A randomized phase III post-operative trial of platinum-based chemotherapy (P) versus capecitabine (C) in patients (pts) with residual triple-negative breast cancer (TNBC) following neoadjuvant chemotherapy (NAC): ECOG-ACRIN EA1131

605 Background: Pts with TNBC who have residual invasive disease (RD) after completion of NAC have a very high risk for recurrence, which is reduced by adjuvant capecitabine (C). Pre-clinical models support the use of platinum agents (P) in the TNBC basal subtype. EA1131 tested the hypothesis that invasive disease-free survival (iDFS) would not be inferior but improved in pts with basal subtype TNBC after NAC with the adjuvant use of a P instead of C (primary objective). Methods: Pts with clinical stage II/III TNBC post neoadjuvant taxane +/- anthracycline-based chemotherapy with at least 1 cm RD in the surgical specimen were randomized (1:1) to receive P (carboplatin or cisplatin once every 3 weeks for 4 cycles) or C (14/7d every 3 weeks for 6 cycles). TNBC subtype (basal vs. non-basal) was analyzed in the surgical specimen by PAM50. A non-inferiority design (non-inferiority margin of hazard ratio [HR] of 1.154) with superiority alternative (alternative HR of 0.754) was chosen, assuming a 4-year iDFS of 67% for the C arm. Non-inferiority was tested first. If non-inferiority was shown, a formal test for superiority of P compared to C would be conducted. Results: 401 participants were randomized to P or C between 2015 and 2020 (recruitment goal, 775), 310 (77%) had TNBC basal subtype disease (primary analysis population). Pts’ median age was 52 years, 71% were White and 19% Black. At diagnosis, most tumors were high grade (78%), T2 (59%), 47% N0, and 40% N1. Residual tumors were 37% ypT1, 44% ypT2, and 47% ypN0. Overall incidence of any toxicity was similar (83% with P, 80% with C), but grade 3 and 4 toxicities (no grade 5) were more common with P (25% vs 15%). After median follow-up of 18 months, 113 iDFS events (58% of full information) had occurred. 3-year iDFS for P arm was 40% (95%CI, 29%-51%) and 44% (95%CI, 32%-55%) for C arm. The HR for arms P/C was 1.09 (95% Repeated Confidence Interval, 0.62-1.90) and the probability of eventually rejecting the null of inferiority (i.e., conditional power) was 6%. The Data Safety and Monitoring Committee recommended stopping the trial at the 5th interim analysis in March 2021 since it was unlikely that the trial would be able to show non-inferiority or superiority of the P arm. Conclusions: Participants with TNBC with RD after NAC had a lower than expected 3-year iDFS regardless of study treatment. Available data show that it is very unlikely that the study would be able to establish non-inferiority of P to C. In addition, severe toxicities were more common with P. In pts with TNBC, particularly basal subtype, with at least 1 cm RD after NAC and high-risk of recurrence, adjuvant P use does not improve outcomes. Correlative analyses of RD tissue (NGS), circulating markers (ctDNA and CTC pre/post treatment), and patient-reported outcomes (PRO) questionnaires will now occur. Clinical trial information: NCT02445391.