RÉSUMÉ
The effect of race on breast cancer outcome is confounded by tumor and treatment heterogeneity. We examined a cohort of women with stage II-III breast cancer treated uniformly with neoadjuvant chemotherapy to identify factors associated with racial differences in chemotherapeutic response and long-term survival. Using a prospective database, we identified women with stage II-III breast cancer treated with neoadjuvant chemotherapy from 1998 to 2011. Race was categorized as African-American (AA) or non-AA. Preplanned subtype analyses were stratified by hormone receptor (HR) and HER2. Pathologic response to chemotherapy (pCR), time to recurrence (TTR), and overall survival (OS) were assessed using logistic regression, Kaplan-Meier method, and Cox proportional hazards regression analyses. Of 349 women identified, 102 (29 %) were AA, who were younger (p = 0.03), more obese (p < 0.001), and less likely to have HR+/HER2- tumors (p = 0.01). No significant differences in pCR rate by race were found. At median follow-up of 6.5 years, AA had worse TTR (hazard ratio 1.51, 95 % CI 1.02-2.24), which was attenuated in multivariable modeling, and there was no significant difference in OS. When stratified by HR, worse outcomes were limited to HR+AA (TTR hazard ratio 1.85, 95 % CI 1.09-3.14; OS hazard ratio 2.42 95 % CI 1.37-4.28), which remained significant in multivariable analysis including pCR rate and BMI. With long-term follow-up, racial disparity in outcome was limited to HR+ breast cancer, with no apparent contribution of chemotherapy sensitivity. This suggests that disparity root causes may be driven by HR+ factors such as unmeasured molecular differences, endocrine therapy sensitivity, or adherence.