Objective: It was investigated whether the immunocytochemical analysis of pleural effusions and ascites with a new antibody can improve sensitivity and specificity in the demonstration of carcinoma cells.
Material and methods: Cytological and immunocytochemical analyses were performed on 111 unselected consecutive aspirated samples (from 97 patients) of pleural effusions (73) and ascites (38). They were stained with Ber-EP-4 antibodies against epithelial antigen and leukocyte common antigen. Information indicated the presence of malignancy underlying 59 of the effusions.
Results: Native cytology demonstrated tumour cells in 21 of these 59 samples (36%), while Ber-EP-4 positive cells were present in 34 (58%). 41 samples were judged as "suspect" by native cytology, 29 of which were actually malignant, while 16 out of 48 samples "without evidence of tumour" were actually known to be malignant. Demonstration of tumour cells succeeded by immunocytochemical analysis in 10 of the 41 judged "suspect" by native cytology and in four of the 48 "unremarkable" samples. There were no false-positive reactions with mesothelial cells or macrophages: no Ber-EP-4 reactivity occurred in the 45 benign effusions and the seven malignant, but not epithelial, ones. Tests with one LCA antibody was of no additional diagnostic value, but proved valuable as positive controls of the staining technique.
Conclusion: These results indicate that analysis of effusions with the Ber-EP-4 antibody has a higher specificity and sensitivity than native cytology in the demonstration of carcinoma cells.