Estrogen receptor (ER) content in breast cancer specimens is correlated with a prolonged disease free survival and increased likelihood of response to hormone therapy. Relatively few anti-ER antibodies are currently available for use in formalin-fixed, paraffin-embedded tissue. Recently, a new anti-ER monoclonal antibody (ERID5; AMAC, Westbrook, Maine) was generated which requires antigen retrieval by microwave oven heating for detection in routinely processed tissue. The specific aim of this study was to compare the ERID5 antibody with the commercially available rat monoclonal (ER-ICA; Abbott, Chicago, IL) which requires proteolytic enzyme digestion for detection in paraffin-embedded tissue. Sections from 20 cases of primary breast carcinoma previously assayed by dextran-coated charcoal (DCC) analysis were examined. Quantitation of ER antibody staining was performed without knowledge of the DCC values. Specimens containing > or = 20% specifically stained malignant cells were considered ER positive. The sensitivity and specificity of visual ER-ICA immunostaining were 57% and 83%, respectively. The sensitivity and specificity of visual ERID5 immunostaining were 93% and 50%, respectively. The predictive value of positive staining was 89% for the ER-ICA antibody and 81% for the ERID5 antibody. The predictive value of negative staining was 45% for the ER-ICA antibody and 75% for the ERID5 antibody. Previous studies have demonstrated a linear correlation between DCC values and the positive nuclear area (PNA) generated by image analysis for ER-ICA immunostaining. In the present study, a similar correlation between DCC value and ERID5 percentage PNA was observed (R = 0.670; P < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)