Racial/ethnic disparities in locoregional recurrence in hormone-receptor positive node-negative breast cancer patients enrolled in the TAILORx trial.

Abstract

515 Background: Whether racial/ethnic disparities in locoregional recurrence (LRR) exist in patients (pts) with similar access to care treated on clinical trials is uncertain. We examined racial/ethnic differences in LRR in hormone-receptor positive HER2-negative (HR+HER2-) node-negative pts enrolled in the TAILORx trial. Methods: 10,273 pts age 18-75 were enrolled in TAILORx, which assigned pts with an OncotypeDx Recurrence Score (RS) <11 to endocrine therapy (ET) alone, those with RS >25 to chemotherapy + ET (CET), and randomized pts with RS 11-25 to ET or CET. Pts with unknown race/ethnicity (n=323) or incomplete treatment adherence information (n=1,168) were excluded from this analysis. Race/ethnicity was self-reported. LRR was defined as ipsilateral invasive in-breast, chest wall, or regional nodal recurrence without distant recurrence. Kaplan-Meier curves were used to estimate 8-year LRR. Cox proportional hazards analysis adjusted for clinical and treatment factors was used to determine factors associated with LRR. Results: 8,782 pts with T1-2N0 HR+HER2- breast cancer were included. Race/ethnicity was non-Hispanic White (NHW) in 6,932 (78.9%), non-Hispanic Black (NHB) in 629 (7.2%), Hispanic in 818 (9.3%), and Asian in 403 (4.6%). Treatment adherence was high across groups over time, with a 9.1% crossover in treatment arms. Average duration of ET was 63.8 +/- 0.3 months. Radiation therapy was planned in 96.0% of pts after breast conservation and 12.7% after mastectomy. At a median follow-up of 8 years, LRR rates were 1.9% in NHW, 4.2% in in NHB, 3.2% in Hispanic, and 3.9% in Asian pts (p<0.01). LRR rates broken down by RS are shown in the Table. On adjusted analyses, NHB and Asian (vs. NHW) pts were more likely to have LRR (HR 1.94 for NHB, HR 2.04 for Asian, p<0.05 for both). Additional statistically significant factors associated with LRR included age <50 (HR 1.85), T2 tumors (HR 1.43), higher grade (HR 2.30 for grade 3), and high RS (HR 3.13 for RS>25). Treatment receipt (chemotherapy, ET duration, and radiation) was not associated with LRR in this population. Conclusions: Racial/ethnic differences in LRR were seen in T1-2N0 HR+HER2- breast cancer pts enrolled in the TAILORx trial despite high rates of treatment adherence in this clinical trial population, with highest LRR rates in NHB and Asian pts. Further study in needed to understand racial/ethnic patterns in LRR by breast cancer subtype and if failure to rescue after LRR may contribute to differences in breast cancer mortality.[Table: see text]