**3. Cytological diagnosis of mesothelioma**

#### **3.1 Cytological features of mesothelioma in routinely stained smears**

Evaluating the cytomorphology of pleural and peritoneal effusions in routinely stained smears enables in most cases to identify malignant cells and suspicious for malignancy. In either case, to discriminate reactive proliferative mesothelium from mesothelioma and other malignancies, ancillary IC studies are required (see later). Some cases cannot be diagnosed by cytology like cases with minimal cell shedding, typically almost all sarcomatoid mesotheliomas. However, sarcomatoid mesothelioma can be overlaid by the reactive epithelioid mesothelial cells, which may readily shed into fluids and mislead the pathologist. Sarcomatoid mesothelioma can be successfully diagnosed only histologically by using core biopsy (or larger tissue samples) [21]. Since the cells in effusion are exfoliative from the tumor surface, and cytology material is lacking access to the deep structures, assessment of invasion of preexisting tissues and its correlation to the clinical and imaging findings are not possible.

Cytological features of mesothelioma are outlined in abundance for pathology specialists [9], but this information is based on histologically confirmed retrospective studies. There is significant overlap between mesothelioma, reactive mesothelial cells, and adenocarcinoma or anaplastic tumors [8, 22]. Also, a rare WDPMT has considerable cytological overlap with mesothelioma [23–25].

**Figure 1** represents an example of the peritoneal fluid cytology with confirmed epithelioid mesothelioma by later histological studies. The basic general cytomorphological criteria indicating possible mesothelioma are: (1) material containing large numbers of mesothelial cells, including large ball-shaped or papillary cell aggregates with knobby outlines (scalloped borders) and (2) presence of overtly malignant cells, either as single cells or in tissue fragments [9].

The malignant mesothelial cells can be significantly larger than normal, and each of the components of the whole cell is enlarged: cytoplasm, nucleus, and nucleolus.

#### **Figure 1.**

*Cytomorphology of the peritoneal epithelioid mesothelioma in effusion. Peritoneal effusion cytospin in epithelioid mesothelioma stained with Leishman-Giemsa (A) and Papanicolaou (PAP) stain (B), original magnification ×400. The specimen is highly cellular, containing large cell cluster (A) and papillary-shaped aggregates (B). Large mesothelial cells with macronucleoli and multinucleated cells (A and B).*

*Mesothelioma: Overview of Technical, Immunochemical and Pathomorphological Diagnosing… DOI: http://dx.doi.org/10.5772/intechopen.106570*

The cells may be multinucleated, contain prominent macronucleoli or there are vacuoles overlapping with cell nuclei. Protrusions from the cell membrane or blebbing and prominent degree of cell-within-cell arrangements are also characteristics. Background may be acidophilic due to large amounts of hyaluronan and contain granular extracellular matrix fragments of collagen and basement membrane cores, as well as multinucleated giant cells and small pyknotic eosinophilic cells [9].

#### **3.2 General aspects of immunochemistry**

Effusion cytology work-up mostly faces discrimination of epithelioid mesothelioma since sarcomatoid subtype rarely exfoliates in the fluids. The recent guidelines of epithelioid mesothelioma cytologic diagnosis and reporting emphasize the role of IC in conjunction with the cytomorphologic evaluation because it substantially increases diagnostic accuracy [9, 21]. IC on cell blocks is mandatory whenever a diagnosis of malignancy is clinically entertained and/or cytologically suspected [21].

There is no fixed IC panel or absolute number of antibodies that can be recommended for the diagnosis of mesothelioma. Workup can be done in stages. It is recommended that a panel of at least four antibodies should be used, two in favor and two against mesothelioma. The diagnosis should never be based on one single IC reaction. Numerous antibodies for mesothelioma are commercially available, but most are not entirely specific and may show cross-reactivity with other tumors [9]. It has to be emphasized that only validated antibodies should be used for clinical diagnosis and different antibody clones have to be carefully tested with appropriate controls in the labs. If possible, antibodies should be chosen with a sensitivity or specificity of at least 80% [9]. The staining patterns (i.e., nuclear, cytoplasmic, and membranous) are important for most antibodies, and since these may differ with the new antibody clones, up-to-date information has to be followed and the tests performed with appropriate controls. There is no standard for the percentage of tumor cells that should be positive, but some have used a 10% cutoff for membranous and cytoplasmic staining [9]. IC results should be interpreted in complexity and in the context of morphological and clinical data. Of notice, the cell blocks can be also used for molecular studies, which is beyond the scope of this review.

#### **3.3 Immunochemical workup of mesothelioma**

The antibodies used for mesothelioma IC workup are largely similar in effusion cell blocks and in histological tissue blocks, however, some extra advice is added for antibody application in tissues.

The diagnosis in effusions is more challenging, comprising the following tasks: 1) confirm the mesothelial or epithelial origin of isolated atypical cells and cell clusters; 2) delineate their benign or malignant nature; and 3) discriminate mesothelioma from other malignancies and metastatic disease, which can show diffuse pleural or peritoneal spread.

Summary of the most widely clinically used IC markers will be given and illustrated by the examples in **Figure 2**. For the rest of markers, only brief references are given [8]. The paraffin-embedded cell blocks are sectioned and stained similarly to the histological specimen and, therefore, a routine hematoxylin and eosin (H&E) staining is also applied, which provides additional cytomorphological evaluation (**Figure 2A**).

#### **Figure 2.**

*Malignant mesothelioma in peritoneal fluid cytoblock. A staining panel confirming mesothelial origin, malignancy, and discriminating from gastrointestinal and gynecologic tumors. All antibodies are applied as ready-to-use (RTU) solutions, the producers are shown in the brackets. A, H&E stain to assess cytomorphology: Highly cellular specimen, enlarged atypical cell aggregates, with hyperchromatic pleomorphic nuclei and vacuolated cytoplasm could be seen (original magnification ×400). B, Calretinin expression both in nuclei and cytoplasm (Ventana, RTU, ×400). C, WT1 specific staining is nuclear (Ventana, RTU, ×400). D, D2–40 strong membranous expression (Dako, RTU, ×400). E, BAP-1 shows nuclear loss of expression in mesothelioma cells, whereas reactive mesothelial cells and background lymphocytes retain nuclear staining (BioSB, RTU, ×400). F, CEA negative (Dako, RTU, ×400). G, Ber-Ep4 negative with minimal nonspecific stain (Dako, RTU, ×400). H, CDX2 negative in mesothelioma cells (nonspecific background stain) (Dako, RTU, ×400).*

#### *3.3.1 Markers used to confirm mesothelial origin*

Markers of mesothelial cells are immunoreactive with both benign and malignant cells.

*Mesothelioma: Overview of Technical, Immunochemical and Pathomorphological Diagnosing… DOI: http://dx.doi.org/10.5772/intechopen.106570*

#### *3.3.1.1 Calretinin*

The recent Calretinin antibodies (**Figure 2B**) require both nuclear and cytoplasmic staining to support a diagnosis of mesothelioma [26]. There are earlier reports of only nuclear staining with "fried egg appearance" [27, 28]. Cytoplasmic staining alone should be interpreted negatively [27]. In effusions, the sensitivity of calretinin in detecting mesothelioma ranges from 81 to 100% [26, 29, 30].

Calretinin can be expressed in breast carcinomas [31], and a weak cytoplasmic staining is reported in variety if other adenocarcinomas [27, 28]. Some studies have shown calretinin positivity in squamous cell carcinoma (SCC) of the lung ranging from 40 to 100% [27, 32].

### *3.3.1.2 Wilms tumor-1 (WT1)*

Specific WT1 staining in mesothelioma is only nuclear (**Figure 2C**). WT1 frequently cross-reacts with cytoplasmic proteins in a variety of benign and malignant entities [33]. WT1 nuclear reactivity was reported in more than 90% of mesothelioma effusion specimens versus 20−30% of metastatic adenocarcinomas, particularly of pulmonary and breast origin [34–36]. In contrast, WT1 is not useful to distinguish peritoneal mesothelioma from ovarian/Mullerian tumors in effusions, since it is expressed in 80%−90% of ovarian malignancy [35, 37], and of notice, not recommended as a carcinoma-specific marker of these tumors either [8].

### *3.3.1.3 D2-40/podoplanin*

D2–40 and podoplanin are specific lymphatic endothelial markers [38].

D2–40 immunostain shows strong membranous staining pattern in mesothelial cells (**Figure 2D**), with reported sensitivity of 83−100% and specificity of 49−100% [30, 39, 40].

Podoplanin has been shown to be even more specific than D2–40, but the number of studies is limited. Podoplanin is expressed in 94% of mesothelioma, 97% of reactive mesothelial cells, and 7% ovarian adenocarcinoma, while it is nonreactive in lung and breast adenocarcinoma, with an overall sensitivity and specificity of 94% and 97%, respectively, for mesothelioma [38]. While podoplanin showed strong membranous reactivity in mesothelioma cells, ovarian adenocarcinoma exhibited weak membranous staining [38].

#### *3.3.2 Markers differentiating benign from malignant mesothelial proliferations*

Many of the markers supposedly differentiating mesothelioma from benign reactive mesothelial cells have limited sensitivity or a too broad spectrum of reactivity. For example, relevance of EMA, p53, IMP-3, CD146, or glucose transporter 1 in defying benign and malignant cases is questioned, especially in histology materials [21].

#### *3.3.2.1 BRCA1-associated protein (BAP1)*

BAP1 is a nuclear ubiquitin hydrolase involved in various cellular processes, including chromatin remodeling. *BAP1* behaves as a true tumor suppressor gene. *BAP1* double-hit inactivation is a key driver event in about half of all mesotheliomas [41, 42]. Loss of BAP1 expression by IC can be a useful adjunct to distinguish mesothelioma from reactive mesothelial proliferations in some cases [43]. However, BAP1 is not very sensitive, with a reported loss of nuclear staining only in 27–57% of mesothelioma but in none of the reactive mesothelial cells [41, 42]. For correct interpretation, only nuclear loss of staining is accepted as true loss of expression [8]. Reactive mesothelial cells and background lymphocytes should express nuclear staining and can serve as internal control (**Figure 2E**).

BAP1 use has more limitations since it is preserved in many non-mesothelial malignancies, frequently encountered in effusion cytology, and BAP1 loss may be also encountered in other malignancies rarely seen in effusions such as malignant melanoma and urothelial carcinoma [44].

### *3.3.2.2 Enhancer of zeste 2 homolog (EZH2)*

EZH2 is a member of the family of polycomb group genes (PcGs), which is a group of important epigenetic regulators that repress transcription. BAP1 loss can promote cell proliferation *in vitro* through up-regulation of EZH2 [45]. High EZH2 expression was observed in 66% of malignant mesothelioma cases, whereas none of the benign lesions showed high EZH2 expression. The combination of BAP1 loss and high EZH2 expression as markers to differentiate epithelioid/biphasic malignant mesothelioma from benign mesothelial lesions was highly sensitive (87−90%) and specific (100%) [46, 47]. Using IC alone for EZH2 also yielded a good sensitivity of 86.9%; this level is high enough for routine diagnostics [47].
