Estrogen receptors (ER) in human breast cancer. The significance of a new prognostic factor based on both ER protein and ER mRNA contents.

Abstract

The response to endocrine therapy is not entirely predictable from the estrogen receptor (ER) and progesterone receptor (PgR) status of primary breast tumors. The authors previously proposed a new prognostic factor, ER.R, which was based on both ER protein and mRNA levels. A previous analysis of 88 primary breast carcinomas showed that ER.R permits the identification of a subset of ER-positive women with a higher risk of early relapse. The purpose of the present study was to confirm the prognostic significance of ER.R. Estrogen receptor protein levels were determined for 171 patients with primary breast cancer either by radio-ligand binding assay (ER-LBA) or enzyme immunoassay (ER-EIA). Estrogen receptor, pS2, and c-erbB-2 mRNA were measured by Northern blot analysis. ER.R factor is determined by calculating the ratio of the values (ER protein in fentomoles per milligram of total proteins) to (ER mRNA in picograms per 4 micrograms of total RNA). A cutoff value of 1.5 (protein levels measured by ER-LBA) or 3 (protein levels measured by ER-EIA) discriminate the two ER.R1 (lower ratio) and ER.R2 (higher ratio) subgroups, which present a significantly lower and higher risk of early relapse, respectively. No association was found between ER.R status and either PgR status or c-erbB-2 and pS2 expression. According to a Cox multivariate analysis for disease free survival, the two stronger factors in predicting a poor prognosis were c-erbB-2 overexpression and ER.R2. In the present analysis, ER.R2 was a stronger predictor of recurrence than was ER negativity. In accordance with the authors' first published data, the analysis of a larger population with a longer follow-up showed that ER.R2 keeps its significance to predict a poorer outcome for a patient, regardless of which assay was used to quantify ER.