Oestrogen receptor assay by ER-D5 immunocytochemistry fails to correlate with ligand-binding assay in breast cancer.

Results obtained with a conventional reference ligand-binding assay were compared with those from an immunocytochemical method employing ER-D5 antigen, a supposedly oestrogen-receptor-related protein. In a group of 144 histologically proven breast carcinomas, no direct correlation could be established. Only 52% of cases would have corresponded in diagnosis. The incidence of false-negatives by ER-D5 immunocytochemistry is of the order of 70% while 32% of patients were false-positive by this method. Since both methods are employed to predict therapeutic responses to hormonal manipulation, the discrepancy between the methods is alarming. While the prognostic and predictive index of conventional oestrogen-receptor assays is well documented, correlation of these assays with ER-D5 immunocytochemistry has not been demonstrated. On the basis of these results, and pending further investigation, the replacement of conventional oestrogen-receptor methods by immunocytochemistry using this antibody is not recommended.