SIMULTANEOUS IMMUNOHISTOCHEMICAL AND BIOCHEMICAL HORMONE-RECEPTOR ASSESSMENT IN BREAST-CANCER PROVIDES COMPLEMENTARY PROGNOSTIC INFORMATION

The prognostic value of the biochemical and the immunohistochemical as sessment of estrogen-and progesterone receptor (ER, PR) status was tes ted in 111 breast cancer patients, mostly focusing on whether the resu lts reveal complementary prognostic information. The biochemical recep tor analysis was performed on snap-frozen tumor tissue using a standar d protocol (ER-DCC, PR-DCC). The immunhistochemical staining was done on 4 mu m thick paraffin sections and was evaluated semiquantitatively (ER-IHC, PR-IHC) and immunohistometrically by means of image analysis (ERMEAN, PRMEAN). 74% of the ER-DCC and 50% of the PR-DCC assays were interpreted as positive. The positivity rates of the immunohistochemi cal reactions ranged between 78% and 81% for ER and between 66% and 82 % for PR, depending on the interpretation mode. The concordance rate f or the DCC method was 68%, and ranged between 77% and 85% for the immu nohistochemical results on paraffin sections. ER-DCC and PR-DCC showed a better survival for receptor-positive patients; however, this tende ncy was only statistically significant for the PR-DCC (p = 0.0294). Pa tients with immunohistochemically determined ER-or PR-positivity revea led a significantly better survival than receptor-negative patients, t he effect being stronger for the progesterone receptor (ER: p = 0.0253 , PR: p = 0.0005). Combining the different methods and receptors in a multivariate analysis, we observed that a) ER and PR reveal complement ary prognostic information to each other after immunohistochemical det ermination (p less than or equal to 0.0018) and that, b) complementary prognostic information was also obtainable by comparing the biochemic al and the immunohistochemical PR-analysts (p less than or equal to 0. 0084); slightly more significant results were obtained for ERMEAN and PRMEAN compared to ER-IHC and PR-IHC. Considering the lymph node statu s and a combined receptor analysis (PR-DCC, ERMEAN, PRMEAN) as the two strongest prognosticators in multivariate Cox models, the combined re ceptor analysis was able to discover for each of the three groups of N O- and N1-patients different survival probabilities (p<0.0001). In con clusion, the ER-DCC appears to be dispensable in all patients. In lymp h node-negative patients, the PR-DCC has no outstanding merit, indicat ing that the neccessity of this method is also controversial. In prima ry tumors of lymph node-positive patients, however, all three remainin g types of receptor analysis should be evaluated for their therapeutic implications.