Malignant Mesothelioma Diagnosis

In a biopsy, which is first determined to be adequate, the histologic findings must be first correlated with clinical and radiologic features. The key indicator of malignancy remains invasion of preexisting tissue, particularly adipose tissue.^6–9  One must be careful not to mistake “fake fat,” which is an artifact and is S100 negative (Figure 1, A and B), for real invasion into preexisting adipose tissue.¹⁰  Two antibodies, glucose transporter 1 (GLUT-1) and insulin-like growth factor II messenger RNA–binding protein 3 (IMP3), have recently been shown to have utility in differentiating benign from malignant mesothelioma, independently from tissue invasion. GLUT-1 shows focal positivity in 67% of mesotheliomas and in 3% of reactive lesions, while IMP3 shows more diffuse positivity in 73% of mesotheliomas and negativity in reactive lesions.^1,11,12 

The most common genetic alteration in malignant mesothelioma is homozygous deletion of the 9p21 locus within a cluster of genes that includes CDKN2A, CDKN2B, and MTAP. p16/CDKN2A deletions are reported in up to 70% of primary epithelioid and 90% to 100% of sarcomatoid pleural mesotheliomas. The presence of this homozygous deletion is the best marker of malignancy in a mesothelial lesion, since it has not been reported so far in any of the benign lesions.¹³  The fluorescence in situ hybridization (FISH) assay using a commercially available dual-color FISH probe (Abbott Molecular, Des Plains, Illinois) can be reliably performed on archival paraffin-embedded tissue and it is relatively less expensive than other molecular assays.

Immunohistochemistry for loss of p16 protein expression does correlate with p16 deletion but it can be positive in FISH-proven homozygous deletion and it can be negative in the absence of deletion. The staining pattern depends on the type of antibody, assay conditions, and interpretation criteria. Therefore, immunohistochemistry is not recommended as a surrogate method for detection of p16 deletion.¹⁴ 

The microscopic features of various histologic patterns have been well described in the past. Recently, the clinical and prognostic significance of many of these patterns have been demonstrated. Two studies have shown that the pleomorphic pattern (Figure 2), when present in more than 10% of tumor, portends a highly aggressive behavior and a poor survival similar to that of sarcomatoid mesotheliomas.^2,3  In one study of 232 patients,²  the pleomorphic subtype was associated with the worst survival, followed by solid, micropapillary, tubulopapillary, and trabecular patterns. The micropapillary pattern was significantly associated with lymphatic invasion, which has also been demonstrated in other tumors such as lung carcinoma. A study of 21 deciduoid mesotheliomas⁴  has shown that the high-grade subgroup (showing wide variation in size and shape of cells, frequent loss of cell cohesion, marked nuclear atypia, and mitotic activity >5 per 10 high-power fields) had a much more aggressive clinical course, when compared to those deciduoid tumors that were more cohesive, less pleomorphic, and had low mitotic activity (Figure 3, A and B).

A study of 232 patients from a single institution has proposed a nuclear grading system for epithelioid mesothelioma irrespective of histologic pattern.⁵  Nuclear atypia and mitotic count proved to be independent prognostic factors and were used to create a 3-tier nuclear grade, which correlated with MIB-1 labeling index and was an independent predictor of overall survival. Further studies are needed to validate this grading system.

All variants of epithelioid mesothelioma react with multiple antibodies used to diagnose mesothelioma. This has been verified recently in series of mesothelioma cases with rare histologic patterns, including small cell, signet ring, adenomatoid, clear cell, and with crystalloid structures.^15–19  There is still a wide variation among laboratories as to which antibodies are selected for testing. It is recommended that, when all clinical, radiologic, and histologic features are concordant, at least 2 mesothelial and 2 carcinoma markers with greater than 80% sensitivity and specificity be used. Of course, additional markers would be indicated when any of the above features are discordant. The most useful mesothelial markers are calretinin, WT-1, cytokeratin 5/6 (CK5/6), and D2-40 (podoplanin). The sarcomatoid component of biphasic tumors and pure sarcomatoid mesotheliomas may lose immunoreactivity for most markers in a majority of the cells; however, calretinin and D2-40 are more likely to remain immunoreactive. The most useful general carcinoma markers are MOC31, BG8, carcinoembryonic antigen, and BerEp4. The reader is referred to the consensus statement from the International Mesothelioma Interest Group¹  as well as another recent review²⁰  for further details.

The differential diagnosis between malignant mesothelioma and metastatic carcinoma to the pleura varies with the histologic pattern of mesothelioma, which guides the selection of immunohistochemistry panel. For differentiating lung adenocarcinoma, thyroid transcription factor-1 (TTF-1) and napsinA are most useful; squamous cell carcinoma is p40 positive while WT-1 stains positively in malignant mesothelioma (note: squamous cell carcinoma is positive for calretinin and CK5/6). For renal cell carcinoma, PAX8 or PAX2 nuclear positivity, and for ovarian carcinoma, PAX8 nuclear positivity, differentiates them from malignant mesothelioma. CDX2 is a good nuclear marker for gastrointestinal carcinoma. Estrogen and progesterone receptors are helpful, if positive markers, in breast carcinoma. Genetic alterations such as deletions of 9p are some of the most frequent events in other nonmesothelioma tumor types including non–small cell carcinomas and sarcomas of the lung. Therefore, 9p deletions alone cannot be used to differentiate these neoplasms from mesothelioma.