P2-14-01: Race, Response to Chemotherapy, and Outcome within Clinical Breast Cancer Subtypes.

Abstract

Abstract Introduction: Racial disparity in breast cancer (BC) outcome has been attributed to access to care, socioeconomic factors, and biologic factors. In particular, the prevalence of triple negative BC is significantly higher among African-American (AA) women than white women (non-AA). However a recent population-based study suggests that racial disparities persist even if stratified by subtype, although it is possible that bias in stage at presentation may have influenced these results. In this study, we examined 464 women of uniform clinical stage II-III breast cancer, all treated with neoadjuvant chemotherapy, and examined pathologic complete response (pCR), recurrence-free survival (RFS), and overall survival (OS) to see if racial disparity exists within a more homogeneous stage and treatment cohort. Methods: The UNC Neoadjuvant Breast Cancer Database is a prospectively maintained cohort with clinical and outcome annotation. We identified women diagnosed between 1991–2011 and assigned BC subtype by clinical assay for estrogen receptor and progesterone receptor (hormone receptors, HR) and HER2. Fisher's Exact and Wilcoxon Rank sum tests compared characteristics between AA and non-AA. The Kaplan-Meier method and Log Rank test compared OS and RFS curves. Results: 142 (31%) of 464 women were AA, 294 (63%) were white, and 28 (6%) were other races (for analysis white and “other” were categorized as non-AA). 191 (41%) were clinical stage II and 273 (59%) stage III at diagnosis. Stage at diagnosis did not differ between AA and non-AA (p=0.07). AA were slightly older (median age 49 v 45, p=0.01) and were more likely to be HR-/Her2- (44% v 25%, p=0.003). All patients received preoperative chemotherapy. AA received less biologic therapy such as trastuzumab or on-protocol bevacizumab (10% v 18%, p=0.03) and received significantly less endocrine therapy (46% v 59%, p=0.01) than non-AA patients. 102 (22%) of patients had a pCR to neoadjuvant chemotherapy; pCR rates did not differ significantly by race (p=0.7) or within subtype by race (all p>0.14). At a median follow-up of 4.5 years, 121 (26%) of patients have died. Among HR-/HER2− (n=114) and HR-/HER2+(n=47) patients, OS and RFS did not differ between AA and non-AA (all p>0.5). Conversely, among HR+/HER2− patients (n=159) AA had worse OS (p=0.01) and RFS (p= 0.03) and among HR+/HER2+ patients (n=53), AA had worse OS (p=0.0004) and RFS (p=0.005). Conclusion: In this cohort of clinical stage II-III BC patients, AA women were over 75% more likely to have triple negative BC than non-AA patients. Pathologic reponse to chemotherapy did not differ between AA and non-AA patients. Examining the entire cohort, AA women had worse RFS and OS, however there was no difference among ER-negative subtypes. The difference in RFS and OS was driven by worse outcome among women with ER or PR+ tumors (regardless of HER2). It is possible that this racial disparity in outcome among hormone receptor-positive BC patients reflects unknown biologic differences by race, or differences in access to or receipt of endocrine therapy. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-14-01.