IMMUNOHISTOCHEMICAL DETECTION OF ESTROGEN AND PROGESTERONE RECEPTORS IN FORMALIN-FIXED, PARAFFIN-EMBEDDED TISSUES AFTER MICROWAVE TREATMENT- COMPARISON WITH BIOCHEMICAL ASSAY IN A SERIES OF 123 BREAST CARCINOMAS WITH DETERMINATION OF THE POSITIVITY CUTOFF

The use of immunohistochemical detection of Estrogen receptors (ER) an d Progesterone receptors (PR) in formalin-fixed, paraffin-embedded tis sue after microwave treatment is growing but is still not sufficiently validated in relation to classic biochemical techniques. The anti-ER and anti-PR monoclonal antibodies, clone ID5 and 1A6 respectively were applied. The immunohistochemical cut-off which is the closest to rite biochemical reference laboratory level was investigated by comparing the percentage of stained cells to the biochemical assay (Radio-Ligand Assay, DCC, cut-off 10 fmol/mg cytosol protein). The balance of value s values of concordance, specificity, sensitivity, positive and negati ve predictive value calculated for a series of 123 invasive mammary ca rcinomas by varying the immuno-histochemical cut-off between I and 55 % of stained, invasive carcinomatous cells allowed us to choose the im munohistochemical qualitative cut-off of 10 % of positive positive car cinomatous cells for ER and PR. Immunohistochemical detections were in qualitative agreement (+/-) with the biochemical assay in 89.5 % of t he cases for RE and 75% for RP (specificity 88 % for ER 78 % for PR, s ensibility 90 % for ER and 73 % for PR). Staining quantification was a lso performed by a score (IHC Score) which is the sum of the grade att ributed to tile % of stained cells and the grade attributed to the deg ree of staining. A highly significant correlation exists between the I HC score and tile biochemical concentration (Spearman correlation coef ficient for ER : r = 0,73, p<10(-4), and for PR : r = 0,65 p<10(-4)). Ln conclusion, with a positivity cut-off of 10% of stained carcinomato us cells, immunohistochemical detection of RE and PR is fast, reliable , nor expensive and clinically useful, particularly if interlaboratory reproducibility and quality controls are being developed.