Optimization of a battery using nine immunocytochemical variables for distinguishing between epithelial mesothelioma and adenocarcinoma

APMIS. 1997 Nov;105(11):889-94. doi: 10.1111/j.1699-0463.1997.tb05099.x.


A battery of immunocytochemical analyses, previously established to distinguish between malignant mesothelioma and metastatic adenocarcinoma, was extended by analysing the same cases with three other commercially available antibodies. Altogether, 11 antibodies were studied in mesotheliomas diagnosed by other means, using 14 different immunocytochemical parameters. Logistic regression analysis indicated that the following parameters were of importance for this diagnostic problem: vimentin reactivity in epithelial cells (1), cytokeratin (CAM 5.2) reactivity in spindle-shaped (fibrous) cells (2), cell membrane-associated reactivity of EMA (3), HBME-1 (4) and thrombomodulin (5), and absence of reactivity to CEA (6), CD15 (7), BerEp4 (8) and Sialyl-TN (9). The analysis gave an algorithm with which a specific diagnosis of mesothelioma could be made in 80% of the cases-i.e., some improvement compared to the 55% sensitivity using the previous battery.

MeSH terms

  • Adenocarcinoma / diagnosis*
  • Antigens, Neoplasm / analysis
  • Humans
  • Immunologic Techniques
  • Mesothelioma / diagnosis*
  • Neoplasm Metastasis
  • Pleural Neoplasms / diagnosis
  • Pleural Neoplasms / secondary
  • Regression Analysis


  • Antigens, Neoplasm