Abstract

HER2 overexpression in breast cancer is associated with worse clinical outcome. To select patients for anti-Her2-based therapy immunohistochemistry is commonly performed as a first step to assess Her2 status. However, interobserver and interlaboratory variability can significantly compromise adequate assessment of Her2 status. In addition, immunohistochemistry does not always result in an unambiguous test result requiring additional testing for Her2 gene amplification. This study aimed to improve the reliability of Her2 immunohistochemistry by using rabbit monoclonal antibody 4B5 as an alternative to mouse monoclonal antibody CB11 routinely used in our laboratory. Therefore, 283 breast adenocarcinomas were included in a tissue microarray. Immunohistochemistry using the 4B5 and CB11 antibodies, and fluorescence and chromogenic in situ hybridization (FISH or CISH) were performed. Immunohistochemistry was scored by two independent investigators. We found that 4B5 staining was more distinct than CB11 staining. For CB11 staining, there were 12% (BV) and 5% (JW) 2+ scores compared with 4% (BV) and 2% (JW) for 4B5. There was a strong trend towards higher interobserver agreement for 4B5 compared with CB11 (4B5: kappa 0.87, 95% CI 0.79-0.96; CB11: kappa 0.77, 95% CI 0.66-0.88). There were no significant differences in sensitivity, specificity and predictive values between CB11 and 4B5. Our results indicate that the 4B5 antibody provides more robust assessment of immunohistochemical Her2/neu status and will reduce the number of gene amplification tests compared with CB11. However, for tumours with a 2+ score additional gene amplification measurement using FISH or CISH remains necessary.