THERAPEUTIC CORTICAL STIMULATION

Nova Biomedical Books

Copyright 2009 by Nova Science Publishers, Inc.

Published by Nova Science Publishers, Inc. New York

To Marco and Serena Who dares wins!

viii Clinical Conditions Chronic Pain Chapter 8 Chapter 9 Chapter 10

Contents 113 115 117 139

169 181 183

217 229 231

249 257 273 275

307 309 319

335 337 353

Box. Milestones of Cortical Stimulation

In: Textbook of Therapeutic Cortical Stimulation Editor: Sergio Canavero

ISBN 978-1-60692-537-9 2009 Nova Science Publishers, Inc.

ANATOMICAL METHODS OF CORTICAL LOCALIZATION

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

METHODS OF LOCALIZATION OF THE CENTRAL AND LATERAL SULCI

Anatomical Methods of Cortical Localization

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Anatomical Methods of Cortical Localization

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Anatomical Methods of Cortical Localization

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Anatomical Methods of Cortical Localization

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Anatomical Methods of Cortical Localization

Michael C. Park, Marc A. Goldman, Gerhard M. Friehs et al.

Anatomical Methods of Cortical Localization

In: Textbook of Therapeutic Cortical Stimulation Editor: Sergio Canavero

ISBN 978-1-60692-537-9 2009 Nova Science Publishers, Inc.

FROM LOCALIZATION TO SURGICAL IMPLANTATION

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

LOCALIZATION BY FUNCTIONAL MRI (FMRI)

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

From Localization to Surgical Implantation

Youichi Saitoh and Koichi Hosomi

In: Textbook of Therapeutic Cortical Stimulation Editor: Sergio Canavero

ISBN 978-1-60692-537-9 2009 Nova Science Publishers, Inc.

PRINCIPLES OF SURGICAL IMPLANTATION AND COMPLICATION AVOIDANCE

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.

Principles of Surgical Implantation and Complication Avoidance

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.

Principles of Surgical Implantation and Complication Avoidance

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.

PRINCIPLES OF COMPLICATIONS AVOIDANCE

Principles of Surgical Implantation and Complication Avoidance

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.

Principles of Surgical Implantation and Complication Avoidance

Benoit JM Pirotte, Philippe Voordecker, Marc Levivier et al.