Abstract 3938: Predicting first line tamoxifen response of recurrent ER+ breast cancer patients based on transcriptional activity of signaling pathways

Abstract

Abstract Introduction Generally, estrogen receptor positive (ER+) breast cancer patients are eligible for hormonal treatment, but roughly half of them respond, and identifying those that do remains a challenge. Predicting hormonal therapy response with higher specificity (without losing sensitivity) is of clinical value, to avoid overtreatment and enable consideration of alternative, more effective (targeted) drugs. We developed Bayesian computational model-based mRNA tests [1] to analyze oncogenic signal transduction pathway activity in cancer tissue samples, for selection of personalized therapy. Here, we show that such a test for the ER pathway can predict hormonal therapy response independent of traditional predictive factors. Methods Gene expression levels (Affymetrix HG-U133+PM microarrays) were used to determine transcriptional activity of the ER, AR, PI3K, Wnt, Hedgehog (HH), TGFβ and NFκB pathways, as described earlier [1], in 132 ER+ primary breast tumor samples of patients who did not receive adjuvant hormonal therapy and subsequently relapsed and were treated with first line tamoxifen treatment. Univariate and multivariate Cox proportional hazards regression was used to associate the pathways’ activities with progression free survival (PFS) on tamoxifen treatment, together with traditional response prediction factors. Results In 102 (77%) of the 132 samples, at least one of the seven pathways was found active. ER pathway activity was observed in 36 samples (27%). Strikingly, 76 samples (58%) had an active NFκB pathway. In 37 samples (28%), two or more pathways were active, predominantly ER with NFκB. The probability of ER pathway activity was associated with a longer PFS, with a hazard ratio of 0.41 in a univariate analysis (p = 0.0039), while activity of the TGFβ pathway was associated with a shorter PFS (HR = 2.2, p = 0.044). In a multivariate analysis using the seven pathways, ER pathway activity remained significant (HR = 0.44, p = 0.011), while TGFβ did not (HR = 1.67, p = 0.24). A Kaplan-Meier analysis as a function of ER activity gave a log-rank p-value of 0.002. A multivariate analysis for PFS including ER pathway activity and the traditional predictive factors disease free interval (<2yr vs. >2yr), site of relapse, age, menopausal status, ESR1 and PGR mRNA expression and Her2 status, showed that ER activity was independently associated with a favorable PFS (HR = 0.43, p = 0.047). Conclusion In ER+ breast cancer patients who did not receive adjuvant hormonal therapy, transcriptional ER pathway activity as measured by our computational ER pathway mRNA test was associated with a favorable outcome upon first line tamoxifen treatment. The ER pathway test can hence be used as an independent predictor of hormonal therapy response in ER+ breast cancer patients with recurrent disease. [1] Verhaegh W et al. Cancer Res. 2014;74:2936-45. Citation Format: Anieta M. Sieuwerts, Marcel Smid, Márcia A. Inda, John A. Foekens, Anja van de Stolpe, John W.M. Martens, Wim F.J. Verhaegh. Predicting first line tamoxifen response of recurrent ER+ breast cancer patients based on transcriptional activity of signaling pathways. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3938.