HER-2/neu gene amplification compared with HER-2/neu protein overexpression and interobserver reproducibility in invasive breast carcinoma

We compared the detection of HER-2/neu gene amplification by fluorescence in situ hybridization (FISH) with detection of HER-2/neu protein overexpression by immunohistochemistry using 2 antibodies on 100 archival invasive breast carcinomas. Protein overexpression for each marker was scored independently by 4 pathologists using standardized criteria, and consensus was compared with results obtained from gene amplification. The concordance rate between FISH and immunohistochemistry was 76% for e2-4001 and 91% for the HercepTest. Of the 37 cases positive by e2-4001, 21 demonstrated no gene amplification; 7 of 24 cases positive by the HercepTest demonstrated no gene amplification. However, 1 of 61 cases negative by e2-4001 showed gene amplification; none of the cases negative by the HercepTest showed amplification. The predictive values of gene amplification based on 0-1+, 2+, and 3+ immunohistochemical staining were best for cases scored as 3+ (75% for e2-4001 and 89% for the HercepTest). Complete agreement among observers for immunohistochemical scoring of e2-4001 and the HercepTest was achieved in 75 and 85 cases, respectively. The pairwise kappa agreement values were substantial for e2-4001 and substantial to almost perfect for the HercepTest. Immunohistochemical staining may be considered a useful screening test. While negative staining almost always correlated with a lack of gene amplification, positive membranous staining, especially 2+, did not predict gene amplification. The low interobserver reproducibility in separating 2+ from 3+ cases necessitates further confirmation by FISH before treatment decisions are made.