An Updated Contextual Approach to Mesothelial Proliferations in Pleural Effusion Cytology Leveraging Morphology, Ancillary Studies, and Novel Biomarkers

On histology, mesothelial cells lining the thoracic cavity's pleural surface are typically flat, elongated, and arranged in a single layer. Similar to their appearance at other body sites such as the liver and spleen, these mesothelial cells can also take on a more cuboidal shape.¹²  In these more cuboidal mesothelial cells, the cell's eosinophilic and granular cytoplasmic quality is more evident. The nucleus is round to ovoid and typically displays an even, granular chromatin distribution with a variably distinct nucleolus. Irritation of the mesothelial surface lining can induce reactive and hyperplastic changes. Mesothelial hyperplasia can manifest as a proliferation of mesothelial cells in solid sheets, nests, papillary and tubulopapillary structures, glandlike structures, cordlike arrangements, or as single cells with variable nuclear changes.¹³  Concurrent reactive changes induced by various etiologies can yield nuclear enlargement (with subsequently increased nuclear to cytoplasmic [N:C] ratio), nuclear contour irregularity, coarse chromatin distribution, prominent nucleoli, and multinucleation.^12,14  Additionally, other cytologic features raising the concern for malignancy may be displayed in reactive mesothelial cells, such as mitotic figures, nuclear atypia, and background necrosis (such as that seen in rheumatoid effusions).¹⁴  While some reactive changes in mesothelial cells may result in the formation of papillary arrangements entrapped in fibrotic stroma, hyperplastic mesothelial cells show no evidence of invasion into underlying supporting tissues and are typically monotonous in appearance.

On cytologic evaluation of benign pleural effusion specimens, mesothelial cells are typically low in quantity and characterized by low N:C ratios, smooth to slightly irregular nuclear contours, granular chromatin distribution, and moderately abundant amounts of densely granular cytoplasm. The cytoplasm may demonstrate characteristic “two-tone” cytoplasm in which there is a denser central cytoplasm (endoplasm) due to the perinuclear accumulation of filaments and a paler peripheral cytoplasm (ectoplasm).¹⁵  Multinucleation can be seen in benign and reactive mesothelial cells. In pleural effusions, mesothelial cells are typically arranged singly and in small clusters. Owing to their surface long, slender microvilli, outlines of neighboring mesothelial cells can be sharply demarcated by the presence of clear spaces or intercellular “windows.”¹⁶  Reactive mesothelial cells on cytology retain many of the features of benign mesothelial cells, but like the prior description of mesothelial cells on histology, they can also display morphologic features that can raise the concern for malignancy, including increased N:C ratios, coarse chromatin distribution, and prominent nucleoli (Figure 1, A).¹⁷  Reactive mesothelial cells can also display cytoplasmic blebbing and/or cytoplasmic vacuolization.¹⁸  These vacuoles can take a variety of morphologic appearances depending on their etiologies. A subset of this cytoplasmic vacuolization is associated with cellular degeneration in hydropic change, which can manifest as multiple vacuoles or a single large vacuole that peripherally displaces the nucleus.¹⁹  Other cytoplasmic vacuoles within mesothelial cells can be due to the accumulation of glycogen, lipid, or hyaluronan (Figure 1, B).^15,19  Beyond these cytologic changes, reactive mesothelial cell proliferations can demonstrate variable architectural arrangements, including small, 3-dimensional cell clusters, papillary-like aggregates, and pseudoacini. When clustered in crowded cell groups, reactive mesothelial cells frequently display a scalloped or “knobby” border (Figure 1, B). Background inflammatory cells may also be seen in a subset of pleural effusions with reactive mesothelial cells.^5,18,20 

Overtly malignant pleural effusions largely represent metastatic disease, often from adenocarcinomas, which have been reported to comprise a majority of malignant pleural effusions.^14,21–23  Often, there are clinical histories for which a malignant pleural effusion is expected as part of clinical progression/widespread metastasis from a known primary. However, it is not infrequent for malignant effusions to be the initial sign of malignancy in cases of unknown primaries.²⁴ 

Adenocarcinomas in pleural effusion specimens are typically characterized by a biphasic population with a discrete population of malignant cells admixed with background benign elements.^19,25  The malignant cells frequently display overt pleomorphism with enlarged nuclei (and subsequently increased N:C ratios), irregular nuclear contours, coarse chromatin distribution, and prominent nucleoli (Figure 2, A). Although singly dispersed malignant cells may be apparent in some tumor types (ie, lobular carcinoma of the breast, signet ring carcinoma of the gastrointestinal tract), many adenocarcinomas display 3-dimensional cell clusters with variable degrees of nuclear overlap and smooth “community” borders^25–27  (Figure 2, B). These smooth borders stand in contrast with the prototypically scalloped borders of mesothelial cell clusters (Table 1).

However, the degree of morphologic overlap between mesothelial cell proliferations and adenocarcinoma can preclude accurate delineation of cell etiology on morphology alone. As mentioned previously, pleomorphism can be identified in malignant processes and reactive mesothelial cells. Conversely, some subtypes of adenocarcinoma may not display overt pleomorphism and instead feature more monotonous malignant cell populations^12,18  (ie, subsets of breast and endometrioid adenocarcinomas). Although frequently associated with cell clusters of adenocarcinoma, the aforementioned “community” borders of cell groups can occasionally be seen in mesothelial proliferations (Figure 3). Other features displayed by mesothelial cells can further complicate the morphologic delineation between mesothelial proliferations and adenocarcinoma. Although cytoplasmic vacuolization is more commonly associated with adenocarcinoma, as mentioned previously, both reactive and malignant mesothelial cells may also display variable degrees of vacuolization ranging in size and location (peripheral or central).^18,28  Large vacuoles can indent mesothelial cell nuclei and mimic the appearance of signet ring adenocarcinoma.²⁹  Additionally, as is more evident in Diff-Quik?stained smear preparations, peripheral cytoplasmic blebs can be appreciated in mesothelial cells.²⁷  “Collagen balls” (more commonly featured in pelvic washing cytology specimens) featuring attenuated benign mesothelial cells surrounding a dense, acellular rounded fragment of collagen can also confound the diagnosis of reactive versus malignant entities, as these structures may resemble the 3-dimensional clusters seen in adenocarcinoma.^30,31 

These overlapping morphologic features highlight the importance of leveraging ancillary studies in addition to obtaining relevant clinical and radiologic information in making accurate cytologic assessments. Immunohistochemistry can be of great use when trying to distinguish cells of mesothelial and epithelial (ie, adenocarcinoma) origin. However, there is a variable frequency of cell marker positivity among mesothelial and epithelial cells, and no single marker exhibits 100% sensitivity and specificity in delineating the 2 cell etiologies. As such, using panels composed of multiple immunohistochemical markers can help elucidate the origin of atypical epithelioid cells in effusion specimens. In these cases, the International Mesothelioma Interest Group (IMIG) has suggested to use at least 2 mesothelial cell markers and 2 other markers relevant to the working diagnosis.^12,14,32  Frequently used markers for mesothelial cells include Wilms tumor 1 (WT-1), calretinin, cytokeratin 5/6 (CK5/6), and D2-40 (podoplanin). Additional mesothelial cell markers have been used in varying degrees, including HMBE-1, thrombomodulin, and mesothelin (Table 1). The use of multiple mesothelial markers is necessary, particularly in the workup for a malignant mesothelial proliferation, which may be negative for 1 or several of these markers.³³  However, it is important to highlight potential interpretative pitfalls. For example, calretinin can highlight breast carcinomas of basaloid types³⁴  as well as lung adenocarcinomas³⁵  and ovarian carcinomas.¹²  In addition to its known positivity for squamous cell carcinoma, CK5/6 has also been reported in a significant subset of salivary gland tumors, urothelial carcinomas, endometrial adenocarcinomas, pancreatic adenocarcinomas, breast adenocarcinomas, and ovarian adenocarcinomas.³⁶  Similarly, D2-40 positivity is also featured in cutaneous carcinomas,³⁷  subsets of squamous cell carcinomas derived from the lung,^38,39  and subsets of ovarian and breast adenocarcinomas.⁴⁰  However, newer immunohistochemical markers have been highlighted as being more specific in identifying mesothelial cells. Studies investigating the utility of HEG1 (heart development protein with EGF-like domains 1) have showcased its higher sensitivity and specificity in the delineation of mesothelial from epithelial cells in comparison to WT-1 and calretinin.^32,41,42  Traditionally, general epithelial markers such as MOC-31, BerEP4, B72.3, Leu-M1, and carcinoembryonic antigen (CEA) among others have been used to highlight carcinomas in effusion specimens. However, similar to the aforementioned mesothelial markers, their use also presents potential interpretative pitfalls, as Ber-EP4, B72.3, and MOC-31 positivity have been reported in small subsets of mesotheliomas.^38,43,44  However, a newer marker, claudin-4, has been shown to have a higher sensitivity³⁸  and specificity than MOC-31 and Ber-EP4 in distinguishing metastatic adenocarcinoma from mesothelial proliferations. As noted by Lepus and Vivero,⁴⁵  reported specificities for BerEP4, MOC31, and claudin-4 range from 95% to 98%, 87% to 97%, and 99% to 100%, respectively.^45 ,46  Correlating sensitivities have been reported to range from 74% to 89%, 86% to 92%, and 91% to 100%.⁴⁶  Organ-specific immunohistochemical markers are also useful in highlighting metastatic adenocarcinomas in patients with known or suspected extrapleural primaries. Nuclear site–specific markers may be particularly helpful in these cases (such as TTF-1 for lung adenocarcinomas, p63 or p40 for squamous cell carcinomas, and CDX-2 for adenocarcinomas with intestinal differentiation). However, some may show positivity in mesothelial proliferations. GATA3 is commonly used to highlight breast and urothelial cancers but has been reported to be diffusely positive in 33% to 50% of epithelioid mesotheliomas.^47,48  PAX8, which is commonly used to delineate renal, thyroid, or Müllerian origins, has been reported as positive in benign mesothelial cells and peritoneal mesotheliomas.⁴⁹  These overlapping expression patterns of markers between epithelial and mesothelial-derived cells highlight the importance of using multiple immunohistochemical markers to appropriately designate an atypical epithelioid proliferation as mesothelial or epithelial in origin.

Mesothelioma is an aggressive disease with an overall survival of 4 to 17 months after initial diagnosis.^32,50–52  Patients typically present with recurrent unilateral effusions in addition to a combination of symptoms that can include dyspnea, chest pain, persistent cough, and weight loss. Historically, former asbestos exposure has been the key piece of patient history that arouses suspicion for mesothelioma. Although this is the most common risk factor for the development of mesothelioma, other causes have been identified, including prior radiation exposure in the treatment of lymphoma,^53,54  exposure to other environmental nonasbestos mineral fibers,^55,56  and the presence of predisposing mutations for the development of mesothelioma.^8,12,57  On imaging, diffuse pleural thickening and nodularity or pleural plaques/calcifications can be seen.¹²  In some cases, mesothelioma can present as a single localized mass⁵⁷  or a lung-based nodule.⁵⁸  It is important to note that although risk factor exposure and the aforementioned radiologic features help raise suspicion for mesothelioma, there are clinical and radiologic pitfalls to consider. While asbestos exposure is most commonly associated with the development of mesothelioma, exposure can also cause benign pleural diseases, which can manifest as pleural plaques and thickening on imaging and radiologically mimic the appearance of mesothelioma.^12,14  Other inflammatory conditions, diffuse pulmonary fibrosis, and adhesions may also give rise to benign mesothelial proliferations that radiologically mimic the appearance of mesothelioma. Furthermore, diffuse serosal growth patterns along the pleura can also be a feature of other neoplasms (termed pseudomesotheliomas) including carcinomas, sarcomas, and lymphomas.⁵⁹  As such, pathologic confirmation of mesothelioma is needed to facilitate appropriate patient management pathways.

The stage at presentation is one of the most important prognostic factors in mesothelioma. Therefore, early diagnosis can improve clinical outcomes for patients through earlier treatment. However, most patients are diagnosed late in the disease's course, limiting potential therapeutic options. Given that 70% to 95% of mesothelioma patients present with pleural effusions,^60,61  effusion cytology can improve patient outcomes by providing earlier definitive diagnoses, which allows patients to avoid more invasive thoracoscopic procedures (and their potential complications, including tumor seeding) for histologic sample collection, as well as unnecessary delays to treatment.⁸  However, making a definitive diagnosis of mesothelioma on cytology specimens has been historically challenging, and, as a result, clinicians did not have confidence in cytology's utility in mesothelioma workups. In 2009, consensus guidelines from the IMIG stated that a mesothelioma diagnosis should only be made on biopsy or histologic material to identify tissue invasion. However, there has been an acknowledgment of accumulating data highlighting how cytologic examination of pleural fluid may provide a safe and accurate alternative to tissue biopsy diagnosis, particularly with the help of ancillary testing.⁸  As a result, the 2012 and 2017 IMIG consensus statements have pointed to cytology's potential utility in clinical evaluation for mesothelioma. However, diagnostic difficulties due to mesothelioma's morphologic heterogeneity and the resultant overlap of cytologic features with benign mesothelial proliferations, in addition to its association with litigation, can still make pathologists reluctant to render a definitive diagnosis. As such, the spectrum of morphologic features of mesothelioma and use cases for ancillary studies will be reviewed.

In pleural effusion specimens in which there may be (1) an architecturally and/or cytologically atypical epithelioid population delineated as mesothelial in origin by immunohistochemistry, (2) numerous bland-appearing mesothelial cells, or (3) a high clinical suspicion for mesothelioma, the next phase of evaluation is to differentiate benign from malignant mesothelial proliferations. Studies comparing cytologic features in pleural effusions of mesothelioma cases with benign mesothelial proliferations showed that mesotheliomas are more frequently associated with the formation of “cell balls,” numerous 3-dimensional clusters with “knobby” cell group boundaries (Figure 4, A), and cell cannibalism manifesting in “cell in cell” arrangements, whereas benign mesothelial proliferations are most frequently associated (particularly in pelvic washings) with monolayer aggregates (Table 2). Some publications have noted the presence of acidophilic extracellular matrix cores within clusters of malignant mesothelial cells, known as collagen or basement membrane cores⁵  (Figure 4, B).

However, as many studies have noted, there can be significant overlap between reactive and malignant mesothelial proliferations, which can make morphologic differentiation on pleural effusion specimens extremely difficult.^14,62,63  Morphologic features such as high specimen cellularity in addition to the presence of a uniform cell population, multinucleation, hyperchromatic nuclei, coarse chromatin distribution, acinar/papillary structures, necrosis, and variable degrees of nuclear pleomorphism/cytologic atypia can be seen in both benign and malignant mesothelial proliferations.^12,18,49  Additionally, other morphologic features more commonly seen in mesothelioma can also be identified in benign proliferations (and vice versa). For example, while mesotheliomas frequently show malignant cells arranged in small to large clusters, some cases show a discohesive single cell dispersion pattern.⁶⁴  Also, although the presence of numerous small lipid-containing vacuoles⁵  and “collagen cores”⁶⁵  has been reported to be associated with mesothelioma, as per the previous section on normal mesothelial cell cytology, various etiologies of cytoplasmic vacuolization and “collagen balls” can be identified in benign mesothelial cell proliferations.

The degree of morphologic overlap contributes to the relatively low sensitivity of pleural effusion cytology in detecting mesotheliomas, which is lower than that of detecting other malignancies and typically ranges from 32% to 53% in published literature.^5,6  False-negative rates can be high and can be attributed to various factors that extend beyond interpretation errors.⁴⁹  In some cases, obscuring elements, such as blood or inflammatory cells, can preclude optimal assessment of the malignant population. In other cases, concurrent disease processes such as fibrinous pleuritis can prevent the exfoliation of tumor cells in mesothelioma and thereby preclude their cytologic evaluation on subsequent effusion sampling.¹⁸  Also, the subtype of mesothelioma can impact detection rates of pleural effusion cytology. Mesothelioma is subdivided into 3 subtypes: epithelioid, sarcomatoid, or mixed (biphasic).⁴⁹  These subtypes are differentiated not only by their distinct morphologic features but also by prognostic features (as the sarcomatoid and biphasic subtypes are associated with a poorer prognosis than the epithelioid subtype^66,67) and the likelihood of shedding into pleural effusion samples for cytologic evaluation (sarcomatoid subtypes have less tumor cell shedding, resulting in lower sensitivities for their detection on effusion cytology specimens^68,69).

When morphology fails to provide a clear distinction between mesothelioma and reactive mesothelial proliferations, ancillary studies can be useful, especially when they leverage information about the molecular characteristics of mesothelioma. Molecular alterations involved in the development of mesothelioma accumulate over several decades, leading to relatively unique sets of genetic alterations within individual patients. However, more detailed molecular characterization has highlighted that molecular profiles of mesotheliomas are different from those seen in metastatic adenocarcinoma^70,71  and often consist of multiple chromosome losses culminating in the loss or inactivation of tumor suppressor genes, particularly those in chromosomes 3p, 9p, and 22q, which correlate to loss of BRCA1-associated protein (BAP-1), p16INK4A-p14ARF (CDKN2A), and neurofibromatosis type 2 (NF2), respectively.^72–74  Pathology laboratories equipped to perform fluorescence in situ hybridization (FISH) assays can detect CDKN2A (p16) deletions in mesotheliomas with a high degree of specificity, reaching up to 100%.^75,76  While the reported sensitivity for detection in sarcomatoid mesotheliomas approaches 100%, the published sensitivity is lower for the more common epithelioid subtype, ranging from 33% to 86%.²³  In addition to this lower sensitivity in detecting epithelioid mesotheliomas, the utility of FISH can be limited given its time and resource requirements when compared to more widely accessible ancillary testing modalities such as immunohistochemistry. Although markers such as epithelial membrane antigen (EMA), desmin, and GLUT-1 have previously been used to help differentiate malignant and benign mesothelial proliferations,^76,77  their use was predominantly based on empiric evidence, with no known underlying molecular mechanisms to support their findings. In some studies, EMA, desmin, and GLUT-1 staining patterns associated with mesotheliomas (positive EMA, positive GLUT-1, and/or loss of expression of desmin) were seen in reactive mesothelial cells.^77,78  However, newer markers leveraging information about the elucidated genetic profile of mesothelioma have enhanced the diagnostic accuracy of pleural effusion cytology.

Methylthioadenosine phosphorylase (MTAP) is a gene close to CDKN2A at chromosome band 9p21 and is thereby codeleted in more than 90% of mesothelioma cases in which CDKN2A is deleted.⁷⁹  As such, loss of immunohistochemical MTAP expression has been showcased as an effective surrogate marker for deletion of CDKN2A (p16) by FISH^80,81  (Figure 5, A and B), surpassing the use of p16 immunohistochemistry.⁸²  While this has been noted to have a high degree of specificity (approaching 100%) in classifying mesothelial proliferations as malignant, its sensitivity has been variable, ranging from 46% to 86%.⁸³  Reports of higher sensitivity were seen in study cohorts with higher proportions of sarcomatoid subtypes of mesothelioma, which more frequently have CDKN2A deletions. Immunohistochemical loss of BAP-1 expression (Figure 5, A and C) (which correlates to biallelic mutations in BAP-1) has also been deemed a highly specific marker (with specificity ranging from 96% to 100%) for mesothelioma.^22,23,84,85  Loss of BAP-1 expression can also be seen in other tumors, including melanoma, renal cell carcinomas, and adenocarcinomas derived from the breast and biliary tract. Like MTAP, while the specificity of BAP-1 is quite high, the reported sensitivity is significantly lower. However, in contrast to MTAP, the sensitivity of BAP-1 in detecting epithelioid subtypes of mesothelioma (56%–81%) is typically higher than that for sarcomatoid subtypes (15%–63%).⁸³  Studies focused on BAP-1 immunohistochemistry performance on effusion cytology samples have reported a sensitivity of about 58% in the detection of mesothelioma.⁸⁶  While the individual sensitivities of MTAP and BAP-1 may be moderate, their combined use has been reported to yield significantly higher sensitivities of 75% to 80% for mesothelioma.⁸⁷  When immunohistochemical use of MTAP and BAP-1 is added to FISH for CDKN2A deletion, the sensitivity can increase to a range of 80% to 90%.⁸⁸ 

Additional immunohistochemical markers are under evaluation for their utility in differentiating mesothelioma from reactive mesothelial proliferations (Table 2), including Merlin and enhancer of zeste homologue 2 (EZH2).^89,90  Merlin is the protein encoded by neurofibromatosis 2 (NF2), which as mentioned previously, is frequently deleted in mesothelioma. FISH for hemizygous NF2 deletion has been reported to have approximately 50% sensitivity and 100% specificity for differentiating mesothelioma from benign mesothelial proliferations.^91,92  As such, studies have investigated the utility of detecting the immunohistochemical loss of Merlin expression in diagnosing mesothelioma. While an earlier study showed limited utility for its use,⁹³  another more recent publication leveraging a newer commercially available antibody clone has yielded more promising results (52% sensitivity, 100% specificity) for its use, especially when used in combination with MTAP and BAP-1.³³  EZH2 overexpression has been reported in several cancers, including those of prostatic, breast, and uterine primaries.⁹⁴  While there is limited literature available, reports have shown moderate sensitivity (45%–66%) and high specificity (approaching 100%) for high EZH2 expression patterns in immunohistochemistry.^95,96  Like Merlin, combining EZH2 with MTAP and BAP-1 can significantly enhance the sensitivity in detecting mesothelioma. Another marker that has shown promise in differentiating reactive from malignant mesothelial proliferations on histologic samples recently is 5-hydroxymethylcytosine (5-hmc).^32,97  While some studies have highlighted its use in combination with BAP-1 and MTAP,^97,98  additional studies will be able to further elucidate its effectiveness on cytologic samples.

In addition to immunohistochemical and FISH studies, the utility of other ancillary tests has been studied. Soluble biomarkers in effusion supernatants include mesothelin, soluble mesothelin-related peptides, and fibulin-3.^32,99,100  While chemistry assays may easily detect these biomarkers, they have not yet demonstrated high sensitivity in differentiating benign and malignant mesothelial proliferations on their own. However, they have been reported to enhance diagnostic sensitivity in the evaluation for mesothelioma when combined with other test results.^32,99  The utility of measuring these biomarkers in serum has also been evaluated as a potential screening modality for asbestos-exposed patients as well as a method to monitor response to treatment. An example of this is MESOMARK (Fujirebio Diagnostics Inc), an immunoenzymatic assay that measures soluble mesothelin-related peptides in blood samples. Studies have shown that MESOMARK's best use case is in monitoring for recurrence after initial treatment.¹⁰¹  Gene expression arrays have also been the focus of some publications, which have shown that they have a higher sensitivity and specificity than the performance of the combination of BAP-1 immunohistochemistry and CDKN2A FISH.¹⁰²  This has been a promising development, particularly noting the degree of morphologic overlap between benign and malignant mesothelial proliferations and the molecular heterogeneity of mesotheliomas (which may not display BAP-1 or MTAP loss by immunohistochemistry or CDKN2A deletion by FISH).

Various features of mesothelioma may play a role in patient prognostication. As mentioned previously, sarcomatoid mesotheliomas tend to have lower survival rates in comparison to epithelioid and biphasic subtypes. Additionally, they are less frequently identified on effusion cytology owing to decreased rates of tumor cell shedding. While epithelioid mesotheliomas generally have better clinical outcomes in comparison to sarcomatoid subtypes, architectural and cytologic features identified on histologic evaluation have been identified as unfavorable prognostic factors. These features include solid and micropapillary architectural growth patterns as well as rhabdoid and pleomorphic cytologic features.¹⁰³  World Health Organization and College of American Pathologists synoptic reporting systems now include a 2-tiered nuclear grading system for epithelioid mesotheliomas, based on scores reflective of the degree of nuclear atypia, mitotic rate, and presence of necrosis in the tumor. This grading system stratifies epithelioid mesotheliomas into low- or high-grade groups. High nuclear grade has been shown as an independent poor prognostic factor.^104,105  While these features are currently only reported in the histologic evaluation of biopsy or resection specimens, it may be important to take note of their potential prognostic implications in the cytologic evaluation of malignant mesothelial cells.

In addition to subtype and grading, there are other special considerations to note in the realm of pleural mesothelial proliferations. Well-differentiated papillary mesothelial tumor of the pleura (WDPMTP), previously classified as well-differentiated papillary mesothelioma, is a rare disease entity of uncertain etiology with an even lower prevalence than its peritoneal counterpart (well-differentiated papillary mesothelial tumor of the peritoneum). Its clinical course typically features slow growth and recurrence, with survival extending over the course of several years.¹⁰⁶  Histologically, it is characterized by a flat to thin papillary proliferation of bland mesothelial cells. Mitotic figures are rare to absent, and unlike mesothelioma, there is no evidence of invasion into underlying tissues. Given this disease entity's rarity, it is no surprise that there is scant literature available about its cytologic features in effusion specimens. Like its peritoneal counterpart, WDPMTP may result in numerous tubulopapillary and spheroid clusters of uniform mesothelial cells in effusion specimens.³⁰  However, a definitive diagnosis of WDPMTP is not feasible on cytologic evaluation, as a thorough histologic examination is required to exclude the potential of invasion (and subsequent diagnosis as mesothelioma). Nonetheless, it is important to recognize this as part of the differential considerations of mesothelioma on effusion cytology. In contrast to mesothelioma, BAP1 germline mutations in WDPMTP are rare,¹⁰⁷  and homozygous deletion of CDKN2A has not been observed.^108,109  This highlights the utility of ancillary studies on cytologic material with mesothelial proliferations, particularly in preventing the overdiagnosis of mesothelioma.

While it is unclear if WDPMTP represents a precursor lesion to mesothelioma, mesothelioma in situ (MIS) is clearly designated as a preinvasive single-layer proliferation of malignant mesothelial cells that can show a spectrum of cytologic atypia. Similar to WDPMTP, a definitive diagnosis of MIS is not feasible on cytologic assessment, as a thorough histologic evaluation for invasion is required. However, MIS is an important diagnostic consideration that can be mentioned as part of a differential, particularly in effusion cytology cases in which an atypical population of mesothelial cells with evidence of BAP-1 (by immunohistochemistry) and/or CDKN2A loss (by MTAP immunohistochemistry or FISH) is identified in the absence of suspicious radiologic findings.^110,111  With earlier identification of malignant mesothelial proliferations, treatment may be implemented before the development of an invasive component to positively impact patient prognosis. However, MIS can have bland cytologic findings that are similar to those of normal or reactive mesothelial cells. As such, pathologists have raised questions about what practice workflows should be generally applied for pleural effusion specimens with bland mesothelial cells. After all, evaluating each pleural effusion specimen with BAP-1 and MTAP immunohistochemistry and/or CDKN2A FISH may not be practical. Additionally, there are some malignant mesothelial cells that do not feature BAP1 and/or CDKN2A alterations. Therefore, ancillary studies may be more efficiently and effectively leveraged in patient cohorts with persistent pleural effusion(s) and/or high clinical suspicion for mesothelioma (eg, history of extensive asbestos exposure, radiation, and/or genetic predisposition).¹⁰³  However, a collaborative effort is required to collect sufficient data that inform optimal clinical practice in the cytologic evaluation of mesothelial proliferations.⁸⁵ 

Mesothelial cell proliferations in pleural effusion cytology can lead to diagnostic pitfalls due to misinterpretation of their nature and cell etiology. Recognizing the morphologic spectrum of both reactive mesothelial cell proliferations and mesothelioma can facilitate the exploration of differential diagnostic considerations and the appropriate use of ancillary studies. While older ancillary tests (particularly immunohistochemistry for EMA and desmin) have not been deemed reliable for accurate categorization of mesothelial proliferations on cytology, more recently available ancillary tests leveraging information about the elucidated genetic profile of mesothelioma (BAP-1, MTAP, and Merlin immunohistochemistry and CDKN2A FISH) can enhance the diagnostic accuracy of pleural effusion cytology in this realm, especially when used in combination. Leveraging suggested algorithmic approaches for morphologic evaluation and ancillary testing can facilitate the diagnostic categorization of mesothelial proliferations in cytology specimens.^62,63  The continual investigation of ancillary tests in addition to emerging immunohistochemical markers will continue to build a substantial artillery for future cytologic diagnoses of mesothelioma, which can impact patient survival through earlier detection and diagnosis⁹⁸  and reduce the need for more invasive procedures with a higher potential for morbidity and tumor seeding.