The use of histological and immunohistochemical markers to distinguish pleural malignant mesothelioma and in situ mesothelioma from reactive mesothelial hyperplasia and reactive pleural fibrosis

Distinguishing malignant mesothelioma from reactive mesothelial hyperplasia and reactive fibrosis can be a diagnostic problem in small pleural biopsie s, made more difficult by the recent recognition of mesothelioma-in-situ. A ntibodies to epithelial membrane antigen (EMA), p53, and bcl-2 have all bee n advocated for differentiating reactive from neoplastic conditions, but re ports are inconsistent. These antibodies have therefore been applied to 31 cases of malignant mesothelioma, 34 cases of reactive pleural disease (20 r eactive mesothelial hyperplasia and 14 reactive pleural fibrosis) and four small biopsies that were initially coded as suspicious, from patients who l ater developed frank mesothelioma. Thirty out of 31 (97 per cent) cases of mesothelioma showed positive nuclear staining for p53, with a higher incide nce of positivity in epithelioid than in sarcomatoid elements and 30/31 (97 per cent) showed diffuse linear membrane staining for EMA, again more inte nse in the epithelioid elements. No mesothelioma was positive for bcl-2. In seven cases that contained both in sire and invasive mesothelioma, the in situ elements showed similar staining patterns to the invasive epithelioid elements. Thirteen out of 20 (65 per cent) cases of reactive mesothelial hy perplasia showed occasional nuclear positivity for p53 and 5/20 (25 per cen t) cases showed focal weak membrane staining for EMA. Three out of 14 (21 p er cent) cases of reactive pleural fibrosis showed positive nuclear stainin g for p53 and 6/14 (43 per cent) cases showed focal membrane staining with EMA. No reactive cases stained for bcl-2. All four suspicious cases showed diffuse linear staining with EMA and three showed focal staining for p53. I t is concluded that strong diffuse linear staining for EMA is a good marker of malignancy when differentiating epithelioid malignant mesothelioma and mesothelioma-in-situ from reactive mesothelial hyperplasia, although weak f ocal staining may occur in reactive conditions. Nuclear staining for p53 is also suggestive of epithelioid mesothelioma, but should be regarded as no more than suspicious. The antibodies used in this investigation are less he lpful in differentiating sarcomatoid mesothelioma from reactive pleural fib rosis. Copyright (C) 1999 John Wiley & Sons, Ltd.