[A comparative evaluation of immunohistochemical markers for the differential diagnosis between malignant mesothelioma, non-small cell carcinoma involving the pleura, and benign reactive mesothelial cell proliferation].

INTRODUCTION Histopathological diagnosis of malignant mesothelioma (MM) and differentiating it from tumors infiltrating the pleura is very difficult. Distinguishing benign reactive mesothelial cell proliferation from MM also presents problems. The objective of this study was to evaluate the significance of selected immunohistochemical stains in differentiating MM from non-small cell lung cancers infiltrating the pleura and from benign reactive mesothelial cell proliferation. MATERIAL AND METHODS The material encompassed 86 cases of MM, 54 cases of NSCLC infiltrating the pleura, and 43 cases of benign reactive mesothelial cell proliferation. The MM cases were reclassified according to the WHO criteria (2004): epithelioid, 61 cases (71%), including well-differentiated papillomatous, 3 cases; sarcomatous, 6 cases (6.8%); fibrous, 4 cases (4.7%); biphasic, 15 cases (17.5%). A panel of immunohistochemical stains was used in this study. It included broad-spectrum antibodies to cytokeratins (CKAE1/AE3, CKMNF116), vimentin, epithelial membrane antigen (EMA), mesothelial cells (HBME1, CK5/6, calretinin), adenocarcinoma cells (BerEp4, B72.3, CEA, TTF1), antibodies enabling the assessment of proliferation (Mib1) and cell-cycle regulating proteins (p53). RESULTS Coexpression of cytokeratins and vimentin was found in 63.9% of MM cases and cell-membrane reactions with EMA were seen in 58.9%. Positive staining for HBME1, CK5/6, calretinin, BerEp4, B72.3, CEA and p53 was obtained in 76.7%, 51.2%, 66.7%, 1.2%, 6.2%, 1.2% and 51% of the cases, respectively. None of the MM cases stained for TTF1. MM by WHO subgroups: Coexpression of cytokeratins and vimentin occurred in 55.7% cases of epithelioid MM, 93.3% of biphasic MM, 66.6% of sarcomatous MM, and in 100% of fibrous MM cases. Positive staining for HBME1, CK5/6, and calretinin was seen only in the epithelioid and mixed subtypes of MM; the respective percentages of positive reactions were: HBME1, 90.2% and 73.3%; CK5/6 58.2% and 53.3%; calretinin, 72% and 75%. Non-small cell lung cancers infiltrating the pleura: Coexpression of cytokeratin and vimentin was found in 17.6% of the cases, positive staining of membranes for EMA, in 13% cases. Positive staining for HBME1 was observed in 22.6% of the cases, for CK5/6, in 9.3%, for calretinin, in 2%, for BerEp4, in 72.2%, for B72.3, in 64.1%, for CEA, in 58.5%, and for TTF1, in 43.8%. Benign reactive mesothelial cell proliferation: Protein p53 was present in 9.3% of cases, whereas no positive staining for EMA was found. Differentiation of MM from non-small cell carcinomas: Among the antibodies used in the study, anti-HBME1 had the highest sensitivity (76.7%) but lowest specificity (77.4%). Staining for calretinin showed high specificity (99.8%), as did CEA and TTF1 (98.8% and 100%), with moderate sensitivity (66.7%, 58.5% and 43.8%, respectively). BerEp4 showed the highest sensitivity (72.2%) and specificity (98.8%). CONCLUSION In diagnosing mesothelioma it is necessary to use a panel of immunohistochemical stains, which should contain antibodies to markers for adenocarcinoma and mesothelioma. Due to the high costs of such a study, a two-stage method is advantageous. The best combination of sensitivity and specificity was found for BerEp4, CEA, and TTF1 and for calretinin and HBME1. In the diagnosis of spindle-cell pleural tumors and the fibrous form of MM and benign reactive mesothelial cell proliferation , markers of mesothelial cells are noncontributory. Immunohistochemical staining fails to identify a reactive process, but a diffuse, positive stain for EMA and the presence of protein p53 support the diagnosis of MM.