Background: Historically, estrogen receptor (ER) determinations have been made by the ligand-binding assay of tumor homogenates, primarily by the dextran-coated charcoal method (DCC). Immunocytochemical assays (ICA) for ER are more recent and have been executed mostly on frozen sections with the monoclonal antibody H222Sp gamma (H222). Lately, new monoclonal antibodies derived by recombinant ER technology have been developed that work well on paraffin embedded, formalin fixed tissue sections. However, there is little information as to whether such assays prognosticate endocrine response.
Methods: Using antigen retrieval, the immunoglobulin G1 monoclonal antibody ER1D5, and the streptavidin-biotin detection system, 74 patients with breast cancer in whom endocrine response was known were assayed and the results compared with ER by DCC and ER by ICA in frozen section with H222.
Results: ER1D5 in paraffin provided the highest correlation with endocrine response (Kendall's tau [r] = 0.57; P<0.001) whereas ER by DCC failed to correlate (r= -0.002; P<0.99). ER1D5 in paraffin correlated weakly though significantly with DCC (Kappa Statistic [K] = 0.204; P<0.02). H222 in frozen sections also correlated moderately with endocrine response (r = 0.34; P<0.001).
Conclusions: ER can be detected in routine tissue sections processed with antigen retrieval and ER1D5, and can be relied upon to provide accurate prognostic information regarding response to endocrine therapies in breast cancer patients.