BEGONIA: Phase 1b/2 study of durvalumab (D) combinations in locally advanced/metastatic triple-negative breast cancer (TNBC)—Initial results from arm 1, d+paclitaxel (P), and arm 6, d+trastuzumab deruxtecan (T-DXd).

1023 Background: Chemotherapy with immune checkpoint inhibitors can improve outcomes vs chemotherapy alone in patients (pts) with metastatic TNBC; however, many still have poor clinical outcomes. BEGONIA is an ongoing 2 part, multicenter, multiarm, open-label platform study evaluating safety and efficacy of D (anti–PD-L1)+P and D±P combined with novel therapies as first-line (1L) treatment for metastatic TNBC (NCT03742102). We report initial results from Part 1 of Arm 1, D+P, and Arm 6, D+T-DXd, an antibody-drug conjugate comprising an anti-HER2 antibody, tetrapeptide-based cleavable linker, and topoisomerase I inhibitor payload. Methods: Eligible pts had untreated unresectable locally advanced or metastatic TNBC. Pts with HER2-low–expressing tumors (IHC 2+/ISH–, IHC 1+/ISH–, or IHC 1+/ISH untested) per local testing were assigned to the D+T-DXd arm. Pts received D (1500 mg IV Q4W)+P (90 mg/m 2 IV Day 1, 8, 15 Q4W) in Arm 1 and D (1120 mg IV)+T-DXd (5.4 mg/kg IV) Q3W in Arm 6, until progression or unacceptable toxicity. Primary objectives are safety and tolerability. Secondary endpoints include objective response rate (ORR), duration of response (DoR), and progression-free survival (PFS). Tumors were assessed Q8W (D+P) or Q6W (D+T-DXd). The first 6 pts treated with D+T-DXd were evaluated for dose-limiting toxicities (DLTs), with additional pts enrolled if D+T-DXd was tolerated. Study arms are noncomparable due to differing eligibility criteria, treatments, and data maturity. Results: Arm 1 D+P (data cutoff Sep 2020): 23 pts received D+P (7 ongoing); 2 discontinued D+P due to AEs. Median follow-up time was 16.6 (range 8.5–19.8) mos. Any Grade 3/4 AEs and SAEs were experienced by 10 (44%) and 1 (4%) pts, respectively. D dose was delayed for 7 (30%) pts. Confirmed ORR was 13/23 (57%) with 54% of those remaining in response at 12 mos (median DoR not reached). Median PFS was 7.3 (95% CI 5.4–13.8) mos in the D+P arm. Arm 6 D+T-DXd (data cutoff Nov 2020): 11 pts received D+T-DXd to date (all ongoing). Median follow-up time was 2.3 (0–6) mos. Any Grade 3/4 AEs and SAEs were experienced by 4 (36%) and 1 (9%) pts, respectively. Pts who received D+T-DXd had no DLTs and 1 had a Grade 1 troponin increase. D dose was delayed and T-DXd dose reduced for 2 (18%) pts each. Confirmed ORR was 4/4 (100%; only 4 pts had the opportunity to complete 2 on-treatment disease assessments) with all 4 remaining in response at data cutoff (median DoR not reached). Conclusions: D+P demonstrated a tolerable safety profile and response rate as expected for a 1L TNBC IO/taxane combination. D+T-DXd showed promising early safety and efficacy in 1L HER2-low–expressing TNBC; pt evaluation and enrollment for D+T-DXd are ongoing. D+T-DXd data will be updated and the impact of PD-L1 expression in both arms will be examined. Clinical trial information: NCT03742102 .