Standardization of steroid receptor assays in human breast cancer--I. Reproducibility of estradiol and progesterone receptor assays.

Four different lyophilized cytosols were analyzed for estrogen and progesterone receptor content by 13 laboratories. Three of the cytosols were assayed on 9 consecutive working days. The fourth cytosol was analyzed 3 times. All laboratories used the same methods of receptor and protein assays. Estradiol and progesterone receptors were measured by Scatchard analysis with centrally provided radioactive ligands and employing charcoal adsorption of free steroid. Protein was assayed by the Bradford technique with the same lots of Coomassie brilliant blue and human serum albumin standard. All participating groups except one produced receptor results (fmol/ml cytosol) with less inter-laboratory variation than in previous trials. Recalculation of the raw data of all participants by a common computer program further reduced this inter-laboratory variation. The discrepancies between the reported and recalculated receptor binding data are discussed. The intra-laboratory variation was sometimes surprisingly high and occurred at random. Single-dose saturation assays showed good agreement with multipoint Scatchard assays for the high receptor-positive samples, while poor agreement was observed for the heat-inactivated, receptor-negative sample. The use of common reagents and methodologies diminished the inter-laboratory variation coefficients of the protein assay to 14-17%; however, the protein estimation still needs to be improved.