Disease outcomes by race in patients with high-risk triple negative breast cancer with residual disease after neoadjuvant chemotherapy: A post-hoc analysis of the EA1131 randomized clinical trial

586 Background: Black women with breast cancer have the highest breast cancer mortality (BCM) rates of any racial and ethnic group. The higher burden of triple negative breast cancer (TNBC) in Black women contributes to disparities in BCM, but the extent to which disparate outcomes exist in patients with similar tumor biology is uncertain. The objective of this study was to examine racial differences in disease outcomes and BCM between Black and White patients in a clinical trial population of high-risk TNBC with residual disease after neoadjuvant chemotherapy (NAC). Methods: From 2015-2021, the ECOG-ACRIN EA1131 clinical trial randomized 415 clinical stage II-III TNBC patients with residual disease after NAC to either adjuvant capecitabine or platinum. For this analysis, 366 patients with known self-reported Black or White race and follow-up data were included. The aim of this unplanned exploratory analysis was to examine locoregional recurrence (LRR), distant recurrence (DR), and BCM by race using Kaplan-Meier curves for unadjusted estimates and Cox modeling for adjusted analyses. Results: Racial distribution of the analysis population included 66 (18.0%) Black and 300 (82.0%) White patients. 239 (65.3%) patients presented with clinical stage II and 127 (34.7%) with stage III disease. All patients had residual pathological disease after NAC: 86 (23.5%) stage I, 176 (48.1%) stage II, and 102 (27.9%) stage III (n=2 unknown). Median age was 49 years in Black and 54 years in White patients, p=0.003. Disease presentation, response to chemotherapy, and treatment arm were similar by race. Median follow up was 20 months. 2-year estimated LRR occurred in 31.6% of Black and 18.6% of White patients although this was not statistically significant (p=0.387), and 2-year estimated DR and BCM were similar by race. Estimated 2-year outcomes are outlined. Models adjusted for race, age, stage, grade, treatment arm, and locoregional therapy did not identify race as an independent predictor of LRR, DR, or BCM. Higher pathologic stage was the only consistent independent predictor of recurrent disease or BCM (stage III with HR 6.14 for LRR; HR 5.16 for DR; HR 6.84 for BCM; all p<0.001). Conclusions: In this population of high-risk patients with residual TNBC after NAC with similar tumor biology treated on a clinical trial, there was a trend towards increased LRR in Black patients and similar DR and BCM between Black and White patients at 2 years of follow-up. This suggests that differences in BCM in other breast cancer populations may be at least partially driven by differences in tumor biology. Clinical trial information: NCT02445391 . [Table: see text]