The prognostic significance of steroid receptor activity in tumor tissues of patients with primary breast cancer.

Abstract

The prognostic significance of steroid-receptor activity is still debatable. Discrepancies in results are probably attributable to few patients, heterogeneous patient populations, and short follow-up. We investigated the prognostic significance of estrogen- and progesterone-receptor (ER and PgR, respectively) activity as a continuous variable in a homogeneous patient population. The prognostic significance of steroid-receptor activity was examined in 329 node-negative and 320 node-positive unselected breast cancer patients. In node-negative patients, ER values of primary tumors between 100 and 400 fmol/mg protein appeared to be a significant predictor for low risk of recurrence, whereas high ER (> 400) revealed an unfavorable prognosis. The classic cutoff level of ER (< 10 fmol/mg proteins) had no prognostic significance, however. In patients receiving adjuvant chemotherapy--the node-positive breast cancer patients--the classic cutoff value of ER (10 fmol/mg protein) predicts significantly distant metastases-free survival and overall survival only in the first 4 years of follow-up after diagnosis. Progesterone receptor is a time-dependent prognosticator in node-negative breast cancer patients (cutoff point for PgR, 80 fmol/mg). In node-positive breast cancer patients treated with chemotherapy or a combination of chemo- and hormonal therapy, PgR values lower than 60 fmol/mg had a worse prognosis. The results show the poor performance of standard cutoff points for ER and PgR positivity in predicting prognosis. Better prognosis is related to higher receptor levels but this relation is predominantly time-dependent. Moreover, patients who have high ER levels have a prognosis that is worse when compared with intermediate ER levels. Standard cutoff points for steroid receptors should not be used to select patients for prognosis.