Assessment of immunocytochemical and histochemical stainings in the distinction between reactive mesothelial cells and adenocarcinoma cells in body effusions.

BACKGROUND

The accumulation of fluid in body cavities is a common event in both neoplastic and non-neoplastic diseases. However, the distinction between cells of reactive process and those of malignancy in cytology is not always possible. It is especially difficult when reactive mesothelial cells and adenocarcinoma cells are encountered. The aim of the present study was to find out the most accurate or reliable immunocytochemical and histochemical stains to distinguish reactive mesothelial cells from adenocarcinoma cells, and to serve as a standard method in the future when dealing with equivocal cases.

METHODS

Ninety-nine cases of malignant epithelial effusion were collected from 755 cases of effusion obtained from 3 large body cavities in the past one-year period. Among them, 71 cases were histologically as adenocarcinoma and 13 cases as non-adenocarcinoma. The other 15 cases were carcinoma proved by image as well as clinical symptoms and signs. These 99 cases, plus 10 cases of non-malignant effusion, underwent immunocytochemical and histochemical stainings. Five common commercial antibodies used in this immunocytochemical study were epithelial membrane antigen (EMA), carcinoembryonal antigen (CEA), cytokeratin, vimentin and Leu-M1. The histochemical study included periodic acid-Schiff diastase (D-PAS) and mucicarmine stains.

RESULTS

The immunocytochemical study showed that EMA had a high frequency of positive staining with malignant epithelial cells and a negative staining with mesothelial cells. Cytokeratin always stained with malignant epithelial cells but it also stained with mesothelial cells. Almost all the proved adenocarcinomas expressed CEA which was not expressed in the proved non-adenocarcinomas. However, CEA occasionally stained mesothelial cells. Leu-M1 showed a low frequency of staining with the proved adenocarcinoma cells, but it did not stain proved non-adenocarcinoma cells and mesothelial cells. On the contrary, vimentin stained with all mesothelial cells but occasionally with malignant epithelial cells, especially the proved adenocarcinoma cells. In the histochemical study, both mucicarmine and D-PAS showed a low sensitivity but high specificity in detecting the adenocarcinoma cells.

CONCLUSIONS

No single marker is absolutely reliable to distinguish exfoliated, reactive mesothelial cells from adenocarcinoma cells in effusions. However, a panel of 3 antibodies containing EMA, CEA and vimentin, together with D-PAS and mucicarmine stains may help solve this problem.