P1-07-02: Discordance between Central and Local Laboratory HER2 Testing from a Large HER2−Negative Population in VIRGO, a Metastatic Breast Cancer Registry

Background: HER2 overexpression is associated with unfavorable prognosis and is reported in 18–25% of breast cancers (BC). HER2 testing is often performed using immunohistochemistry (IHC) or fluorescence in situ hybridization (FISH). Because of the significant benefit of HER2−directed therapies, it is critical to accurately identify women whose tumors are HER2+. Reports have noted discordance between HER2+ test results from local vs. large reference labs in patients with HER2+ BC evaluated for trastuzumab-based clinical studies. There are little published data on central testing of BC found to be negative locally. Patients and Methods: VIRGO is an observational cohort of N=1,287 women with primarily HER2−negative metastatic BC. An optional tissue collection substudy was conducted, and 776 patient samples were received and centrally retested. Central testing was performed at 2 reference labs and tumors were deemed HER2+ if IHC 3+ and/or FISH positive (HER2:CEP17 ratio ≥2.0). Tumors with unknown/missing local HER2 status (n=68) were excluded from primary analyses. Number of patients potentially affected based on BC incidences from the American Cancer Society (ACS) 2011 estimates and the World Health Organization (WHO) 2008 report were calculated. Testing on the remainder of the HER2−negative cohort is in process. Results: Central retesting has been performed on tumor samples from n=373 patients to date: HER2−negative locally evaluable tumors (n=301), n=4 HER2−negative locally with no evaluable tumor, and HER2 unknown (n=68). A total of 301 unique patient samples were included in the primary analysis. Of these, 15 (4.98% [95% CI (2.7%, 7.9%)] were found to be HER2+ by central testing (Table). Based on sensitivity analyses assuming all 68 tumors with unknown HER2 status to be negative locally, 4.07%(15 /369) would be centrally HER2+. Of the 15 HER2+ tumors, 4 tumors tested positive centrally by both IHC and FISH; 6 IHC positive/FISH negative; and 5 FISH positive/IHC negative. 14/15 tumors were tested locally by only one testing methodology, and 11/15 were determined to be HER2+ centrally based on the testing methodology not performed locally. Investigators for all 15 patients have been notified of central HER2 testing results. Conclusion: Based on ACS estimates of 232,620 new cases of invasive BC diagnosed in the US in 2011 (assuming 80% testing HER2−negative); a discordance rate of 4–5% equates to 7,444 - 9,305 patients’ tumors diagnosed as HER2+ by central testing. Based on WHO global BC incidence estimates, 44,274 - 55,342 patients could be impacted worldwide as reported in this study. Inaccurate HER2 testing has significant clinical impact, both in denying appropriate treatment or leading to inappropriate use of HER2−targeted therapies. This study suggests testing by both IHC and FISH may be of benefit to accurately identify HER2 status, consistent with the Herceptin® USPI. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-07-02.