Infectious Complications in Patients with Hematologic Malignancies Receiving CD19 Vs. BCMA Targeted CAR-T Therapy

Background Infections result in significant morbidity and mortality among CAR-T recipients. While there are limited data on infections in patients with R/R B-cell HMs undergoing CAR-T, there is no evidence comparing infection rates between BCMA-targeted CAR-T for RRMM vs CD19 patients for other HMs. We examined and compared the infection patterns and related mortality in patients receiving CAR-T in a tumor-associated antigen (TAA)-specific manner. Methods A single center retrospective study including adults >18 years receiving CAR-T therapy for R/R B-cell HMs receiving CAR T-cells targeting CD19 and BCMA between September 2018 through March 2022. The primary outcome was non-relapse mortality (NRM) compared between CD19 vs BCMA. The secondary outcomes included infection density and cumulative incidence of overall, bacterial, viral, and fungal infections, stratified by TAAs. Results Among 107 eligible patients in the study cohort, 64% (n=68) received anti-CD19 and 36% (n=39) BCMA-directed CAR-T. With a median follow-up among survivors of 15 months, there were no differences in neutropenic fever events within the first 4 weeks between the 2 CAR TAA arms (59% in CD19+CAR-T vs 44% in BCMA+CAR-T, p=0.13). Rates of overall, bacterial, viral, and fungal infections were not significantly different. Although the NRM rates did not reach statistical significance, these were greater with anti-CD19 compared to BCMA-directed CAR-T at 6 (17% vs 8%), 12 (22% vs 8%), and 24 months (24% vs 12%) (p=.16). The cause-specific hazard of NRM was not statistically different between TAAs on multivariate analyses (CD19 vs BCMA HR 3.9, 95% CI, 0.44-34, p=0.22).There was no statistically significant difference in the infection density between BCMA (66 infection events) and CD19 CAR-T recipients (39 infection events), with an infection density of 0.28 events in BCMA vs 0.31 events in CD19 per 100 person-days (p=.75). The overall infection density significantly decreased over time across both TAA arms with 1.6 (95% CI, 1.12-2.29) events per 100 person-days in the first 30 days, dropping to 0.08 between D+366-D+730 (p<.0001). Multivariate analysis did not show a statistically significant difference between TAAs, antibiotic receipt in the first 30 days post-CAR-T, hypogammaglobulinemia at the time of CAR-T infusion, prior infection, or ICANS. However, CRS was independently associated with an increased risk of overall infections (HR 2.5; 95% CI, 1.05-5.7; p=.038). Hypogammaglobulinemia at the time of CAR-T infusion independently conferred an increased risk for viral infections (HR 2.6; 95% CI, 1.2-5.6; p=.013). Conclusions Our results suggest that anti-CD19 CAR-T is possibly associated with an overall heightened infection risk, resulting in an increased NRM, compared to BCMA-directed CAR-T, and aid in the development of risk-adapted and target antigen-specific infection prevention guidelines.