Duffy-null status and cytopenias following BCMA-directed CAR T therapy in multiple myeloma.

Abstract

7549 Background: Anti-BCMA chimeric antigen receptor T-cell (CAR T) therapies are highly effective in relapsed or refractory multiple myeloma (RRMM). The Duffy null phenotype, common among individuals of African ancestry and associated with lower baseline neutrophil counts, may influence hematologic toxicity following CAR T therapy. We evaluated outcomes of anti-BCMA CAR T by Duffy antigen status. Methods: We performed a retrospective analysis of 173 RRMM patients receiving ciltacabtagene autoleucel (cilta-cel; n=83) or idecabtagene vicleucel (ide-cel; n=90) between July 2021 and May 2025. Primary endpoints were PFS, OS, and response rate. Secondary endpoints included cytopenias and infections. Analyses were stratified by Duffy status and race/ethnicity and adjusted for age, prior therapies, and baseline ANC. Results: Cilta-cel: Duffy null patients (n=11, 13.3%) had comparable baseline characteristics to Duffy positive patients, except lower ANC (1.71 vs 2.44, p=0.01). Duffy null status was associated with prolonged neutropenia, with lower median ANC at day 30 (0.56 vs 1.77, p=0.005) and day 90 (1.06 vs 2.45, p=0.0002), and persistent grade ≥3 neutropenia at day 90 in 44.4% vs 4.3% (p=0.002). Infection rates were similar between groups (36.4% vs 41.7%, p=1.0). Response rates were high in both groups (100% vs 92%) and PFS did not differ significantly (HR 0.54, 95% CI 0.12–2.43, p=0.42), with a nonsignificant trend toward improved PFS among Duffy null patients after adjustment (HR 0.38, p=0.24). There were no significant differences in OS. Ide-cel: Duffy null status (n=16, 17.8%) was not associated with delayed cytopenia recovery (day 90 grade ≥3 neutropenia 33.3% vs 16.4%, p=0.2). Duffy null patients had higher infection rates (68.8% vs 39.2%, p=0.03) and a trend toward inferior OS (median 19.1 vs 40.4 months, log-rank p=0.11). These patients were more heavily pretreated (median 7 vs 5 prior lines, p=0.04); in multivariable analysis, prior lines of therapy predicted OS (HR 1.13, p=0.046) while Duffy status did not (HR 1.63, p=0.30). Response rates and PFS were similar between groups. Race/ethnicity alone was not associated with cytopenia recovery, infection rates, or survival. Conclusions: Duffy null status predicts severe, prolonged cytopenias following cilta-cel, without compromising efficacy or survival. Despite profound neutropenia, similar infection rates suggest benign cytopenia rather than true immunosuppression. Pre–CAR T Duffy phenotyping may prevent unnecessary neutropenia-directed interventions such as growth factors or stem cell boost. In ide-cel recipients, higher infection rates and inferior survival trends likely reflect heavier pretreatment and diminished marrow reserve. Future studies should evaluate Duffy status in earlier treatment lines and assess post-CAR T supportive care utilization by phenotype. Christen Dillard and Erneisha Brown contributed equally to this work.