Abstract PD02-01: Her2 expression measured by AQUA analysis on BCIRG-005 and BCIRG-006 predicts the benefit of Herceptin therapy

Abstract

Abstract Introduction: There have been disparate results reported in breast cancer testing for HER2 assessment as measured by protein expression or DNA amplification, yet both tests are routinely used to prescribe the drug Herceptin (trastuzumab, Genentech, So San Francisco, CA). Typically, immunohistochemistry (IHC) staining intensity of 3+ or FISH copy ratio of ≥2.0 are used to establish the cutoff between a negative and a positive result. However, it is unclear whether positivity is correlated with differential response to therapy. We used Automated Quantitative Analysis (AQUA) and a fluorescent immunohistochemical assay to measure HER2 expression in cases scored by central laboratory FISH and also receiving Herceptin therapy. The intentions of the study were two-fold: first, to provide further validation of the AQUA technology as applied to the clinical measurement of HER2 expression in breast cancer and second, to examine the potential of drug response stratification within those patients that are considered positive. Methods: AQUA fluorescence IHC staining was performed on a multi-cohort tissue microarray (TMA) set. The assay was constructed in the Genoptix CLIA laboratory per ASCO/CAP guidelines and with a cutpoint that was validated against IHC (with FISH reflex). The trial specimens tested were from the BCIRG-005/006 studies. BCIRG-005 had n=1544 cases all assessed as FISH- while the 006 cohort had n=1477 cases all assessed as FISH+. Disease free survival (DFS) was used as the variable in subsequent modeling and analysis. Results: The BCIRG 005 and 006 cohorts, examined in aggregate, allowed for an initial examination of agreement relationships between HER2 levels as assessed by AQUA scoring and HER2 levels as assessed by central lab FISH. Results indicated a 77% negative agreement, a 97% positive agreement and an 87% overall concordance agreement for a total of n=3021 cases. Additional Cox modeling of the patients that were enrolled as FISH+ and stratified for those who did or did not receive Herceptin treatment demonstrated a significant overall hazard ratio (HR = 0.75, CI=0.60,0.93) and when stratified for response to Herceptin, cases determined to be positive by AQUA showed significant benefit from treatment (HR = 0.66, CI = 0.52,0.85) in contrast to those who were scored as negative by AQUA that did demonstrate benefit from therapy (HR = 1.19, CI=0.71,1.97). Conclusions: Analysis of the cases in this study originally determined to be HER2+ by FISH indicates that AQUA may improve predictions of which patients will benefit from Herceptin therapy. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr PD02-01.