Immunohistochemical staining of estrogen and progesterone receptors: aspects for evaluating positivity and defining the cutpoints.

BACKGROUND Estrogen (ER) and progesterone (PgR) receptors are predictive and prognostic factors in breast cancer. The most suitable immunohistochemical cutpoints for dividing the tumors in hormone receptor-negatives and -positives may, however, need more consideration. We examined the association between breast cancer survival and cutpoints assessed by four different models. We looked for evidence for which patient subgroups could be handled best through applying different cutpoints. MATERIALS AND METHODS Three hundred and twenty-four samples of invasive breast cancer were immunohistochemically stained for ER and PgR, bcl-2 and erbB2. Fractions of ER- and PgR-positive cells and also ER and PgR staining scores were assessed. The fractions of stained cells and staining scores, respectively, were determined on the whole section area, and the area of most intense staining. Candidate cutpoints, dividing the patients into good and poor prognosis groups, were tested among all patients group, N+ and N- groups, premenopausal and postmenopausal patient groups. The correlation between immunohistochemistry results of ER, PgR, bcl-2 and erbB2 as well as SMI (standardized mitotic index), patient age, tumor size and axillary lymph node status were tested. RESULTS The ER score was correlated with age, SMI and bcl-2 positivity. The PgR score was correlated with erbB2 and bcl-2. Lobular carcinomas had higher staining scores of ER and PgR than ductal carcinomas. CONCLUSION In this material, ER was correlated with factors reflecting the differentiation of the tumor. On the basis of the ER and PgR immunohistochemistry cutpoint analysis, we found that the optimal cutpoints in different patient groups may not be the same.