Variation of receptor status in cancer of the breast.

One hundred and nineteen patients with breast cancer had 2 or more lesions removed for oestrogen (REc) or progesterone receptor (RPc) assay, either synchronously (on 38 occasions) or after an interval (on 91 occasions). In all but 7 both receptors were assayed for each lesion. The assays did not agree on the presence or absence of REc alone, RPc alone or the combination of both receptors in 11, 13 and 16% respectively of the synchronous samples, compared with 23, 30 and 43% of the asynchronous samples. The differences between the synchronous and asynchronous samples were significant for the combined receptors (P = 0.007) but not for REc (P = 0.176) or RPc alone (P = 0.077). Variation between asynchronous biopsies was greater when the earlier lesion contained RPc (18/37 disagreed) than when it did not (8/50) disagreed, P = 0.0023). This was not true for oestrogen receptor. In those remaining receptor positive there was only a weak correlation between the first and second values (Spearman rank correlation coefficient, rho = 0.39 for REc, P less than 0.02, and 0.45 for RPc, 0.05 less than P less than 0.1). Receptor levels and receptor status may change with time. Biopsy is most appropriate at the time when systemic treatment is proposed.