Demonstration of estrogen and progesterone receptors in breast cancers with monoclonal antibodies. Different results with enzyme-immunoassay and immunohistochemical methods.

Estrogen (ER) and progesterone receptors (PR) were measured in 120 breast cancer biopsies using the same tissue block and the same monoclonal antibody either immunohistochemically on frozen sections or quantitative by means of an enzyme-immunoassay (EIA) on the cytosol-fraction of a tumor homogenate. The immunohistochemical staining was performed by a standard PAP-method and the samples were classified according to the "immuno-reactive score" (IRS). IRS values > = 2 were regarded as receptor-positive; a cut-off-value of 15 fmol/mg protein was chosen for the EIA-test. In 90.8% of cases the findings of immunohistochemical and biochemical methods were concordant when only receptor-positivity or -negativity was regarded. Thus the results were discordant in 9.2%. Immunohistochemically positive and biochemically negative findings were observed more often (8.2% for the ER and 5.8% for the PR) than vice versa (0.8% for the ER and 3.3% for the PR). In all cases of discordant findings the relevant histological and immunohistochemical preparations were reevaluated. In total, there were 5 cases with a positive biochemical result which were classified as receptor-negative by immunohistochemistry. Of these, 3 had borderline findings by EIA together with a massive inflammatory infiltration. The other 2 cases must be interpreted as technical problems with the immunohistochemical staining. In 17 cases with a biochemically negative result receptor-positive cancer cells could be demonstrated by immunohistochemistry. In 13 of these 17 cases only a small fraction of tumor cells showed a positive reaction. Furthermore all these cases were rather acellular and partially necrotic. So the negative results of the enzyme immunoassay can be explained by a low concentration of receptor positive cells in the biopsy specimen or the measurements of necrotic tissue areas. It can be concluded that immunohistochemical receptor analysis is superior to methods using tumor-homogenates. The prognostic and predictive superiority of the immunohistochemical method is in accordance with published data. However, for immunohistochemical staining high internal and external quality standards must be applied as for similar laboratory procedures.