19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Review Article

Ancillary Studies in Pleural, Pericardial, and Peritoneal

Effusion Cytology
Kaitlin E. Sundling, MD, PhD1 ; and Edmund S. Cibas, MD2

Pleural, pericardial, and peritoneal effusion specimens present diagnostic challenges and clinical opportunities for cytol-
ogy. For the patient, these specimens may be the first diagnosis of malignancy or the first sign of disease recurrence. This
review aims to provide useful approaches with which to resolve key difficulties in cytologic diagnosis through the use of
ancillary studies, focusing on immunohistochemistry. These approaches are suggested in concert with clinical, radio-
graphic, and morphologic findings. The differentiation of reactive mesothelial cells from malignant mesothelioma and
adenocarcinoma is a recurring theme, and Wilms tumor 1 (WT1)/AE1/AE3, claudin 4, and BRCA1-associated protein 1
(BAP1) immunostains are useful new tools in the armamentarium. A targeted workup is suggested for patients with no
known primary site or with multiple prior malignancies. Molecular and other biomarker testing can be performed on even
modestly cellular fluid specimens and may allow patients to benefit from targeted therapy without the need for addi-
tional tissue biopsies. Cancer Cytopathol 2018;126:590-8. © 2018 American Cancer Society.

KEY WORDS: ascites, BRCA1-associated protein 1 (BAP1), claudin 4, cytopathology, immunohistochemistry, mesotheli-
oma, neoplasms, pericardial effusion, pleural effusion, precision medicine, unknown primary

INTRODUCTION
Body cavity fluids, including pleural, pericardial, and peritoneal fluids, present an important diagnostic
­opportunity for cytopathologists. Pleural effusions and peritoneal effusions (ascites) usually are collected using
minimally invasive methods, and effusions often are drained for therapeutic reasons. These fluids may rep-
resent the first diagnosis of cancer in patients with an occult tumor, and an effusion may be an early sign of
disease recurrence.
Although cytomorphology is the fundamental basis of a cytologic diagnosis, ancillary studies are playing
a growing role in the workup of body fluid specimens. Immunohistochemical (IHC) stains on formalin-fixed,
paraffin-embedded (FFPE) cell block sections are the most popular and readily available method and are the
focus of this review. Specific applications of IHC include: 1) resolving the mesothelial or epithelial origin of
isolated atypical cells and cell clusters; and 2) identifying the primary site of malignancy in a patient with a
history of multiple malignancies or an unknown primary site. In all cases, the morphology, clinical history,
and radiographic findings help guide the choice of ancillary studies to optimize small specimens and to reduce
costs while providing the most diagnostically useful information. This review provides an overview of helpful
ancillary methods and tips for approaching challenging cases.

Preanalytic Issues and Specimen Preparation for Ancillary Studies

We recommend upfront preparation of a FFPE cell block whenever the specimen has adequate volume and
cellularity. In our practice, this is determined by the judgement of the cytopreparatory technician based on the
Corresponding author: Kaitlin E. Sundling, MD, PhD, Wisconsin State Laboratory of Hygiene, 465 Henry Mall, Madison, WI 53706, kaitlinsundling@gmail.com
1
Wisconsin State Laboratory of Hygiene, Department of Pathology, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin;
2
Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts
The opinions or views expressed in this special issue are those of the authors and do not necessarily reflect the opinions or recommendations of the
journal editors, the American Cancer Society, or John Wiley & Sons, Inc.
Received: February 28, 2018; Revised: April 25, 2018; Accepted: April 27, 2018
Published online August 29, 2018 in Wiley Online Library (wileyonlinelibrary.com)
DOI: 10.1002/cncy.22021, wileyonlinelibrary.com

590 Cancer Cytopathology  August 2018


volume and turbidity of the specimen, as well as whether a
sufficient cell button is formed after centrifugation of the
fluid. One group has suggested a minimum fluid volume
of 60 to 75 mL for the adequate evaluation of fluids.1,2
Creating a cell block upfront reduces any delay should
IHC be needed and decreases the number of times the
specimen needs to be stored and located in the cytopre-
paratory workflow.
Several cell block preparations are available, each
with distinct advantages and limitations3‒5; a detailed
discussion is beyond the scope of this review. Important
considerations when choosing a cell block technique in-
clude the cost, workflow impact, technical skill required,
scalability, cellular yield, preservation of cytomorphol-
ogy, and performance with IHC stains.
Cell block preparation provides a means of concen-
trating paucicellular specimens and may capture larger tis-
sue fragments. Cut sections of a cell block may reveal the
internal structure of thick, 3-dimensional groups (eg, fi-
brovascular cores, hollow “cannon ball” spheres, solid cel-
lular groups, or psammomatous calcifications). In larger
cellular groups and tissue fragments, cell blocks also may
improve the visualization of nuclear and cytoplasmic fea-
tures. When ordering an initial panel of IHC stains on
a cell block, cutting extra unstained slides upfront may
be helpful for conserving small cell block specimens to
reduce the amount of specimen loss due to multiple block
facings. An initial batch of 10 unstained slides usually is
sufficient for a diagnostic workup in all but the most chal-
lenging cases. For thoracic specimens and other specimens
likely to require subsequent molecular or other biomarker
testing, 15 unstained slides may be considered.
Preparation of additional cytospin slides, ThinPrep
(Hologic, Marlborough, Massachusetts) or SurePath
(Becton, Dickinson and Company, Franklin Lakes, New
Jersey) slides, or smears for immunocytochemistry or
fluorescence in situ hybridization (FISH) can be an al-
ternative or adjunct to cell blocks. These tests should be
validated as appropriate for the specimen type, fixation
conditions, and cytologic preparation, or a disclaimer
added to the interpretation.
Specific Applications of Ancillary Studies
An overview of an algorithm for the workup of challeng-
ing effusions is presented in Fig. 1. A limited initial panel
of immunostains is helpful in a variety of scenarios, with
more specific stains performed in subsequent rounds.
Cancer Cytopathology  August 2018
19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Effusion Cytology Ancillary Studies/Sundling and Cibas
Individual IHC stains and their applications and pitfalls
are discussed below.
Distinguishing Among Adenocarcinoma,
Malignant Mesothelioma, and Reactive
Mesothelial Proliferations
Wilms Tumor 1/AE1/AE3 cytokeratin dual stain
Multiplexed IHC stains offer both efficiency and
improved diagnostic information in histopathologic
and cytopathologic applications. For example, in breast
pathology, a dual myoepithelial IHC stain allows for
the assessment of the myoepithelial layer by 2 different
markers on a single slide.6 Using a prostate carcinoma
triple stain, one slide can be used to assess simultaneously
for the presence of alpha-methylacyl-CoA racemase
(AMACR) in adenocarcinoma and the intactness of the
basal layer.7 In effusion cytology, the Wilms tumor 1/
AE1/AE3 cytokeratin (CK) (WT1/AE1/AE3) dual stain
is a multipurpose tool for colocalizing the expression
of these 2 proteins within the same cells. WT1 is a
transcriptional regulator used for its nuclear staining in
benign mesothelial cells and malignant mesothelioma; it
also is expressed in the nuclei of serous neoplasms (see
later discussion of BRCA1-associated protein 1 [BAP1]),
but the majority of other carcinomas are negative (Fig.
2).8 AE1/AE3 is a CK antibody that highlights the
cytoplasm of both epithelial cells and mesothelial cells.
Thus, dual staining with WT1/AE1/AE3 distinguishes
cells of mesothelial origin from the majority of epithelial
cells,8 which is particularly useful when clusters of cells
with indeterminate morphology are observed on the
cell block. The dual stain is ideally implemented with
different chromogens for WT1 and AE1/AE3 (eg, a
brown chromogen for WT1 and a red chromogen for
AE1/AE3). This dual stain has 2 additional diagnostic
uses: cytoplasmic staining for WT1 is characteristic
of melanoma and some sarcomas,8 and AE1/AE3
demonstrates a weak, perinuclear “dot-like” pattern in
small cell carcinoma.9 In the majority of cases, this stain
is sufficient on its own to differentiate mesothelial from
epithelial cells, and to identify background macrophages
by negative staining.
Additional stains for mesothelial
differentiation
Of single stains for mesothelial origin, WT1 combines
high specificity with good sensitivity, with positive stain-
ing reported in approximately 70% to 95% of cases with
cells of mesothelial origin.10,11 However, in cases in which
591
 19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Review Article

Figure 1. Algorithm for immunohistochemistry (IHC) of challenging effusion specimens. Wilms tumor 1 (WT1)/AE1/AE3 dual
stain and claudin 4 comprise the initial basic panel, with additional stains added either upfront or in a stepwise fashion, based
on clinical, radiographic, and morphologic features. Retained BRCA1-associated protein 1 (BAP1) staining in background
inflammatory cells is depicted as an internal control for BAP1 loss.

Cytokeratin (CK) 7* and other keratins may be positive in mesothelial cells, and therefore these stains should be interpreted with
caution and within the context of other stains and morphological features. 

+ indicates positive; -, negative; CDX2, caudal type homeobox 2; GATA3, GATA-binding protein 3; PAX8, paired box 8; SATB2,
SATB homeobox 2; TTF-1, thyroid transcription factor 1.

mesothelial origin is suspected but WT1 is negative or
equivocal, a panel of additional markers may be useful.
Other markers that have lower specificity and may stain
some adenocarcinomas include calretinin (nuclear and
cytoplasmic), D2-40 (also known as podoplanin; mem-
branous pattern), and mesothelin (membranous).10
Claudin 4
Claudin 4 is a tight junction-associated protein,12 and
staining for claudin 4 now is a preferred pan-adenocarci-
noma marker. A positive result demonstrates crisp mem-
branous staining (Fig. 2). Claudin 4 shows improved
sensitivity and specificity over Ber-EP4 and MOC-31
for distinguishing adenocarcinoma from cells of meso-
thelial origin.13‒15 Although Ber-EP4 and MOC-31 are
reasonably sensitive as adenocarcinoma markers, both
demonstrate positive staining in a subset of epithelioid
mesotheliomas, which represents a potential diagnos-
tic pitfall in effusion specimens.14,15 A study found no
membranous staining in 75 malignant mesotheliomas,
although 11% (8 of 75 cases) had weak cytoplasmic
staining, which is considered nonspecific.12 In our prac-
tice, we no longer routinely use Ber-EP4 or MOC-31
because we have found claudin 4 to be more useful. In
laboratories in which claudin 4 is not available, Ber-EP4
and MOC-31 still are of use, perhaps together with a
mesothelial marker such as WT1 when malignant meso-
thelioma enters the differential diagnosis.
BAP1, a new marker that is helpful in
distinguishing reactive mesothelial cells from
malignant mesothelioma
The distinction between reactive and malignant mesothe-
lial cells has long been a challenge in effusion cytology.
In biopsy specimens, pleural invasion aids in the diag-
nosis of malignant mesothelioma; however, in cytologic
specimens, in which invasion per se cannot be assessed,
malignant mesothelioma is difficult to diagnose with
certainty, except perhaps when a patient with a robust
asbestos exposure history, recurrent pleural effusions, and

592 Cancer Cytopathology  August 2018


Figure 2.  (A-C) Serous borderline tumor (peritoneal fluid). (A) H & E staining. (B) Wilms tumor 1 (WT1)/AE1/AE3 dual stain (WT1
in brown and AE1/AE3 in red). (C) Membranous claudin 4 staining. (D-F) Metastatic poorly differentiated carcinoma (pleural
fluid). (D) H & E staining. (E) WT1/AE1/AE3 dual stain (WT1 in brown and AE1/AE3 in red; note the background mesothelial cell
in the upper left). (F) Membranous claudin 4 staining.
characteristic radiographic findings has a clearly malig-
nant effusion with abundant large (“mulberry-like”) cell
clusters with a mesothelial immunophenotype. Until re-
cently, epithelial membrane antigen was perhaps the best
marker for distinguishing benign mesothelial cells from
malignant mesothelioma, but some studies reported a
lower sensitivity.16 Thus, it was not widely used for this
purpose. In addition, the specificity of epithelial mem-
brane antigen staining varies by the specific antibody
clone that is used. Other markers, such as desmin and
glucose transporter type 1 (GLUT-1), also have been
proposed to distinguish between benign and malignant
mesothelial cells, but significant overlap between the 2
has been demonstrated, and therefore these markers do
not provide sufficient specificity for clinical practice.10,16
A big step forward was the recent availability of the
BAP1 immunostain, a relatively new marker that is ex-
pressed in the nuclei of benign mesothelial cells but is
absent in greater than one-half of cases of malignant me-
sothelioma.17 The BAP1 gene is located at 3p21, which is
lost in approximately 30% of mesotheliomas.17 Including
mesotheliomas with BAP1 truncations and other muta-
tions, approximately 60% to 79% of mesotheliomas have
genetic alterations in BAP1.17 Accordingly, IHC stain-
ing for BAP1 is lost in a substantial subset of malignant
Cancer Cytopathology  August 2018
19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Effusion Cytology Ancillary Studies/Sundling and Cibas
mesotheliomas but retained in benign mesothelial prolif-
erations, and its performance correlates well with 9p21
FISH (p16 deletion).18,19 It is important to note that
patchy BAP1 staining may be noted in benign and reac-
tive mesothelial populations, and thus caution should be
observed when making an interpretation of BAP1 loss in
effusion specimens containing rare atypical cells.20 Special
attention should be paid to background inflammatory
cells, an essential internal control for BAP1 staining (Fig.
3). A major differential diagnostic consideration within the
context of malignant mesothelioma is serous carcinoma of
the gynecologic tract, both of which are WT1 positive.
BAP1 is retained in nearly all serous carcinomas,21 making
it a useful addition to a panel to differentiate between ma-
lignant mesothelioma and serous carcinoma, which also
could include calretinin (usually positive in mesothelial
cells) and paired box 8 (PAX8) (usually positive in serous
carcinoma), as discussed in more detail below.
Additional studies to differentiate
reactive mesothelial cells from malignant
mesothelioma
In addition to the aforementioned IHC stains, several nucleic
acid-based tests are available for the diagnosis of malignant
593
 19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Review Article

Figure 3. BRCA1-associated protein 1 (BAP1) loss in mesothelial populations favors a malignant origin. (A and B) Malignant
mesothelioma demonstrating a single-cell pattern (A: H & E staining; B: BAP1 immunohistochemical stain). (C and D) Malignant
mesothelioma demonstrating mulberry-like cell clusters (C: H & E staining; D: BAP1 immunohistochemical stain). Note the
retained BAP1 staining in inflammatory cells as an essential internal control.

mesothelioma. These include conventional cytogenetics (re-
quiring the allocation of fresh specimen for culture), FISH,
UroVysion (Abbott Molecular Inc, Des Plaines, Illinois),
and comparative genomic hybridization arrays.22‒25 The
combination of karyotyping and 9p21 and 22q deletions
by FISH in patients with malignant mesothelioma detected
diagnostic abnormalities in approximately 54% of patients
(13 of 24 patients), including 8 whose specimens were nega-
tive or suspicious for malignant mesothelioma on cytology.22
Potential pitfalls include the confirmation of the mesothelial
origin of the cells of interest, because other malignancies
may share these chromosomal alterations.
Workup of Malignant Effusions to Determine
Site of Origin
The most cost-effective “studies” for the initial evalu-
ation of malignant effusions are a thorough review
of the clinical history and radiographic findings and
cytomorphologic examination of the fluid. A personal
history of malignancy is a key consideration in guiding
the diagnostic workup. Additional detailed history might
include a history of carcinogenic exposures (eg, tobacco
use or asbestos exposure), symptoms and signs indicating
a possible primary tumor localization (eg, melena), and
a family history of malignancy. A high level of clinical
concern for a specific primary site may lower the thresh-
old for ordering confirmatory IHC stains. A discussion
of specific IHC workup strategies by effusion site follows.
Morphologic evaluation of the fluid may suggest a
particular lineage or differential diagnosis. Mesothelial
features such as microvillous “lacy skirt” borders, windows
between adjacent cells, and multinucleation suggest meso-
thelial origin and allow targeting of the IHC panel. Isolated
atypical cells may represent reactive mesothelial cells, me-
sothelioma, adenocarcinoma, melanoma, lymphoma, or
less common entities such as metastatic sarcoma.

594 Cancer Cytopathology  August 2018


Pleural and pericardial effusions
In men, the most frequent causes of malignant pleural
effusions are lung carcinoma, lymphoma/leukemia, and
gastrointestinal carcinoma.26 In women, breast carci-
noma is the most frequent cause (although malignant
pleural effusions very rarely are the initial presentation
of breast cancer), followed by lung carcinoma and lym-
phoma/leukemia.26 Lung adenocarcinoma usually can
be identified through positive IHC staining for thyroid
transcription factor 1 (TTF-1) and napsin A, with a few
important caveats. Mucinous lung adenocarcinoma may
lack both TTF-1 and napsin A expression27 and, in such
cases, a metastasis from another site such as the gastroin-
testinal tract should be considered. TTF-1 is not entirely
specific for lung origin. It also is expressed in most thy-
roid malignancies (including medullary thyroid carcino-
mas), but it is rare for a thyroid malignancy to present
clinically as a malignant effusion of unknown primary.
A minority of endometrial and endocervical carcinomas
also are strongly positive for TTF-1.28 In women, breast
carcinoma is an important consideration with pleural flu-
ids and, less often, pericardial fluids. Metastatic breast
carcinoma in effusions often forms “cannonball” spheres,
which may be hollow on cell block preparations.26 A use-
ful IHC panel for breast origin includes GATA-binding
protein 3 (GATA3) and mammaglobin. Estrogen recep-
tor (ER) should be used cautiously as a diagnostic stain in
patients with known ER-positive breast carcinoma. Other
malignancies can be ER positive, including carcinomas
of Mullerian origin, lung adenocarcinoma, and papillary
thyroid carcinoma.29 However, the biomarkers ER, pro-
gesterone receptor, and human epidermal growth factor
receptor 2 (HER2) can be performed on effusion speci-
mens for the prediction of clinical therapeutic response.30
Peritoneal effusions
In men, the most common causes of malignant peritoneal
effusions are lymphoma/leukemia, gastrointestinal carci-
noma, and pancreatic carcinoma.26 In women, the most
common causes of malignant peritoneal effusions are
carcinomas of the gynecologic tract, breast carcinoma,
and gastric carcinoma.26 In the workup of a potential
gastrointestinal primary site, CK7 and CK20 should be
interpreted with caution because mesothelial cells are
strongly positive for CK7. It may be helpful to compare
these stains with staining for a mesothelial marker such
as WT1 or calretinin as well as claudin 4 to determine
the percentage and distribution of mesothelial cells and
adenocarcinoma. More specific stains for gastrointestinal
Cancer Cytopathology  August 2018
19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Effusion Cytology Ancillary Studies/Sundling and Cibas
origin such as caudal type homeobox 2 (CDX2) or
SATB homeobox 2 (SATB2) may be included. The dif-
ferential diagnosis between mesothelial cells and a serous
neoplasm often is straightforward. Although mesothelial
cells and serous neoplasms both demonstrate nuclear
WT1 expression (Fig. 1), serous neoplasms usually lack
other mesothelial markers (calretinin and D2-40) and
express PAX8 (nuclear staining) and claudin 4 (membra-
nous staining), which usually are absent in mesothelial
cells. It is interesting to note that PAX8 also is expressed
in carcinomas of renal and thyroid origin.29 If malignant
mesothelioma is in the morphologic and clinical differ-
ential diagnosis, BAP1 may be added to the panel.21
Effusions With Increased Lymphocytes or
Other Hematolymphoid Cells
In fluids containing increased numbers of hematolym-
phoid cells, correlation of cytologic preparations with
peripheral counts as well as fluid cell counts and slide
preparations in the hematology laboratory may help in
triaging the fluid appropriately and in providing consist-
ent reporting across the laboratory. In effusions contain-
ing distinctive populations of lymphocytes and plasma
cells, flow cytometry can be a very useful diagnostic
study. Flow cytometry can be performed only on the
fresh specimen, before alcohol or formalin fixation. A
limited hematologic workup can be performed on a cell
block specimen, including CD3 and CD20, with tar-
geted B-cell or T-cell panels to follow.31 Unstained levels
from the cell block, ThinPrep or SurePath slides, or un-
stained cytospin slides prepared directly from the fluid
may be used for subsequent FISH testing.
A pleural effusion occurs in approximately 20%
to 30% of patients with lymphoma, both Hodgkin and
non-Hodgkin types.31 The majority of patients have a
known history of lymphoma or a known predisposing
condition, such as human immunodeficiency virus in-
fection, which can aid greatly in targeting the workup.31
Lymphomatous ascites and pericardial effusions are less
common. When lymphoma is suspected, a cell block
should be prepared in addition to flow cytometry, if there
is sufficient specimen. Diffuse large B-cell lymphoma
may be difficult to characterize on flow cytometry, and
the neoplastic cells of Hodgkin lymphoma typically are
rare, leading to negative results. If large Reed-Sternberg-
like cells are identified on the cell block, appropriate
IHC workup including CD30, CD15, BOB1, and oct-
amer transcription factor-2 (OCT2) can be performed.
Primary effusion lymphoma is a rare large-cell lymphoma
595

Review Article
most often associated with human immunodeficiency
virus together with human herpesvirus 8 viral infec-
tion.31 Morphologically, primary effusion lymphoma is
a large B-cell lymphoma with a plasmablastic, immuno-
blastic, or anaplastic appearance.26 On flow cytometry,
the cells usually are positive for CD45 and negative for
the majority of B-cell markers. The cells may express ac-
tivation-associated antigens, such as human leukocyte
antigen–antigen D related (HLA-DR), and sometimes
are identified by aberrant T-cell marker expression.32
The detection of human herpesvirus 8 viral infection,
which may be performed by IHC staining of a cell block,
is essential to the diagnosis.32
Eosinophil-rich pleural effusions are not infrequent;
the most common cause is pneumothorax (spontaneous
or iatrogenic), although Churg-Strauss syndrome is a po-
tential consideration.26 The presence of increased eosin-
ophils within a background of mixed inflammatory cells
may raise concern for Hodgkin lymphoma, in which
IHC staining as described in the above paragraph may
be of use. Langerhans cell histiocytosis is exceptionally
rare in pleural fluids.33
Molecular and Biomarker Testing in
Effusion Specimens
Effusion cytology specimens are a valuable resource for
molecular testing. Smears, cytospin preparations, liquid-
based preparation slides, and cell blocks all can be used
for molecular testing. FFPE cell blocks fit easily into
existing workflows built around FFPE surgical tissue.
However, alcohol-fixed or air-dried and stained cytology
slides can be scraped and used for molecular testing, and
may yield higher quality nucleic acid due to the lack of
formalin crosslinking artifacts.34 Appropriate validation
of molecular tests for the cytologic preparation and fixa-
tive type is important, particularly because the amount
of input DNA may need to be adjusted. In selected cases,
it may be helpful to reserve a portion of the fresh speci-
men for prompt nucleic acid extraction and next-gener-
ation sequencing to avoid the damage to nucleic acids
caused by formalin fixation. Effusions with even low to
moderate tumor quantity can be used for next-generation
sequencing and for single-gene tests, although a low per-
centage of tumor in relation to background mesothelial
or inflammatory elements may be limiting.35 Anaplastic
lymphoma kinase (ALK) and reactive oxygen species 1
(ROS1) IHC stains may be useful in specimens with a low
596
19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
tumor percentage because the counterstain affords mor-
phologic identification of tumor cells. ALK has been par-
ticularly well studied in cytology specimens.36 ALK IHC
staining is acceptable as an alternative to FISH, as per the
College of American Pathologists guideline for molecular
testing in patients with lung cancer.37 However, ROS1
IHC staining should be confirmed by follow-up FISH or
a molecular method.37 Programmed death-ligand 1 (PD-
L1) has been reported in cytology specimens containing
≥100 well-preserved tumor cells, with good concordance
with surgical specimens.38 To the best of our knowledge,
there are very few data to date regarding reporting PD-L1
in specimens that contain <100 well-preserved tumor
cells. Thus, PD-L1 should be interpreted cautiously in
specimens with low cellularity. If PD-L1 is reported in
lower cellularity specimens, an explanatory note should
indicate that the interpretation might be limited by the
low cellularity. Because of the clinical importance of pre-
dictive biomarkers, the College of American Pathologists
guidelines for validation of these biomarkers are more
stringent than those for other stains.39
Conclusions
Pleural, pericardial, and peritoneal effusions are challeng-
ing yet highly clinically useful specimens for the initial
diagnosis of malignancy and the early identification of
disease recurrence in patients with cancer. The morpho-
logic evaluation of effusion specimens has been greatly
enhanced recently by the addition of new IHC stains that
help with morphologically ambiguous fluids. Targeted
panels of IHC stains in combination with clinical and
morphologic features provide more specific diagnoses
than ever before, which may eliminate the need for more
invasive tissue sampling. In addition, cytology slides and
cell block preparations may be used for molecular and
other biomarker testing. These opportunities continue to
shape the cytopathologist's role as a minimally invasive
diagnostician in the era of precision medicine.
FUNDING SUPPORT
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
AUTHOR CONTRIBUTIONS
Kaitlin E. Sundling: Conceptualization, writing-original
draft, and writing–review and editing. Edmund S. Cibas:
Conceptualization and writing–review and editing.
Cancer Cytopathology  August 2018
 19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Effusion Cytology Ancillary Studies/Sundling and Cibas

REFERENCES diagnosis: comparison with 9p21 FISH and BAP1 immunohisto-

chemistry. Lung Cancer. 2017;104:98-105.
1. Rooper LM, Ali SZ, Olson MT. A minimum fluid volume of 75 mL
19. Walts AE, Hiroshima K, McGregor SM, Wu D, Husain AN,
is needed to ensure adequacy in a pleural effusion: a retrospective
Marchevsky AM. BAP1 immunostain and CDKN2A (p16)
analysis of 2540 cases. Cancer Cytopathol. 2014;122:657-665.
FISH analysis: clinical applicability for the diagnosis of
2. Rooper LM, Ali SZ, Olson MT. A minimum volume of more
malignant mesothelioma in effusions. Diagn Cytopathol.
than 60 mL is necessary for adequate cytologic diagnosis of ma-
2016;44:599-606.
lignant pericardial effusions. Am J Clin Pathol. 2016;145:101-106.
20. McGregor SM, Dunning R, Hyjek E, Vigneswaran W, Husain
3. Saqi A. The state of cell blocks and ancillary testing: past, present,
AN, Krausz T. BAP1 facilitates diagnostic objectivity, classifica-
and future. Arch Pathol Lab Med. 2016;140:1318-1322.
tion, and prognostication in malignant pleural mesothelioma. Hum
4. Balassanian R, Wool GD, Ono JC, et al. A superior method for
Pathol. 2015;46:1670-1678.
cell block preparation for fine-needle aspiration biopsies. Cancer
21. Andrici J, Jung J, Sheen A, et al. Loss of BAP1 expression is very
Cytopathol. 2016;124:508-518.
rare in peritoneal and gynecologic serous adenocarcinomas and can
5. La Fortune KA, Randolph ML, Wu HH, Cramer HM.
be useful in the differential diagnosis with abdominal mesotheli-
Improvements in cell block processing: the Cell-Gel method.
oma. Hum Pathol. 2016;51:9-15.
Cancer Cytopathol. 2017;125:267-276.
22. Factor RE, Dal Cin P, Fletcher JA, Cibas ES. Cytogenetics and
6. Pavlakis K, Zoubouli C, Liakakos T, et al. Myoepithelial cell
fluorescence in situ hybridization as adjuncts to cytology in
cocktail (p63+SMA) for the evaluation of sclerosing breast lesions.
the diagnosis of malignant mesothelioma. Cancer Cytopathol.
Breast. 2006;15:705-712.
2009;117:247-253.
7. Herawi M, Epstein JI. Immunohistochemical antibody cocktail
23. Dal Cin P, Qian X, Cibas ES. The marriage of cytology and cyto-
staining (p63/HMWCK/AMACR) of ductal adenocarcinoma and
genetics. Cancer Cytopathol. 2013;121:279-290.
Gleason pattern 4 cribriform and noncribriform acinar adenocarci-
24. Chirac P, Maillet D, Lepretre F, et al. Genomic copy num-
nomas of the prostate. Am J Surg Pathol. 2007;31:889-894.
ber alterations in 33 malignant peritoneal mesothelioma ana-
8. Conner JR, Cibas ES, Hornick JL, Qian X. Wilms tumor 1/cy-
lyzed by comparative genomic hybridization array. Hum Pathol.
tokeratin dual-color immunostaining reveals distinctive staining
2016;55:72-82.
patterns in metastatic melanoma, metastatic carcinoma, and meso-
25. Flores-Staino C, Darai-Ramqvist E, Dobra K, Hjerpe A.
thelial cells in pleural fluids: an effective first-line test for the workup
Adaptation of a commercial fluorescent in situ hybridization test
of malignant effusions. Cancer Cytopathol. 2014;122:586-595.
to the diagnosis of malignant cells in effusions. Lung Cancer.
9. Khan M, Nizami S, Mirrakhimov AE, Maughan B, Bishop JA,
2010;68:39-43.
Sharfman WH. Primary small cell neuroendocrine carcinoma of
26. Cibas ES. Pleural, pericardial and peritoneal fluids. In: Cibas
paranasal sinuses. Case Rep Med. 2014;2014:874719.
ES, Ducatman BS, eds. Cytology: Diagnostic Principles and
10. Husain AN, Colby TV, Ordonez NG, et al. Guidelines for patho-
Clinical Correlates. 4th ed. Philadelphia, PA: Elsevier Saunders;
logic diagnosis of malignant mesothelioma 2017 update of the
2014:127-153.
consensus statement from the International Mesothelioma Interest
27. Wu J, Chu PG, Jiang Z, Lau SK. Napsin A expression in primary
Group. Arch Pathol Lab Med. 2018;142:89-108.
mucin-producing adenocarcinomas of the lung: an immunohisto-
11. Tandon R, Jimenez-Cortez Y, Taub R, Borczuk AC.
chemical study. Am J Clin Pathol. 2013;139:160-166.
Immunohistochemistry in peritoneal mesothelioma: a sin-
28. Siami K, McCluggage WG, Ordonez NG, et al. Thyroid transcrip-
gle-center experience of 244 cases. Arch Pathol Lab Med. 2018;
tion factor-1 expression in endometrial and endocervical adenocar-
142:236-242.
cinomas. Am J Surg Pathol. 2007;31:1759-1763.
12. Jo VY, Cibas ES, Pinkus GS. Claudin-4 immunohistochemistry
29. Kandalaft PL, Gown AM. Practical applications in immunohisto-
is highly effective in distinguishing adenocarcinoma from ma-
chemistry: carcinomas of unknown primary site. Arch Pathol Lab
lignant mesothelioma in effusion cytology. Cancer Cytopathol.
Med. 2016;140:508-523.
2014;122:299-306.
30. Nakayama Y, Nakagomi H, Omori M, et al. Benefits of using the
13. Oda T, Ogata S, Kawaguchi S, et al. Immunocytochemical utility
cell block method to determine the discordance of the HR/HER2
of claudin-4 versus those of Ber-EP4 and MOC-31 in effusion cy-
expression in patients with metastatic breast cancer. Breast Cancer.
tology. Diagn Cytopathol. 2016;44:499-504.
2016;23:633-639.
14. Mawas AS, Amatya VJ, Kushitani K, et al. MUC4 immunohis-
31. Das DK. Serous effusions in malignant lymphomas: a review.
tochemistry is useful in distinguishing epithelioid mesothelioma
Diagn Cytopathol. 2006;34:335-347.
from adenocarcinoma and squamous cell carcinoma of the lung. Sci
32. Pittaluga S, Said JW. Virally associated B-cell lymphoproliferative
Rep. 2018;8:134.
disease. In: Jaffe ES, Arber DA, Campo E, Harris NL, Quintanilla-
15. Chaouche-Mazouni S, Scherpereel A, Zaamoum R, et al. Claudin
Martinez L, eds. Hematopathology. 2nd ed. Philadelphia, PA:
3, 4, and 15 expression in solid tumors of lung adenocarcinoma
Elsevier; 2017:547-564.
versus malignant pleural mesothelioma. Ann Diagn Pathol.
33. Nagaoka S, Maruyama R, Koike M, et al. Cytology of Langerhans
2015;19:193-197.
cell histiocytosis in effusions: a case report. Acta Cytol.
16. Su XY, Li GD, Liu WP, Xie B, Jiang YH. Cytological differential
1996;40:563-566.
diagnosis among adenocarcinoma, epithelial mesothelioma, and
34. Roy-Chowdhuri S, Chen H, Singh RR, et al. Concurrent fine nee-
reactive mesothelial cells in serous effusions by immunocytochem-
dle aspirations and core needle biopsies: a comparative study of sub-
istry. Diagn Cytopathol. 2011;39:900-908.
strates for next-generation sequencing in solid organ malignancies.
17. Bott M, Brevet M, Taylor BS, et al. The nuclear deubiquitinase BAP1
Mod Pathol. 2017;30:499-508.
is commonly inactivated by somatic mutations and 3p21.1 losses in
35. Roy-Chowdhuri S, Goswami RS, Chen H, et al. Factors affecting
malignant pleural mesothelioma. Nat Genet. 2011;43: 668-672.
the success of next-generation sequencing in cytology specimens.
18. Hida T, Hamasaki M, Matsumoto S, et al. Immunohistochemical
Cancer Cytopathol. 2015;123:659-668.
detection of MTAP and BAP1 protein loss for mesothelioma

Cancer Cytopathology  August 2018 597


Review Article
36. Thunnissen E, Allen TC, Adam J, et al. Immunohistochemistry of
pulmonary biomarkers: a perspective from members of the Pulmonary
Pathology Society. Arch Pathol Lab Med. 2018;142:408-419.
37. Lindeman NI, Cagle PT, Aisner DL, et al. Updated molecular
testing guideline for the selection of lung cancer patients for treat-
ment with targeted tyrosine kinase inhibitors: guideline from the
College of American Pathologists, the International Association
for the Study of Lung Cancer, and the Association for Molecular
Pathology. J Mol Diagn. 2018;20:129-159.
598
19346638, 2018, S8, Downloaded from https://acsjournals.onlinelibrary.wiley.com/doi/10.1002/cncy.22021 by Cochrane Mexico, Wiley Online Library on [09/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
38. Heymann JJ, Bulman WA, Swinarski D, et al. PD-L1 expression
in non-small cell lung carcinoma: comparison among cytology,
small biopsy, and surgical resection specimens. Cancer Cytopathol.
2017;125:896-907.
39. Stuart LN, Volmar KE, Nowak JA, et al. Analytic validation
of immunohistochemistry assays: new benchmark data from
a survey of 1085 laboratories. Arch Pathol Lab Med. 2017;141:
1255-1261.
Cancer Cytopathology  August 2018