Measurement of estrogen and progesterone receptors in human breast tumors: enzyme immunoassay versus binding assay.

To determine whether we could replace our current binding assay (BA) method for measurement of estrogen receptors (ERs) and progesterone receptors (PRs) with the recently developed enzyme immunoassay (EIA) method, we compared simultaneous measurements of ERs and PRs in frozen breast tumor samples by both methods. A value of greater than or equal to 10 fmol/mg cytosol protein was defined as positive. There was agreement between the BA and EIA on whether the sample was positive or negative in 75 of 91 (82%) samples measured for ERs and in 74 of 93 (80%) for PRs. When the threshold value for a positive assay was redefined as greater than or equal to 20 fmol/mg protein, there was agreement in 85 of 91 (93%) samples for ERs and 85 of 93 (91%) for PRs. The numerical value for ERs by EIA was not consistently greater or less than ERs by BA, but the difference between the EIA and BA measurement increased as the size of the measurement increased. We did not see an excess of premenopausal patients whose ERs by BA were negative and whose ERs by EIA were positive. Although we performed a linear regression analysis and determined the Pearson correlation coefficient to compare the BA and EIA as reported by others, we show that this analysis may be misleading when the objective is to demonstrate similarities between these methods. Our study shows that the EIA can be confidently used in place of the BA. However, a threshold value for a positive EIA should be confirmed clinically in future studies.