Prospective randomized phase II trial to assess the efficacy and safety of neo-adjuvant olaparib/carboplatin (OC) in comparison to docetaxel/epirubicin/cyclophosphamide (TAC) in patients with early triple-negative breast cancer (TNBC) with homologous recombination deficiency (HRD): Primary results from the ABCSG 45 trial.

Abstract

510 Background: Carboplatin-based regimen are effective in patients (pts) with eTNBC, and olaparib improves the outcome of pts with BRCA1/2 pathogenic variants (PV), but the safety / efficacy of OC co-treatment in HRD-positive TNBC is unknown. ABCSG 45 (EU CT 2024-512821-10) is a prospective multicenter phase II study investigating the efficacy and tolerability of OC compared to conventional chemotherapy in HRD-positive eTNBC. Methods: Pts with HRD (Myriad genetics)-positive eTNBC were randomized to 6 cycles of olaparib (100 mg bid, days 4-16) / carboplatin (AUC 5) q3w, or 6 cycles of docetaxel/epirubicin/cyclophosphamide (75/50/500) q3w (TAC). In an initial dose-finding phase, 100 mg bid was identified as olaparib combination dose. Stratification factors were tumoral BRCA1/2 and menopausal status. Primary endpoint was centrally assessed residual cancer burden (RCB), pCR and QoL were secondary endpoints. Planned sample size was 90 pts, randomized 1:1 to achieve 80% power (two-sided alpha=0.05) to detect a RCB 0/I difference of 31%. Differences between treatment arms were assessed with a two-sided Cochran Mantel-Haenszel test using stratification factors. Pre-defined subgroup analysis was performed with logistic regression. Results: A total of 90 pts (OC: n=46; TAC: n=44), of whom 42 (47%) were BRCA1/2 PV carriers, were randomized between November 2019 and December 2023. Median age was 50.5 years (range 27.0-80.0). 40% had cT1, 55.6% cT2, and 4.4% cT3/4 tumors, and 60% of pts were clinically N0. 94.4% of tumors were G3, and Ki67 was >60% in 71.1%. Overall, the RCB0/I rate with OC was 52.2% vs. 70.5% with TAC (stratified risk difference = -18.8% (95%CI: -39.6% to 2.0%); p=0.068). In pts with BRCA1/2 PV, RCB0/I rates were comparable: 77.3% (OC) vs. 65.0% (TAC), while in 47 pts with BRCA1/2 wild type (WT), OC was significantly less effective: RCB0/I of 29.2% vs 73.9% in TAC (interaction p=0.008). pCR was achieved in 47.8% (OC) vs 59.1% (TAC; p=0.231). In pts with a BRCA1/2 PV, OC resulted in 77.3% pCR rate, vs. TAC 65.0%, in BRCA1/2 WT pts pCR was achieved by 20.8% (OC) vs. 56.5% (TAC) (interaction p=0.021). OC treatment resulted in more ≥ grade 3 hematologic toxicities with 30% vs 3% thrombocytopenia and 43% vs 18% neutropenia but caused fewer non-hematological toxicities. Conclusions: In this prospective randomized study in HRD-positive TNBC, 6 cycles of TAC resulted in strikingly high RCB0/I and pCR rates, independent of BRCA1/2 status. 6 cycles of OC achieved a pCR rate of >77% in BRCA1/2 PV but were less effective in pts with BRCA1/2 WT disease. These results may help to optimize neoadjuvant treatment strategies in TNBC. This research was conducted with support from AstraZeneca Austria GmbH. Clinical trial information: 2024-512821-10 .