Abstract B031: Prediction of the clinical benefit of adjuvant tamoxifen therapy using Breast Cancer Index (HOXB13:IL17BR)

Abstract

Abstract Background: Breast Cancer Index (BCI) is a gene expression-based biomarker with a novel mechanism of action that provides individual risk assessment of both early (0-5 years) and late (5-10 years) distant recurrence in ER+, node negative (LN-) breast cancer patients. BCI is an algorithmic combination of two complementary biomarkers: Molecular Grade Index (MGI), which recapitulates tumor grade/proliferation status; and HOXB13:IL17BR ratio (H/I), which interrogates estrogen signaling pathways. In a previous analysis of the randomized MA.17 trial, H/I was predictive of extended endocrine therapy benefit (Sgroi et al., JNCI 2013). The current study examined whether H/I predicts benefit of adjuvant tamoxifen (TAM) versus untreated (UNT) in ER+ LN- patients from the prospective, randomized, placebo-controlled Stockholm trial. Materials and Methods: Formalin fixed paraffin embedded (FFPE) tumor blocks from 808 LN- patients were analyzed with 37 cases excluded due to insufficient tumor content (N=771). The primary endpoint was distant recurrence, with 15-year clinical follow-up data. H/I was calculated and risk groups determined using a pre-specified cutpoint. Kaplan-Meier analysis and Cox proportional hazards regression with an interaction term between the treatments and pre-defined H/I risk groups were examined for prediction of benefit of TAM vs UNT. The strength of the relationship between magnitude of TAM benefit and H/I as a continuous variable was assessed by the likelihood ratio test. Results: Of 771 specimens assayed, RT-PCR was successful for 769 cases, among which 600 were from ER+ tumors (317 TAM, 283 untreated). Using the pre-specified cutpoint, 230 patients (38%) were classified as having high H/I ratio while 370 (62%) had a low H/I ratio. High H/I was predictive for benefit from TAM therapy, with a decrease of 17.4% in absolute risk of distant recurrence (90.6% [95% CI: 85.1-96.3%] for the TAM arm versus 73.1% [95% CI: 65.0-82.2%] for UNT arm) and a relative risk of 0.33 (95% CI: 0.17-0.63; p < 0.0001) as measured by hazard ratio (HR). No significant reduction in distant recurrence was observed in patients with low H/I [91.6% (95% CI: 87.7-95.6%) for the TAM arm versus 88.2% (95% CI: 83.3-93.5%) for the untreated arm] and a HR of 0.67 (95% CI: 0.35-1.28; p = 0.204). Multivariate analysis demonstrated that the benefit of TAM was significant in patients with high H/I (HR 0.33; 95% CI 0.17-0.62; p = 0.0007) after adjusting for age, tumor size, grade, PR and HER2 status. A multivariate analysis of Cox models with and without an interaction term between H/I and TAM treatment by likelihood ratio test showed that the interaction was significant (p = 0.003). Consistent with the significant interaction, the 10-year rate of distant recurrence as a function of continuous H/I for the 2 treatment arms demonstrated that the reduction in the rate of distant recurrence by TAM increases as the H/I ratio increases. Conclusions: This prospective-retrospective study of ER+ LN- breast cancer patients from the Stockholm trial demonstrated that H/I was predictive of benefit from adjuvant TAM treatment. The results from this study and the recent data demonstrating that H/I predicts benefit from extended letrozole therapy in the MA.17 trial, indicates that H/I is a predictive biomarker for both adjuvant and extended adjuvant endocrine therapies and further supports the clinical utility of BCI (H/I) for predicting patient benefit of extended endocrine therapy. Citation Format: Yi Zhang, Brock E. Schroeder, Piiha-Lotta Jerevall, Olle Stal, Dennis C. Sgroi, Mark G. Erlander, Catherine A. Schnabel. Prediction of the clinical benefit of adjuvant tamoxifen therapy using Breast Cancer Index (HOXB13:IL17BR). [abstract]. In: Proceedings of the AACR Special Conference on Advances in Breast Cancer Research: Genetics, Biology, and Clinical Applications; Oct 3-6, 2013; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Res 2013;11(10 Suppl):Abstract nr B031.