Bimodal frequency distribution of estrogen receptor immunohistochemical staining results in breast cancer: an analysis of 825 cases.

Immunohistochemical analysis is used routinely to determine the estrogen receptor (ER) status of breast cancers in paraffin sections. However, lack of standardization has raised concerns that weakly ER+ tumors often are classified erroneously as ER-. To determine the frequency of weakly ER+ tumors, we reviewed ER immunostains of 825 breast cancers. For each case, we estimated the proportion of ER+ tumor cells and also determined an Allred score (which results in scores of 0 or 2 through 8, based on staining intensity and proportion of positive cells). In 817 cases (99.0%), tumor cells showed complete absence of staining or staining in 70% or more of the cells. Similarly, 818 cases (99.2%) exhibited Allred scores of 0 or of 7 or 8. Thus, with the immunohistochemical method used in our laboratory, ER staining is essentially bimodal. The overwhelming majority of breast cancers are either completely ER- or unambiguously ER+, and cases with weak ER immunostaining are rare.