617 Background: Triple-negative breast cancer (TNBC) exhibits marked racial disparities in outcomes. However, the contributions of genetic ancestry, self-reported race, and structural determinants remain unclear, particularly in admixed populations. We aimed to disentangle the independent and combined impact of these factors in a Brazilian cohort. Methods: We retrospectively studied 156 women with TNBC treated with neoadjuvant chemotherapy at a tertiary center in São Paulo, Brazil (2007-2022). Genetic ancestry was estimated using ancestry-informative markers (AIMs) from an unpublished dataset. Self-reported race followed national census categories. Healthcare system type (public/private) served as a proxy for socioeconomic status. Primary outcomes were pathologic complete response (pCR), overall survival (OS), and disease-free survival (DFS). Multivariable models were adjusted for age, stage, Ki-67, and year of diagnosis. Statistical techniques included multiple imputation, propensity score matching, LASSO regression, and causal mediation analysis. Results: The cohort was highly admixed: 27% had ≥20% African ancestry (AA). Discordance was observed between genetic and social classifications: 48% of self-identified Black patients had <20% AA, while 24% of non-Black patients had ≥20% AA. Patients with ≥20% AA had higher tumor proliferation (median Ki-67 90% vs 80%) and higher pCR (50.0% vs 41.2%, p=0.42) but higher mortality (35.7% vs 25.4%). Black patients more often to use public healthcare (48.4% vs 15.9%, p<0.001) present with stage III disease (45.2% vs 35.6%), and receive less dose-dense chemotherapy (41.9% vs 76.1%, p<0.001). In adjusted analyses, neither AA nor self-reported race independently predicted pCR, OS, or DFS. Public healthcare use was the strongest independent predictor of worse OS (HR 2.45, 95%CI 1.12-5.36, p=0.025) and DFS (HR 2.18, 95%CI 1.08-4.40, p=0.03). Clinical stage (OR 0.66, p=0.022) and carboplatin use (OR 2.8, p=0.018) were the only independent predictors of pCR. Causal mediation analysis showed that 60.1% of the detrimental effect of public healthcare on pCR was mediated by advanced stage at presentation. Conclusions: In this pre-immunotherapy Brazilian cohort, TNBC outcome disparities are primarily driven by structural determinants related to healthcare access rather than genetic ancestry or self-reported race. While ancestry and race capture relevant biological and social dimensions, delayed diagnosis and suboptimal treatment delivery were the main mechanisms underlying outcomes. These findings underscore the need to address structural inequities and ensuring equitable access to timely, evidence-based cancer care in admixed populations. Future prospective studies integrating genetic ancestry, socioeconomic factors, and tumor immune profiling are needed.
Cerqueira et al. (Wed,) studied this question.