Comparison of histo- and biochemical methods for estrogen and progesterone receptor analysis in primary human breast carcinoma.

One hundred and twenty-three primary breast carcinomas have been examined under optimal procedural conditions for presence and quantity of estrogen (ER) and progesterone (PgR) receptors by a histochemical method using hormones conjugated to fluoresceinated bovine serum albumin and by a conventional biochemical (dextran-coated charcoal) assay. The semiquantitative histochemical results for ER and PgR were strongly positively correlated, as were the corresponding quantitative biochemical results, but there was no significant correlation between histo- and biochemical results for ER, nor between the corresponding results for PgR. By generally accepted cut-off limits for positivity there was an equal frequency of ER- and PgR-positive tumors by the histochemical method (71 and 67% respectively), whereas there was a significantly higher frequency of ER- than PgR-positive tumors by the biochemical method (76 and 59%). The qualitative agreement between the two methods was 76% for ER and 61% for PgR, and there was a correspondingly inferior sensitivity and specificity of the histochemical method in comparison with the biochemical method, whose value is clinically validated.