Key points are not available for this paper at this time.
The aims of this study were as follows: to confirm that the presence of estrogen (ER) and progesterone (PgR) receptors is an indicator of clinical behavior of human breast cancer independent of other known prognostic factors; to seek clinical correlates of those receptor values that best predict the overall disease evolution; and to determine the relative prognostic importance of PgR versus ER. The clinical records of 547 patients, the follow-up of some extending to 6 years, were analyzed in retrospect. Patients were placed into one of four disease stages and also in three groups according to the degree of axillary node involvement. In each group the prognostic value of receptors for patient survival, disease-free interval and, in case of metastasis or local recurrence, the response to endocrine treatment or chemotherapy were studied. The break point in the spectrum of receptor concentrations with regard to survival, disease-free interval, and response to treatment was greater than or equal to 20 for ER+, greater than or equal to 15 for PgR+, and less than 5 for both ER- and PgR- reported in fmol/mg protein. Survival and disease-free interval showed positive correlations with ER and PgR (P less than 0.001-less than 0.0003). When disease stage or node involvement were considered, these correlations were found essentially in Stage II and node involvement in more than three nodes, where patients had longer survival and disease-free interval if ER and PgR were positive (P less than 0.05-less than 0.0003). Estrogen receptor was a more sensitive prognostic indicator than PgR, and the combination ER+/PgR+ showed a correlation equivalent to ER+/PgR-. The correct prediction percentages of the response of patients to endocrine treatment were 77% if ER+, 69% if PgR+, and 79% if both ER+ and PgR+. However, the correct prediction percentage of the response to chemotherapy was of 50%. These results show that ER and PgR are prognostic factors for survival and disease-free interval mainly on patients at Stage II and node involvement of greater than three, with ER demonstrating a better predictive value than PgR. The measurement of these receptors provides a prognostic index for response to endocrine therapy but is without value in predicting the response to chemotherapy.
Vollenweider-Zerargui et al. (1986) studied this question.