Comparison of Evaluations for Hormone Receptors in Breast Carcinoma Using Two Manual and Three Automated Immunohistochemical Assays

The aims of this study were to compare the quality of immunohistochemical assays of estrogen receptor (ER) and progesterone receptor (PR) and to compare the intermethod variability of assays from different manufacturers. immunohistochemical staining was entrusted to the following laboratories in Japan: Kyowa Medex, dealing with the products of BioGenex (Mishima, Shizuoka), DAKO Japan (Kyoto) and Ventana Japan (Yokohama). All slides were semiquantitatively evaluated according to the Allred score. Intermethod variability showed fair to moderate multirater kappa values for ER and PR (for total score, ER, kappa = 0.34; PR, kappa = 0.45). Another scoring system was also applied in which, irrespective of the intensity of nuclear staining, the proportion of cells stained in each specimen was recorded as 0; less than 1%; 1% or more and less than 10%; or 10% or more. Intermethod variability showed substantial multirater kappa values for ER and PR (according to percentage of positive cells, ER, kappa = 0.67; PR, kappa = 0.72). Concerning intermethod consistency, the scoring system based on the percentage of positive cells was advantageous over other scoring systems.