Oestrogen and progesterone receptor determinations in breast cancer: technology and biology

Abstract

Our present state of knowledge regarding oestrogen and progesterone receptors (ER and PgR, respectively) has led to changes in treatment strategies: patients without receptors in their tumour tissues cannot be expected to respond to endocrine therapy; furthermore, groups of patients with specifically good or poor prognoses can be selected. Treatment of the disease is now starting to be rational rather than empirical. However, it is imperative that we achieve a greater level of understanding before we can predict with high probabilities which patients will benefit from endocrine therapy. Only through a coordinated effort by many centres can we hope to attain this goal. In such collaboration several factors must be considered if reproducible conclusions with respect to results from receptor assays are to be reached: sampling of the tumour biopsy for analysis, potential differences in assay procedures which may affect results and the composition of the population studied. Differences between centres in any of these aspects may affect the study conclusions. The causes of and direction of possible biases due to these factors will be discussed here. Approximately 90% of all patients with primary breast cancer have been registered in the national Danish Breast Cancer Cooperative Group (DBCG) project. ER and PgR determinations have been performed on tumour tissue from approximately 30% of these patients in a single laboratory. The results of these analyses are presented here for 3735 patients in relation to age, menopausal status, tumour size, grade of anaplasia, lymph node involvement and parity. Approximately 30% of these patients have been in DBCG protocols in which no adjuvant systemic therapy has been administered. Because of this large number of cases, the recurrence-free survival of the untreated disease in relation to receptor status can be reliably analysed. Both ER status and PgR status were found to be significant prognostic variables for premenopausal women under 50 years of age. In contrast, in the postmenopausal women neither ER nor PgR status was a significant prognostic factor in the low risk group (tumour less than 5 cm, no lymph node involvement). In high risk, postmenopausal women, however, ER status is a significant prognostic factor for recurrence-free survival and it appears to be independent of lymph node status.