How to assess the best immunohistochemical panel in the diagnosis of malignant pleural mesothelioma in a pathology lab.

BACKGROUND: malignant pleural mesothelioma (MPM) is a rare tumor with a challenging diagnosis. Even if, clinical data are mandatory to suspect the diagnosis, the positive diagnosis is based on microscopic features. Morphologic features are still the port of call of the diagnosis but their non specific character and the multiplicity of differential diagnoses made the immunohistochemical markers mandatory for the diagnosis. Many antibodies with a positive diagnostic value including claretinin, mesothelin, WT1 and antibodies with a negative diagnostic value including TTF1, EMA, CD15 are recommended by the scientific societies. This is due to the diagnostic limits of every antibody which necessitate the association of multiple antibodies. In the diagnostic demarch, pathologists deal with different antibodies and clones. Even if many recommendations are available, every pathology lab has to experiment its own antibodies in order to optimize the routine diagnostic demarch especially in low-income country. Our aim was to assess the diagnostic value of different antibodies available in our lab and to recommend a decisional flowchart. PATIENTS AND METHODS: we conducted a retrospective study about 30 MPM diagnosed over a 20-year-period. The different techniques were realized manually. The different antibodies used were anti-calretinin, anti-Epithelial Membrane Antigen (EMA), anti-mesothelin, anti-Thyroid Transcription Factor 1 (TTF1), anti-ACE, anti-cytokeratin, anti-vimentin, anti-CD15, anti-cytokeratin 5/6, anti-bcl2, and anti-CD99 and anti-CD34 antibodies. The sensitivity and specificity of these antibodies were assessed. RESULTS: the microscopic exam concluded to an epithelioid mesothelioma (EM) in 17 cases, sarcomatoid mesothelioma (SM) in four cases and biphasic mesothelioma (BM) in nine cases. The immunohistochemical study was performed in all cases. A mean of eight antibodies was used in every case, average 4 to 20 antibodies. The immunohistochemical study was repeated from 2 to 5 times in 15 cases and concerned a mean of 3 antibodies per case. In EM and BM, the antibodies with positive predictive value and highest sensitivity were calretinin, EMA, cytokeratin, and vimentin reaching respectively a sensitivity of 86.2%, 89.7%, 92.9% and 89.3%. The most valuable antibodies with negative predictive value were TTF1, CD15 and ACE that presented a specificity reaching respectively 100%. In sarcomatoid mesothelima, the most sensitive antibody was the cytokeratin antibody. CONCLUSION: these results yielded to a diagnostic flowchart that we can use in routine practice and that is in accordance with the literature findings. Many diagnostic and technical pitfalls have to be known by pathologists when dealing with MPM.