Impact of standardization of estrogen and progesterone receptor assays of breast cancer biopsies in Denmark.

Estrogen and progesterone receptor (ER and PgR, respectively) data for the three laboratories participating in the Danish Breast Cancer Cooperative Group (DBCG) project for treatment of primary breast cancer are presented for the period 1979-1986. The frequency of ER positivity remained constant for one laboratory throughout this period, while this value changed significantly in the other two laboratories. Inter-laboratory reproducibility (evaluated as the frequency of ER positivity) was poor at the onset of the project (P = 0.0003) but, due to standardization procedures, improved after 1982. Slight but significant differences in the composition of the patient populations at the three centers (menopausal status and tumor size) may account for some of the differences observed both in frequency of PgR positivity as well as PgR concentrations determined. Greater intra- and inter-laboratory differences were observed in all three laboratories for PgR than for ER. Part of this variation is believed to have been alleviated by the addition of 10 mM sodium molybdate to the assay buffer in 1983. From having very divergent frequencies of PgR positivity in the three laboratories (31-71%) in 1981, this divergence has been reduced in 1985 (62-78%). While data regarding ER status significantly distinguish between patients with long versus short recurrence-free survival (irrespective of treatment) in one laboratory in both the 77 and the 82 generations of clinical protocols within the DBCG program, ER status from the other two laboratories makes this distinction only in the 82 protocols. We attribute this inability of ER status to distinguish among patients in the 77 protocols to the suboptimal nature of the assays performed in these two laboratories at that time. The overall improvement in comparability of data from the three laboratories during the standardization procedures as well as the fact that ER status from all three laboratories is now capable of distinguishing different patient groups attest to the fact that standardization procedures are both necessary and useful.