Safety and efficacy of AZD0120, a BCMA/CD19 dual-targeting CAR T-cell therapy, in relapsed/refractory multiple myeloma: Preliminary Results from the DURGA-1 Phase 1b/2 study

Abstract Introduction: Multiple myeloma (MM) accounts for nearly 20% of all hematologic cancers in the US and remains a substantial clinical burden, with patients (pts) requiring multiple lines of therapy (LOT). While advances such as BCMA-directed CAR T-cell therapy have improved outcomes, challenges persist, including disease relapse, treatment toxicities (CRS, ICANS, and non-ICANS neurotoxicities), and access. AZD0120 (formerly GC012F) is a first-in-class autologous BCMA/CD19 dual-targeting CAR T-cell therapy using the FasTCAR rapid manufacturing platform that preserves the naive and central memory T-cell phenotypes with marked in vivo proliferative capacity. Phase 1 investigator-initiated trials in China demonstrated deep, durable responses and a favorable safety profile following single-infusion AZD0120 in newly diagnosed and relapsed/refractory MM (RRMM) (Du J, et al. EHA 2024 and ASCO 2023). Here, we report for the first time preliminary phase 1b results from DURGA-1, an ongoing phase 1b/2 clinical trial evaluating the safety and efficacy of AZD0120 in pts with RRMM. Methods: This open-label, single-arm, US-based, multicenter study (NCT05850234) evaluated the safety/tolerability of 2 dose levels (DL) of AZD0120 in pts with RRMM along with preliminary efficacy. Eligible ptswere aged ≥18 y with RRMM (3+ prior LOT [pLOT] including a proteasome inhibitor, immunomodulatory drug, and anti-CD38 antibody), ECOG PS 0–1, and documented evidence of progressive disease. Exposure to BCMA-directed therapy ≥6 mo with best response of PR or better was permitted. Pts underwent lymphodepletion followed by a single infusion of AZD0120. Pts in DL1 received 1x105 cells/kg; DL2 received 3x105 cells/kg. Dose escalation and recommended phase 2 dose (RP2D) were determined by a safety review committee. Phase 1b primary objectives included safety/tolerability and RP2D determination; secondary objectives included efficacy, cellular kinetics (CK), and pharmacodynamics. Results: As of data cutoff (DCO) on 18 July 2025, a total of 25 pts received infusion of AZD0120 (n=12 DL1; n=13 DL2). The median age was 64 y (range 44–78), median pLOT was 4 (range 3–7), 72% were triple-class refractory, 20% had prior BCMA CAR T-cell therapy, 4% had prior teclistamab, 28% had high-risk cytogenetic features [del(13q), del(17p13), t(4;14), t(14;16), amp(1q)], and 8% had extramedullary plasmacytomas. The median time from apheresis to release was 14 d (range 10–30). Median time from apheresis to infusion was 28 d (range 19–44) with 5 pts receiving bridging therapy. Median follow-up was 1.4 mo (range 0–19.3). No dose-limiting toxicities were reported for either dose. The most common treatment-emergent AEs (TEAEs, any grade) were CRS (64%), neutrophil count decreased (56%), and anemia (32%). The most common grade ≥3 TEAEs were neutrophil count decreased (52%), lymphocyte count decreased (32%), and white blood cell count decreased (32%). CRS was reported in 75% of pts at DL1 (all grade 1) and 54% at DL2 (46% grade 1; 8% grade 2), with no cases of grade ≥3 CRS. Median time to CRS onset (DL1/DL2) was 9 d (range 2–11), with a median duration of 2 d (range 1–4); 12 pts (48%) received tocilizumab for CRS management and 12% received dexamethasone. No cases of ICANS, related non-ICANS neurotoxicity, IEC-colitis, or secondary primary malignancies have been reported. There have been no deaths. For efficacy-evaluable patients (n=15), ORR was 100% (33% sCR, 47% VGPR, 20% PR). CR rates in evaluable pts were 30% in DL1 (n=10) and 40% in DL2 (n=5); median time to response was 0.9 mo for both DLs (range 0.9–1.9 DL1; 0.6–1.8 DL2). All MRD-evaluable pts (n=5 DL1; n=3 DL2; DCO 01 July 2025) were MRD-negative by NGS at a sensitivity of 10-5. Three pts in DL1 with ≥12 mo follow-up maintained MRD negativity. Consistent with robust in vivo expansion, CK analysis from 16 evaluable pts (DCO 12 June 2025) demonstrated a median Tmax of 13 d post-infusion with a median Cmax of 85,266 copies/mg gDNA. Median persistence was 42 d (range 13–273). Updated clinical data with additional follow-up will be presented. Conclusion: Preliminary phase 1b results demonstrated AZD0120 was well tolerated, with a low incidence of serious AEs, no grade ≥3 CRS, and no ICANS. FasTCAR-manufactured AZD0120 had expectedin vivo expansion, which may have resulted in the predictable CRS profile. A single infusion of AZD0120 achieved early, deep responses with 100% MRD negativity from 4L+ triple-class‒exposed pts with RRMM.